ࡱ>  '` 0'bjbjLULU .,.?.?&zzz  8Ěnh[[[dqqq[ q[qq| ƚ!i, z0/8b8$8P.qdF^[[[[SY@Y Sentiment Strength Detection in Short Informal Text Mike Thelwall, Kevan Buckley, Georgios Paltoglou, Di Cai Statistical Cybermetrics Research Group, School of Computing and Information Technology, University of Wolverhampton, Wulfruna Street, Wolverhampton WV1 1SB, UK. E-mail: m.thelwall@wlv.ac.uk, K.A.Buckley@wlv.ac.uk, G.Paltoglou@wlv.ac.uk, caid@wlv.ac.uk Tel: +44 1902 321470 Fax: +44 1902 321478 Arvid Kappas School of Humanities and Social Sciences, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany E-mail: a.kappas@jacobs-university.de Tel: +49 421 200-3441 A huge number of informal messages are posted every day in social network sites, blogs and discussion forums. Emotions seem to be frequently important in these texts for expressing friendship, showing social support or as part of online arguments. Algorithms to identify sentiment and sentiment strength are needed to help understand the role of emotion in this informal communication and also to identify inappropriate or anomalous affective utterances, potentially associated with threatening behaviour to the self or others. Nevertheless, existing sentiment detection algorithms tend to be commercially-oriented, designed to identify opinions about products rather than user behaviours. This article partly fills this gap with a new algorithm, SentiStrength, to extract sentiment strength from informal English text, using new methods to exploit the de-facto grammars and spelling styles of cyberspace. Applied to MySpace comments and with a lookup table of term sentiment strengths optimised by machine learning, SentiStrength is able to predict positive emotion with 60.6% accuracy and negative emotion with 72.8% accuracy, both based upon strength scales of 1-5. The former, but not the latter, is better than baseline and a wide range of general machine learning approaches. Introduction Most opinion mining algorithms attempt to identify the polarity of sentiment in text: positive, negative or neutral. Whilst for many applications this is sufficient, texts often contain a mix of positive and negative sentiment and for some applications it is necessary to detect both simultaneously and also to detect the strength of sentiment expressed. For instance, programs to monitor sentiment in online communication, perhaps designed to identify and intervene when inappropriate emotions are used or to identify at-risk users (e.g., Huang, Goh, & Liew, 2007), would need to be sensitive to the strength of sentiment expressed and whether participants were appropriately balancing positive and negative sentiment. In addition, basic research to understand the role of emotion in online communication (e.g., Derks, Fischer, & Bos, 2008; e.g., Hancock, Gee, Ciaccio, & Lin, 2008; Nardi, 2005) would also benefit from fine-grained sentiment detection, as would the growing body of psychology and other social science research into the role of sentiment in various types of discussion or general discourse (Balahur, Kozareva, & Montoyo, 2009; Pennebaker, Mehl, & Niederhoffer, 2003; Short & Palmer, 2008). A complicating factor for online sentiment detection is that there are many electronic communications media in which text based communication in English seems to frequently ignore the rules of grammar and spelling. Perhaps most famous is mobile phone text language with its abbreviations, emoticons and truncated sentences (Grinter & Eldridge, 2003; Thurlow, 2003) but similar styles are evident in many other forms of computer mediated communication, including chatrooms, bulletin boards and social network sites (Baron, 2003; Crystal, 2006). Widely recognised innovations include emoticons like :-) that are reasonably effective in conveying emotion (Derks, Bos, & von Grumbkow, 2008; Fullwood & Martino, 2007) and word abbreviations like m8 (mate) and u (you) (Thurlow, 2003). Although sometimes seen as poor language use, these are a natural response to the technological affordances and social factors associated with a system (Baron, 2003; Walther & Parks, 2002). These variations cause problems because typical linguistic sentiment analysis programs start with part of speech tagging (e.g., Brill, 1992), which is reliant upon standard spelling and grammar, and/or apply rules that assume at least correct spelling, if not correct grammar. Spelling correction can be useful in this context, but this is based upon the assumption that spelling deviations are likely to be accidental mistakes (Kukich, 1992; Pollock & Zamora, 1984) and so current algorithms are unlikely to work well with deliberately non-standard spellings. Nevertheless, there is a range of common abbreviations and new words that a linguistic algorithm could, in principle, detect. Non-linguistic machine learning algorithms typically predict sentiment based upon occurrences of individual words, word pairs and word triples in documents. These may also perform poorly on informal text because of spelling problems and creativity in sentiment expression, even if a large training corpus is available (see below). The social network site MySpace, the source of the data used in the current study, is known for its young members, its musical orientation and its informal communication patterns (boyd, 2008; boyd, 2008). Probably as a result of these factors 95% of English public comments exchanged between friends contain at least one abbreviation from standard English (Thelwall, 2009). Common features include emoticons, texting-style abbreviations and the use of repeated letters or punctuation for emphasis (e.g., a loooong time, Hi!!!). Comments are typically short (mean 18.7 words, median 13 words, 68 characters) (Thelwall, 2009) but positive emotion is common (Thelwall, Wilkinson, & Uppal, 2010). This article proposes a new algorithm, SentiStrength, which employs several novel methods to simultaneously extract positive and negative sentiment strength from short informal electronic text. SentiStrength uses a dictionary of sentiment words with associated strength measures and exploits a range of recognised non-standard spellings and other common textual methods of expressing sentiment. SentiStrength was developed through an initial set of 2,600 human-classified MySpace comments, and evaluated on a further random sample of 1,041 MySpace comments. Note that in some articles, but not in emotion psychology, the term sentiment refers to affect split into positive, negative and neutral whereas the term emotion refers to more differentiated affect (e.g., happy, sad, frightened). In contrast, the two terms are used as synonyms here, with their meaning effectively defined by the coder instructions described below. The main novel contributions of this paper are: a machine learning approach to optimise sentiment term weightings; methods for extracting sentiment from repeated letter non-standard spelling in informal text; and a related spelling correction method. In addition, the paper introduces a dual 5-point system for positive and negative sentiment, a corpus of 1,041 MySpace comments for this system, and a new overall sentiment strength detection system that combines novel and existing methods. Background and Related Work This literature review section discussed related opinion mining/sentiment analysis research as well as some relevant contributions from emotion psychology. Opinion mining Opinion mining, also known as sentiment analysis, is the extraction of positive or negative opinions from (unstructured) text (Pang & Lee, 2008). The many applications of opinion mining include detecting movie popularity from multiple online reviews and diagnosing which parts of a vehicle are liked or disliked by owners through their comments in a dedicated site or forum. There are also applications unrelated to marketing, such as differentiating between emotional and informative social media content (Denecke & Nejdl, 2009). Opinion mining typically occurs in two or three stages, although more may be needed for some tasks (e.g., Balahur et al., 2010). First, the input text is split into sections, such as sentences, and each section tested to see if it contains any sentiment: if it is subjective or objective (Pang & Lee, 2004). Second, the subjective sentences are analysed to detect their sentiment polarity. Finally, the object about which the opinion is expressed may be extracted (e.g., Gamon, Aue, Corston-Oliver, & Ringger, 2005). Opinion mining normally deals with only positive and negative sentiment rather than discrete emotions (e.g., happiness, surprise), does not detect sentiment strength (but sometimes uses the strength of association of words with positive or negative sentiment, e.g., Kaji & Kitsuregawa, 2007), and does not simultaneously identify both positive and negative emotions. Nevertheless, such opinion mining research can aid the simultaneous assessment of positive and negative sentiment strength both because of its general insights into sentiment analysis and also because most techniques could, in theory, be repurposed for this new task. For example, phrase analysis techniques could be applied to identify both positive and negative sentiment even within individual sentences (Choi & Cardie, 2008; Wilson, 2008; Wilson, Wiebe, & Hoffman, 2009). Opinion mining algorithms often use machine learning to identify general features associated with positive and negative sentiment, where these features could be a subset of the words in the document, parts of speech or n-grams (i.e., the frequency of occurrence of all n consecutive words, where n is typically 1, 2, or 3) (Abbasi, Chen, Thoms, & Fu, 2008; Ng, Dasgupta, & Arifin, 2006; Tang, Tan, & Cheng, 2009). Other features used with some success include: emoticons in online movie reviews (Read, 2005), which seem so be more domain-independent than words; lexico-syntactic patterns (e.g., Riloff & Wiebe, 2003); and artificial features derived from adjective polarity lists (Ng et al., 2006). The additional features typically provide small but significant increases in performance. Rules-based methods have also been used to identify structures in sentences associated with sentiment (Prabowo & Thelwall, 2009; Wu, Chuang, & Lin, 2006). Two recurring machine learning issues are feature selection and classification algorithm choice. Feature selection, data processing to remove the least useful n-grams, has been shown to slightly improve classification performance, for example by choosing a restricted set of features (e.g., 5000) that score highest on a measure like information gain (Riloff, Patwardhan, & Wiebe, 2006), or log likelihood (Gamon, 2004). When using n-grams (and lexico-syntactic patterns) small improvements can also be made by pruning the feature set of features that are subsumed by simpler features that have stronger information gain values (Riloff et al., 2006). For example, if love has a much higher information gain value than I love then the bigram can be eliminated without much risk of loss of power for the subsequent classification. An entropy-weighted genetic algorithm can also perform better than standard feature reduction approaches (Abbasi, Chen, & Salem, 2008). In terms of classification algorithms, support vector machines (SVMs) are widely used (Abbasi et al., 2008; Abbasi et al., 2008; Argamon et al., 2007; Gamon, 2004; Mishne, 2005; Wilson, Wiebe, & Hwa, 2006) because they seem to perform as well or better than other methods in most machine learning contexts. Nevertheless, with a few exceptions (Read, 2005; Wilson et al., 2006), explicit comparisons with other methods have not been included in opinion mining publications. Many other approaches have also been used to detect sentiment in text. One is to have a dictionary of positive and negative words (e.g., love, hate), such as that found in General Inquirer (Stone, Dunphy, Smith, & Ogilvie, 1966), WordNet Affect (Strapparava & Valitutti, 2004), SentiWordNet (Baccianella, Esuli, & Sebastiani, 2010; Esuli & Sebastiani, 2006) or Q-WordNet (Agerri & Garca-Serrano, 2010), and to count how often they occur. Modifications of this approach include the identification of negating terms (Das & Chen, 2001), words that enhance sentiment in other words (e.g., really love, absolutely hate) and overall sentence structures (Turney, 2002). A more sophisticated approach is to identify text features that could potentially be subjective in some contexts and then use contextual information to decide whether they are subjective in each new context (Wiebe, Wilson, Bruce, Bell, & Martin, 2004). An alternative opinion mining technique has used a primarily linguistic approach: simple rules based upon compositional semantics (information about likely meanings of a word based upon the surrounding text) to detect the polarity of an expression (Choi & Cardie, 2008). This gives good results on phrases in newswire documents that are manually coded as having at least medium level positive or negative sentiment. This approach seems particularly suited to cases where there is a large volume of grammatically correct text from which rules can be learned. Nevertheless, a study of poor grammatical quality texts in online customer feedback showed that linguistic approaches could improve classification slightly when added to bag of words (1-grams) approaches, although aggressive feature reduction had a similar impact to adding linguistic features (Gamon, 2004). The improvement was probably due to the large data set available (40,884 documents with an average of 2.26 sentences each), as has been previously claimed for an analysis of informal text (Mishne, 2005). Another approach used a lexicon of appraisal adjectives (e.g., sort of, very) together with an orientation lexicon to detect movie review polarity. This did not perform as well as unigrams but the combined performance was better than that of unigrams alone (Argamon et al., 2007). Linguistic features have also been successfully used to extend opinion mining to a multi-aspect variant that is able to detect opinions about different aspects of a topic (Snyder & Barzilay, 2007). A promising future approach is the incorporation of context about the reasons why sentiment is used, such as differentiating between intention, arguments and speculation (Wilson, 2008). Detecting multiple emotions Psychology of emotion research argues that whilst positive and negative sentiment are important dimensions, there are many different widely socially-recognised types of emotion and the strength of emotions (arousal level) can vary (e.g., Cornelius, 1996; Fox, 2008). In the dimensional model of emotion from psychology (Russell, 1979), sentiment can always be fundamentally split into two axes: arousal (low to high) and valence (positive to negative). Whilst this model is useful, other research has shown that positive and negative sentiment can coexist (e.g., Fox, 2008, p. 127) and are relatively independent in many contexts particularly when sentiment levels are not extreme and over longer time periods (Diener & Emmons, 1984; Huppert & Whittington, 2003; Watson, 1988; Watson, Clark, & Tellegen, 1988) and so it also seems reasonable to conceive sentiment as separately-measureable positive and negative components, as encoded in a popular psychology research instrument (Watson et al., 1988). There have been some previous attempts to develop algorithms to detect the strength or prevalence of sentiment or emotion in text, or to differentiate between several types of emotion. The LIWC (Linguistic Inquiry and Word Count, www.liwc.net) software from psychology, for example, uses a list of emotion-bearing words to detect positive and negative emotion in text in addition to three specific emotions of particular use in psychology and psychotherapy: anger, anxiety and sadness. It uses simple word counting, measuring the proportion of words falling within an extensive predefined list (e.g., 408 positive and 499 negative words or word stems). The list includes some words that are associated with emotions but do not describe them. For example lucky is a positive keyword and loses is a negative keyword. In contrast to the machine learning approaches discussed above, these lists have been compiled and validated using panels of human judges and statistical testing. LIWC calculates the prevalence of emotion in text, rather than attempting to diagnose a texts overall emotion or emotion strength. It is most suited to longer documents, for which its statistics would be useful indicators of the tendency for emotion to occur. The program uses word truncation for simplicity (e.g., joy* matches any word starting with joy), rather than stemming or lemmatisation, but does not take into account booster words like very or the negating effect of negatives (e.g., not happy). LIWC has been used by psychology researchers to investigate the connection between language and psychology (Pennebaker et al., 2003) and also as a practical tool, for example to detect how well people are likely to cope with bereavement based upon their language use (Pennebaker, Mayne, & Francis, 1997). A related emotion detection approach differentiates between happy, unhappy and neutral states based upon words used by students describing their daily lives (Wu et al., 2006). This is similar to the typical positive/negative/neutral objective for opinion mining, however. One computer science initiative has attempted to identify various emotions in text, focussing on the six so-called basic emotions (Ekman, 1992; Fox, 2008) of anger, disgust, fear, joy, sadness and surprise (Strapparava & Mihalcea, 2008). This initiative also measured emotion strength. A human-annotated corpus was used with the coders allocating a strength from 0 to 100 for each emotion to each text (a news headline), although inter-annotator agreement was low (Pearson correlations of 0.36 to 0.68, depending on the emotion). A variety of algorithms were subsequently trained on this data set. For example, one used WordNet Affect lists to generate appropriate dictionaries for the six emotions. A second approach used a Naive Bayes classifier trained on sets of LiveJournal blogs annotated by their owners with one of the six emotions. The best system (for fine-grained evaluation) was one previously designed for newspaper headlines, UPAR7 (Chaumartin, 2007), which used linguistic parsing and tagging as well as WordNet, SentiWordNet and WordNet Affect, hence relying upon reasonably correct standard grammar and spelling. In psychology, the term mood refers to medium and long term affective states. Some blogs and social network sites allow members to describe their mood at the time of editing their status or writing a post, typically by selecting from a range of icons. The results can be used as annotated mood corpora. In theory such corpora ought to be usable to train classifiers to identify mood from the text associated with the mood icon and one system has been designed to do this, but with limited success, probably because the texts analysed are typically short (average 200 words) and there are many moods, some of which are very similar to each other, although even a binary categorisation task also had limited success (Mishne, 2005). A follow up project attempted to derive the proportion of posts with a given mood within a specific time period using 199 words (1-grams) and word pairs (2-grams) derived from the aggregate of all texts, rather than by classifying individual texts (Mishne & de Rijke, 2006). The results showed a high correlation with aggregate self-reported mood. A similar aggregation approach has been applied subsequently in a range of social science contexts (Hopkins & King, 2010). Linguistic processing has also been combined with a pre-existing large collection of subjective common sense statement patterns and applied to relatively informal and domain-independent text in email messages to detect multiple emotions (Liu, Lieberman, & Selker, 2003). This was part of an email support system, however, and the accuracy of the emotion detection was not directly evaluated. Sentiment strength detection In addition to the research discussed above concerning strength detection for multiple emotions (Strapparava & Mihalcea, 2008), there is some work on positive-negative sentiment strength detection. One previous study used modified sentiment analysis techniques to predict the strength of human ratings on a scale of 1 to 5 for movie reviews (Pang & Lee, 2005). This is a kind of sentiment strength evaluation with a combined scale for positive and negative sentiment. Experiments with human judgements led the authors to merge two of the categories and so the final task was a 4 category classification, with a 3 category version also constructed for testing purposes. A comparison of multi-class SVM classification with SVM regression suggested that SVM regression worked slightly better than multi-class SVM classification when all 4 categories were used but not when only 3 categories were used. It seems likely that the relative performance of SVM regression would increase further as the number of categories increases because the ordering of the classes is implicit information that the multi-class SVM does not use but that SVM regression does. Slight improvements were also gained when information about the percentage of positive sentences in each review was added. This may not be relevant to corpora of very short texts, however. Sentiment strength classification has also been developed for a three level scheme (low, medium, and high or extreme) for subjective sentences or clauses in newswire texts using a linguistic analysis converting sentences into dependency trees reflecting their structure (Wilson et al., 2006). Adding dependency trees to unigrams substantially improved the performance of various classifiers compared to unigrams alone, perhaps helped by the fairly large training set (9,313 sentences), the (presumably) good quality grammar of the texts, and the fairly low initial performance on this task (34.5% to 50.9% for unigrams, rising to 48.3% to 55.0% for the three types of classifier applied to level 1 clauses). Here, SVM regression was outperformed by both the rule-based learning Ripper (Cohen, 1995) and BoosTexter, a boosting algorithm combining multiple weak classifiers (Schapire & Singer, 2000). Quite similar to the current paper is one that measured multiple emotions and their strengths in informal text associated with a dialog system using a combination of methods, including seeking symbolic cues via repeated punctuation (e.g., !!), emoticons and capital letters as well as translating abbreviations (Neviarouskaya, Prendinger, & Ishizuka, 2007). The system also measured emotion intensity on a scale of 0-1 and used a dictionary of terms and intensity ratings assigned by three human judges (with moderate agreement rates: Fleiss Kappa 0.58). The reported evaluation on 160 human-coded sentences showed that in 68% of sentences the system agreed with the coder average to within 20%. Data Set and Human Judgement of Sentiment Strength MySpace was chosen as a source of test data for this study because it is a public environment containing a large quantity of informal text language and is important in its own right as one of the most visited web sites in the world in 2009. A random sample of MySpace comments was taken by examining the profiles of every 15th member that joined on June 18, 2007, up to 40,000 and selecting those with a declared U.S. nationality and a public profile not of a musician, comedian or film-maker. Of these, those with less than two friends or no comments were rejected as inactive and those with over 1,000 friends or 4,000 comments were rejected as abnormal. A commenting friend was then identified for each remaining member, satisfying the same criteria above, and a random comment selected from each direction of communication between the two. The comments were extracted in December 2008. This produced a large essentially random sample of U.S. commenter-commentee messages. Spam comments and chain messages were subsequently eliminated, as were comments containing images. Although sentiment analysis is normally concerned with opinions (Pang & Lee, 2008), Wilson (2008) has generalised this to the psychological task of identifying the author's hidden internal state from their text. For the MySpace data, the objective was not to determine opinions or the author's internal state, however, but to identify the role of expressed sentiment for online communication. Hence the focus of the task was to identify the sentiment expressed in each message, whether reflecting the author's hidden internal state, the intended message interpretation, or the reader's hidden internal state. In order to obtain reliable human judgements of a random sample of the MySpace comments, two pilot exercises were undertaken with separate samples of the data (a total of 2,600 comments). These were used to identify key judgement issues and an appropriate scale. Although there are many ways to measure emotion (Mauss & Robinson, 2009; Wiebe, Wilson, & Cardie, 2005), human coder subjective judgements were used as an appropriate way to gather sufficient results. A set of coder instructions was drafted and refined and an online system constructed to randomly select comments and present them to the coders. One of the key outcomes from the pilot exercise was that the coders treated expressions of energy as expressions of positive sentiment unless in an explicitly negative context. For example, Hey!!! would be interpreted as positive because it expresses energy in a context that gives no clue as to the polarity of the emotion, so it would be accepted by most coders as positive by default. In contrast, Loser!!! would be interpreted as more negative than Loser as the exclamation marks are associated with a negative word. Consequently, the instructions were revised to explicitly state that this conflation of ostensibly neutral energy and positive sentiment was permissible. For the final judgements, over a thousand MySpace comments in the data set (20 words and 101 characters per comment, on average) were selected to be judged on a 5 point scale as follows for both positive and negative sentiment. [no positive emotion or energy] 1 2 3 4 5 [very strong positive emotion] [no negative emotion] 1 2 3 4 5 [very strong negative emotion] The coders were given verbal instructions for coding each text as well as a booklet explaining the task (motivated by Wiebe et al., 2005), with the key instructions reproduced in this articles appendix. The booklet also contained a list of emoticons and acronyms with explanations and background context of the task for motivation purposes. An early version of the booklet included examples of comments with associated positive and negative sentiment judgements but these had little impact in practice on coders during the pilot testing phase. The set of examples was therefore not used so that inter-coder reliability could be more realistically assessed without the possibility that some of the comments were too similar to the examples given. Emotions are perceived differently by individuals, partly because of their life experiences and partly because of personality issues (Barrett, 2006) and gender (Stoppard & Gunn Gruchy, 1993). For system development, the judgements should give a consistent perspective on sentiment in the data, rather than an estimate of the population average perception. As a result, a set of same gender (female) coders was used and initial testing conducted to identify a homogeneous subset. Five coders were initially selected but two were subsequently rejected for giving anomalous results: one gave much higher positive scores than the others, and another gave generally inconsistent results. The mean of the three coders results was calculated for each comment and rounded. This was the gold standard for the experiments. Below are some examples of texts and judgements. hey witch wat cha been up too (scores: +ve: 2,3,1; -ve: 2,2,2) omg my son has the same b-day as you lol (scores: +ve: 4,3,1; -ve: 1,1,1) HEY U HAVE TWO FRIENDS!! (scores: +ve: 2,3,2; -ve: 1,1,1) What's up with that boy Carson? (scores: +ve: 1,1,1; -ve: 3,2,1) Table 1 reports the degree of inter-coder agreement. Basic agreement rates are reported here for comparability with SentiStrength. Previous emotion-judgement/annotation tasks have obtained higher inter-coder scores, but without strength measures and therefore having fewer categories (e.g., Wiebe et al., 2005). Moreover, one previous paper noted that inter-coder agreement was higher on longer (blog) texts (Gill, Gergle, French, & Oberlander, 2008), suggesting that obtaining agreement on the short texts here would be difficult. The appropriate type of inter-coder reliability statistic for this kind of data with multiple coders and varying differences between categories is Krippendorffs  (Artstein & Poesio, 2008; Krippendorff, 2004). Using the numerical difference in emotion score as weights, the three coder  values were 0.5743 for positive and 0.5634 for negative sentiment. These values are positive enough to indicate that there is broad agreement between the coders but not positive enough (e.g., < 0.67. although precise limits are not applicable to Krippendorff s  with weights) to suggest that the coders are consistently measuring a clear underlying construct. Nevertheless, using the average of the coders as the gold standard still seems to be a reasonable method to get sentiment strength estimates. Table 1. Level of agreement between coders for the 1,041 evaluation comments (exact agreement, % of agreements within one class, mean percentage error, and Pearson correlation). Comparison+ve+ve +/- 1 class+ve mean % diff.+ve corr-ve-ve +/- 1 class-ve mean % diff.-ve corrCoder 1 vs. 251.0%94.3%.256.56467.3%94.2%.208.643Coder 1 vs. 355.7%97.8%.216.67776.3%95.8%.149.664Coder 2 vs. 361.4%95.2%.199.68268.2%93.6%.206.639The SentiStrength Sentiment Strength Detection Algorithm The SentiStrength emotion detection algorithm was developed on an initial set of 2,600 MySpace classifications used for the pilot testing. The key elements of SentiStrength are listed below. The core of the algorithm is the sentiment word strength list. This is a collection of 298 positive terms and 465 negative terms classified for either positive or negative sentiment strength with a value from 2 to 5. The default classifications are based upon human judgements during the development stage, with automatic modification occurring later during the training phase (see below). Following LIWC, some of the words include wild cards (e.g., xx*) matches any number e"2 of consecutive xs. Some terms are standard English words and others are non-standard but common in MySpace (e.g., luv, xox, lol, haha, muah). The emotion strength is specific to the contexts in which the words tend to be used in MySpace. For example,  love was originally classified as strength 4 positive but was reduced to strength 3 due to many casual uses such as Just showin love 2 ur page. Some of the words explicitly express emotion, such as love or hate but others, normally given a weak strength 2, are indirectly associated with positive or negative contexts (e.g., appreciate, help, birthday). The SentiStrength algorithm includes procedures (described below) to fine-tune the sentiment strengths using a set of training data. The above default manual word strengths are modified by a training algorithm to optimise the sentiment word strengths. This algorithms starts with the baseline human-allocated term strengths for the predefined list and then for each term assesses whether an increase or decrease of the strength by 1 would increase the accuracy of the classifications. Any change that increases the overall accuracy by at least 2 is kept. The minimum increase could also be set to 1 which would risk over-fitting, whereas 2 risks loosing useful changes to rare words. Here 2 was selected to make the algorithm run faster, due to less changes, rather than for any theoretical reason (in fact the algorithm worked better on the test data with 1, as the results show). The algorithm tests all words in the sentiment list at random and is repeated until all words have been checked without their strengths being changed. The word miss was allocated a positive and negative strength of 2. This was the only word classed as both positive and negative. It was typically used in the phrase I miss you, suggesting both sadness and love. A spelling correction algorithm identifies the standard spellings of words that have been miss-spelled by the inclusion of repeated letters. For example hellllloooo would be identified as hello by this algorithm. The algorithm (a) automatically deletes repeated letters above twice (e.g., helllo -> hello); (b) deletes repeated letters occurring twice for letters rarely occurring twice in English (e.g., niice -> nice) and (c) deletes letters occurring twice if not a standard word but would form a standard word if deleted (e.g., nnice -> nice but not hoop -> hop nor baaz -> baz). Formal spelling correction algorithms (see Pollock & Zamora, 1984) were tried but not used as they made very few corrections and had problems with names and slang. A booster word list contains words that boost or reduce the emotion of subsequent words, whether positive or negative. Each word increases emotion strength by 1 or 2 (e.g., very, extremely) or decreases it by 1 (e.g., some). A negating word list contains words that invert subsequent emotion words (including any preceding booster words). For example, if very happy had positive strength 4 then not very happy would have negative strength 4. The possibility that some negating terms do not negate was not incorporated as this did not seem to occur often in the pilot data set. Repeated letters above those needed for correct spelling are used to give a strength boost of 1 to emotion words, as long as there are at least two additional letters. The use of repeated letters is a common device for expressing emotion or energy in MySpace comments, but one repeated letter often appeared to be a typing error. An emoticon list with associated strengths (positive or negative 2) supplements the sentiment word strength list (and punctuation included in emoticons is not processed further for the purposes below). Any sentence with an exclamation mark was allocated a minimum positive strength of 2. Repeated punctuation including at least one exclamation mark gives a strength boost of 1 to the immediately preceding emotion word (or sentence). Negative emotion was ignored in questions. For example, the question are you angry? would be classified as not containing sentiment, despite the presence of the word angry. This was not applied to positive sentiment because many question sentences appeared to contain mild positive sentiment. In particular, sentences like whats up? were typically classified as containing mild positive sentiment (strength 2). The above factors were applied separately to each sentence, with the sentence being assigned with both the most positive and most negative emotion identified in it. Each overall comment was assigned with the most positive of its sentence emotions and the most negative of its sentence emotions. Sentence were split either by line breaks in comments or after punctuation other than emoticons. Some additional modifications were added to SentiStrength but subsequently rejected after additional testing, or were found to be impractical. Phrase identification was not extensively used except for a few frequent examples found in the initial 2,600 development comments. Although idiomatic phrases were common, their variety was such that it did not seem practical to systematically identify them. Future work could perhaps identify booster phrases like so much and a lot, and use phrase identification to separate weak uses of the word love with stronger uses, such as I love you. Semantic disambiguation was not used for ambiguous words because of the problems caused by highly non-standard grammar. This could potentially improve the algorithm but would require considerable computational effort. For example, the word rock was sometimes strongly positive (e.g., you rock!!!) and sometime neutral (e.g., do you listen to rock music?). Experiments SentiStrength was tested on a set of 1,041 MySpace comments that were different from the comments used in the development phase and were classified by three people (see Table 1), and the average was used as the gold standard. A 10-fold cross-validation approach was used. The results were compared to random allocation and to the baseline majority class classification (a positive sentiment of 2 and a negative sentiment of 1). SentiStrength was also compared to a range of standard machine learning classification algorithms in Weka (Witten & Frank, 2005) using the frequencies of each word in the sentiment word list as the feature set. The extended feature set used for the comparisons included n-grams of length 1-3 consisting of all terms extracted from the text, including emoticons, spelling-corrected words (where appropriate), repeated punctuation, question marks and exclamation marks (e.g., one feature was the 3-gram: love-u-!) as well as counts of the total number of 1, 2, and 3-grams in each comment. This extended set of features incorporates most of the elements of text used by SentiStrength. A second test compared different feature sets to see whether alternative smaller feature sets could give better results for machine learning and to discover which features were most useful. A third test used feature reduction with subsumption (see below for details). A fourth test compared different variations of SentiStrength to see which aspects of the algorithm were most powerful. Comparison with machine learning, extended feature set Figures 1 and 2 show the performance of various machine learning algorithms on the 1,041 MySpace comments with different feature set sizes, as selected using the top-ranking features from the information gain metric. Feature selection improved the results for all methods, with one minor exception (Nave Bayes for positive sentiment: 52.0% without feature selection, averaged over 4 10-fold cross-validations). For each method, Table 2 reports comparisons with SentiStrength using the optimal feature set size for each method.  Fig. 1. Positive sentiment classification accuracy against feature set size for different classifiers using the extended feature set; average over 4 classifications.  Fig. 2. Negative sentiment classification accuracy against feature set size for different classifiers using the extended feature set; average over 4 classifications. From Table 2, machine learning classifiers using the extended feature set with the optimal number of features, as selected by information gain, are significantly less accurate than SentiStrength. SentiStrength also has the highest correlation with the gold standard, the lowest mean percentage error and the highest accuracy to within one class. Hence it performs consistently better (at least 2.1%) than the other algorithms. The level of accuracy for SentiStrength is nevertheless moderate at 60.6%. This is similar to the degree of agreement between the human coders (Table 1), suggesting that positive sentiment strength detection in informal short texts is an inherently difficult task. Table 2. Performance of various algorithms on positive sentiment strength detection for 1,041 comments with the extended feature set and 10-fold cross-validation (decreasing order of positive sentiment strength performance). Other than SentiStrength, results are averages over 4 runs of different random test/training splits and for the optimal feature numbers, as selected from Figure 1. AlgorithmOptimal featuresAccuracyAccuracy +/- 1 classCorr.Mean % absolute errorSentiStrength (standard configuration, 30 runs)-60.6%96.9%.59922.0%Simple logistic regression70058.5%96.1%.55723.2%SVM (SMO)80057.6%95.4%.53824.4%J48 classification tree70055.2%95.9%.54824.7%JRip rule-based classifier70054.3%96.4%.47628.2%SVM regression (SMO)10054.1%97.3%.46928.2%AdaBoost10053.3%97.5%.46428.5%Decision table20053.3%96.7%.43128.2%Multilayer Perceptron10050.0%94.1%.42230.2%Nave Bayes10049.1%91.4%.56727.5%Baseline-47.3%94.0%-31.2%Random-19.8%56.9%.01682.5%Bold=sig at 0.01, italic=sig at 0.05 compared to SentiStrength. For negative sentiment strength, most of the methods give quite similar results and some give better results than SentiStrength. Although the SentiStrength accuracy is 72.8%, this is only 2.9% better than the baseline, several of the other methods have similar levels of accuracy and SVM is significantly more accurate. SentiStrength is significantly the most accurate of the methods if up to one class error is allowed, and has significantly the highest correlation with the human coder results. Note that in theory none of the methods ought to be worse than the baseline but this can occur due to optimisation on the training set rather than the evaluation set. Overall, it seems that SentiStrength is not good at identifying negative emotion but that this is a hard task for the short texts analysed here. Note also that the mean percentage absolute error for the random category is over 100% due to the predominance of 1 as the correct category for negative sentiment. Table 3. Performance of various algorithms on negative sentiment strength detection for 1,041 comments with the extended feature set and 10-fold cross-validation (decreasing order of positive sentiment strength performance). Other than SentiStrength, results are averages over 4 runs and for the optimal feature numbers, as selected from Figure 2. AlgorithmOptimal featuresAccuracyAccuracy +/- 1 classCorr.Mean % absolute errorSVM (SMO)10073.5%92.7%.42116.5%SVM regression (SMO)30073.2%91.9%.36317.6%Simple logistic regression80072.9%92.2%.36417.3%SentiStrength (standard configuration, 30 runs)-72.8%95.1%.56418.3%Decision table10072.7%92.1%.34617.0%JRip rule-based classifier50072.2%91.5%.30917.3%J48 classification tree40071.1%91.6%.23518.8%Multilayer Perceptron10070.1%92.5%.34620.0%AdaBoost10069.9%90.6%-16.8%Baseline-69.9%90.6%-16.8%Nave Bayes20068.0%89.8%.31127.3%Random-20.5%46.0%.010157.7% Bold=sig at 0.01, italic=sig at 0.05 compared to SentiStrength. The remainder of the paper focuses on positive sentiment alone, since the results for negative sentiment are not significant. Comparison of feature sets for machine learning positive sentiment strength Figures 3 and 4 compare the impact of using different feature sets with the two best-performing algorithms for positive sentiment strength detection. The feature sets are: 1-3-grams; 1-3-grams with emoticons; 1-3-grams with punctuation; 1-3-grams with misspellings (i.e., including terms before spelling correction in addition to terms after spelling correction, when different); 1-3-grams with emoticons, punctuation and misspellings; 1-3-grams with emotion terms; and 1-grams. The basic bag or words approach (1-grams) performs poorly always the worst feature set for logistic regression and the worst or amongst the worst few feature sets all the time for SVM. For SVM, the best results are achieved with the basic 1-3-grams enhanced by the emotion terms, although most of the time (i.e., for 500-1000 features) the extended feature set (labelled all of the above and the same as used in the results above) performs best, perhaps mainly due to the punctuation component, since this enhancement performs second best for 700-1000 features.  Fig. 3. SVM (SMO) positive sentiment classification accuracy against feature set size for different feature set types; average over 4 classifications. Figure 4 suggests that, other than the basic bag of words, the difference between feature sets is less clear-cut for logistic regression than for SVM but the best performing combination is again the 1-3 grams plus emotion terms. For larger feature sets, the combined feature set performed best, probably due to the punctuation and emotion terms.  Fig. 4. Logistic regression positive sentiment classification accuracy against feature set size for different feature set types; average over 4 classifications. A potential weakness of using bigrams and trigrams in conjunction with unigrams is that there is some redundancy involved. For instance, the trigram I love you will also match the bigrams I love and love you as well as the unigrams I, love and you. In response, subsumption is a feature selection method that eliminates bigrams and trigrams that appear to be redundant in the sense of not giving additional information above that of their constituent unigrams (and bigrams for trigrams). This approach is appropriate here. Subsumption was applied with a logical extension: that word patterns, like happ* could eliminate matching words (e.g., happily, happy in this case) if the appropriate measure was matched. Figures 5 and 6 show the results of subsumption for the two machine learning algorithms for which it performed best: SVM and logistic regression. Subsumption performs best in conjunction with feature reduction, as both graphs show. For the other algorithms, subsumption improved the performance of Jrip by 0.4% ( = 0.005, 100 features), SVM regression by 1.1% ( = 0.02, 100 features), multilayer perceptron by 1.0% ( = 0.02, 100 features) and decision table by 1.0% ( = 0.005, 900 features) but did not improve J48, AdaBoost and Nave Bayes.  Fig. 5. SVM(SMO) positive sentiment classification accuracy against feature set size for subsumption with various  values; average over 5 classifications.  Fig. 6. Logistic regression positive sentiment classification accuracy against feature set size for subsumption with various  values; average over 5 classifications. From Figure 5, SVM with subsumption outperforms SVM without subsumption on the extended feature set by 1.8%, and outperforms SVM on all the other feature sets ( = 0.02, 500 features). Nevertheless, its accuracy is lower than the SentiStrength standard version, although the difference is not statistically significant (accuracy = 59.42%, accuracy +/-1 = 96.60%, correlation = 0.5822, mean absolute error = 22.65%; only the mean absolute error difference is statistically significant from SentiStrength standard configuration). From Figure 6, logistic regression with subsumption outperforms logistic regression without on the extended feature set by a lower margin of 0.7% ( = 0.01, 200 features). It performs less well than 1-3grams with the emotion terms added, however (Figure 4), but this could be a statistical anomaly due to the large number of comparisons performed. Logistic regression performs less well than standard SentiStrength, but the difference is again not significant (accuracy = 59.23%, accuracy +/-1 = 95.79%, correlation = 0.5820, mean absolute error = 22.57%; all except accuracy are statistically significantly different from SentiStrength standard configuration). In terms of  values, 0.02 tends to perform almost uniformly better than other values for this data set. Note that that although SentiStrength is not statistically significantly better than the optimal SVM and logistic regression models using subsumption, the optimal variation of SentiStrength in Table 4, with one simple modification (training needs only increase of 1 to alter word strengths), is statistically significantly better in all respects than SVM and is statistically significantly better in all respects, except accuracy within +/-1, than logistic regression. Comparison of SentiStrength versions Tables 4 and 5 report comparisons of different variations of SentiStrength. Most variations have little influence on the results individually accounting for a maximum of 0.8% of the performance of the algorithm, except for the last two options. These differences are small enough to be attributable to the corpus used and so the table does not provide convincing evidence that any of the variations are better or worse than the standard approach. When removing all the options (but not changing the averaging method) the cumulative effect is more significant, however, reducing performance by 3.4%. Perhaps comments using non-standard features tend to use multiple non-standard features and so if one special rule is ignored then this is frequently compensated for by the other special rules. Compared with tables 2 and 3, the main power of SentiStrength is in the combined effect of its rules to adapt to various informal text variations as well as in the overall approach of using a list of term strengths and identifying the strongest positive and negative terms in any comment. In this context, it seems that the generic classification algorithms in Table 2 were a minimum of 2.1% less effective than SentiStrength mainly due to the 1-3 grams approach being insufficiently flexible to cope with non-standard MySpace language (about 3.4% attributable to this cause). In addition, it seems that they were not able to draw upon a large enough training set to learn effective term strengths and a much larger training set could see some of them approach closer to the performance of SentiStrength. Finally, note that the variations of SentiStrength that apparently improve it are not robustly better: when all these are combined to make a new version of SentiStrength this has exactly the same accuracy as the standard configuration (60.64% correct, 97.07% +/- 1 class, .6071 correlation, 21.62% mean % error). Table 4. Comparison of the positive emotion performance over several algorithm variations: average over 30 10-fold cross-validations for 1,041 classified comments. Type% Correct+/- 1 classcorr.Mean % err. (pred-act)/actSentiStrength standard algorithm (but training needs only increase of 1 to alter word strengths)61.03%96.68%.598321.66%Negating words not used to switch following sentiment (e.g., not happy)60.87%97.50%.620621.28%Multiple consecutive positive words not used as emotion boosters 60.70%96.88%.596221.97%Emoticons ignored60.68%96.87%.597721.95%Booster words ignored (e.g., very)60.68%96.80%.597022.14%SentiStrength standard algorithm60.64%96.90%.598621.96%Exclamation marks not given a strength of 260.51%96.62%.603521.47%Automatic spelling correction disabled60.39%96.88%.596122.05%Extra multiple letters not used as emotion boosters60.21%96.81%.595222.16%The term miss not given a strength of +260.45%96.77%.595322.16%Idiom lookup table disabled60.52%96.88%.605421.62%Neutral words with emphasis not counted as positive emotion60.13%96.79%.596621.90%SentiStrength with all the above changes57.44%96.07%.607321.91%Sentence sentiment is the average of all term sentiments (rather than the maximum)42.40%88.54%.406529.27%Text sentiment is the average of all sentence sentiments (rather than the maximum)39.13%86.96%.329333.19%* Bold=significant at p=0.01, italic=sig. at p=0.05, compared to the standard algorithm. Table 5. Comparison of the negative emotion performance over several algorithm variations: average over 30 10-fold cross-validations for 1,041 classified comments. Type% Correct+/- 1 classcorr.Mean % err. (pred-act)/actNegative sentiment in questions is not ignored73.56%95.14%.592118.11%SentiStrength standard algorithm (but training needs only increase of 1 to alter word strengths)72.95%94.86%.565118.16%Negating words not used to switch following sentiment (e.g., not happy)72.84%94.79%.570618.35%SentiStrength standard algorithm72.83%95.07%.564418.27%Multiple consecutive negative words not used as emotion boosters72.81%95.08%.565318.29%Emoticons ignored72.80%94.97%.561418.28%SentiStrength with all the changes in this table except averaging72.76%94.59%.566819.07%Idiom lookup table disabled72.73%95.03%.555618.63%Extra multiple letters not used as emotion boosters72.72%95.04%.562718.40%Text sentiment is the average of all sentence sentiments (rather than the maximum)72.66%95.83%.548616.81%Automatic spelling correction disabled72.64%95.07%.558618.62%Booster words ignored (e.g., very)72.35%95.03%.555918.50%Sentence sentiment is the average of all term sentiments (rather than the maximum)72.17%95.35%.498016.82%* Bold=significant at p=0.01, italic=sig.t at p=0.05, compared to the standard algorithm. Table 5 shows that there is very little variation in the performance of the different variations of SentiStrength for negative emotion strength detection: the performance differs from the standard configuration by a maximum of 0.83%. It suggests however, that negative sentiment in questions (e.g., Do you hate Tony?) should not be ignored in future. Discussion and Conclusions Recall that the main novel contributions of this paper are: a machine learning approach to optimise sentiment term weightings; methods for extracting sentiment from non-standard spelling in text; and a related spelling correction method. SentiStrength was able to identify the strength of positive sentiment on a scale of 1 to 5 in 60.6% of the time in informal MySpace language, significantly above the best standard machine-learning approaches which had a performance of up to 58.5% - in line with those for a previous 4-category opinion intensity classification task (Wilson et al., 2006). The standard version of SentiStrength was also better then standard machine learning methods when their performance was improved (or not, in some cases) with the use of subsumption and information gain feature reduction, but the difference was not statistically significant. A slightly modified version of SentiStrength was statistically significantly better than the improved machine learning methods, however. This is good evidence of the efficacy of SentiStrength for positive sentiment strength detection given the range of different algorithms and parameters that it was compared against (9 algorithms x 11 feature set sizes, x 7 feature set types = 693 variations, plus 9 algorithms x 10 feature set sizes x 3  values = 270 variations for subsumption), which gives lower-performing algorithms a reasonable statistical chance of outperforming SentiStrength through chance, but none did. The main reason for SentiStrengths relative success seems to be procedures for decoding non-standard spellings and methods for boosting the strength of words, which accounted for much of its performance. Without these factors, the SentiStrength variant based solely upon a dictionary of emotion-associated words and their estimated strengths with 57.5% was only 1.3% better than the most successful machine learning approach on an extended set of 1-3grams. In contrast, SentiStrength was able to identify negative sentiment little better (1.8%) than the baseline, probably due to creativity in expressing negative comments or due to the difficulty in getting significantly above the baseline when one category dominates (Artstein & Poesio, 2008; Krippendorff, 2004). It seems that both positive and negative sentiment detection in informal text language like MySpace comments is challenging because of several factors: language creativity, expressions of sentiment without emotion-bearing words, and differences between human coder interpretations meaning that there is not a genuinely correct classification for most comments. Given the success in generating an algorithm for positive sentiment strength detection and the predominance of positive sentiment in MySpace comments, it seems that future research can apply the sentiment strength detection techniques to automatically identify and classify positive sentiment in informal web communication environments on a large scale. Moreover, there are many commercial applications of sentiment analysis, some of which use informal computer text generate from chatrooms or mobile phone text messages, and this algorithm shows that it is possible to estimate the strength of positive sentiment even in these short messages. In terms of future work, a next logical step is to attempt to improve the performance of the system through linguistic processing, despite the poor grammar of the short informal text messages analysed. Previous work has shown that this approach is promising, particularly via dependency trees (Wilson et al., 2009) and that, given a large enough training sample, improvements may be possible even in poor quality text (Gamon, 2004). Appendix: Coder Instructions (extract) Code each comment for the degree to which it expresses positive emotion or energy. Excitement, enthusiasm or energy should be counted as positive emotion here. If you think that the punctuation emphasises the positive emotion or energy in any way then include this in your rating. The scale for positive emotion or energy is: [no positive emotion or energy] 1 2 3 4 5 [very strong positive emotion] Allocate 1 if the comment contains no positive emotion or energy. Allocate 5 if the comment contains very strong positive emotion. Allocate a number between 2 and 4 if the comment contains some positive emotion but not very strong positive emotion. Use your judgement about the exact positive emotion strength. Code each comment for the degree to which it expresses negative emotion or is negative. If you think that the punctuation emphasises the negative emotion in any way then include this in your rating. The scale for negative emotion is: [no negative emotion] 1 2 3 4 5 [very strong negative emotion] Allocate 1 if the comment contains no negative emotion at all. Allocate 5 if the comment contains very strong negative emotion. Allocate a number between 2 and 4 if the comment contains some negative emotion but not very strong negative emotion. Use your judgement about the exact negative emotion strength. When making judgements, please be as consistent with your previous decisions as possible. Also, please interpret emotion within the individual comment that it appears and ignore all other comments. References Abbasi, A., Chen, H., & Salem, A. (2008). Sentiment analysis in multiple languages: Feature selection for opinion classification in web forums. ACM Transactions on Information Systems, 26(3), 12:11-12.34. Abbasi, A., Chen, H., Thoms, S., & Fu, T. (2008). Affect analysis of Web forums and Blogs using correlation ensembles. IEEE Transactions on Knowledge and Data Engineering, 20(9), 1168-1180. Agerri, R., & Garca-Serrano, A. (2010). Q-WordNet: Extracting polarity from WordNet senses. Proceedings of the Seventh conference on International Language Resources and Evaluation, Retrieved May 25, 2010 from: http://www.lrec-conf.org/proceedings/lrec2010/pdf/2695_Paper.pdf. Argamon, S., Whitelaw, C., Chase, P., Hota, S. R., Garg, N., & Levitan, S. (2007). Stylistic text classification using functional lexical features. Journal of the American Society for Information Science and Technology, 58(6), 802-822. Artstein, R., & Poesio, M. (2008). Inter-coder agreement for computational linguistics. Journal of Computational Linguistics, 34(4), 555-596. Baccianella, S., Esuli, A., & Sebastiani, F. (2010). SentiWordNet 3.0: An enhanced lexical resource for sentiment analysis and opinion mining. Proceedings of the Seventh conference on International Language Resources and Evaluation, Retrieved May 25, 2010 from: http://www.lrec-conf.org/proceedings/lrec2010/pdf/2769_Paper.pdf. Balahur, A., Kozareva, Z., & Montoyo, A. (2009). Determining the polarity and source of opinions expressed in political debates. Lecture Notes in Computer Science, 5449, 468-480. Balahur, A., Steinberger, R., Kabadjov, M., Zavarella, V., Goot, E. v. d., Halkia, M., et al. (2010). Sentiment analysis in the news. Proceedings of the Seventh conference on International Language Resources and Evaluation, Retrieved May 25, 2010 from: http://www.lrec-conf.org/proceedings/lrec2010/pdf/2909_Paper.pdf. Baron, N. S. (2003). Language of the Internet. In A. Farghali (Ed.), The Stanford Handbook for Language Engineers (pp. 59-127). Stanford: CSLI Publications. Barrett, L. F. (2006). Valence as a basic building block of emotional life. Journal of Research in Personality, 40(1), 35-55. boyd, d. (2008). Taken out of context: American teen sociality in networked publics. University of California, Berkeley, Berkeley. boyd, d. (2008). Why youth (heart) social network sites: The role of networked publics in teenage social life. In D. Buckingham (Ed.), Youth, identity, and digital media (pp. 119-142). Cambridge, MA: MIT Press. Brill, E. (1992). A simple rule-based part of speech tagger. Proceedings of the Third Conference on Applied Natural Language Processing, 152-155. Chaumartin, F.-R. (2007). UPAR7: A knowledge-based system for headline sentiment tagging. In Proceedings of the 4th International Workshop on Semantic Evaluations (SemEval-2007) (pp. 422-425). New York, NY: ACM. Choi, Y., & Cardie, C. (2008). Learning with compositional semantics as structural inference for subsentential sentiment analysis. Proceedings of the Conference on Empirical Methods in Natural Language Processing, 793-801. Cohen, W. (1995). Fast effective rule induction. Proceedings of the Twelfth International Conference on Machine Learning, 115123. Cornelius, R. R. (1996). The science of emotion.Upper Saddle River, NJ: Prentice Hall. Crystal, D. (2006). Language and the Internet (2nd ed.). Cambridge, UK: Cambridge University Press. Das, S., & Chen, M. (2001). Yahoo! for Amazon: Extracting market sentiment from stock message boards. Proceedings of the Asia Pacific Finance Association Annual Conference (APFA), Bangkok, Thailand, July 22-25, Retrieved July 17, 2009 from: http://sentiment.technicalanalysis.org.uk/DaCh.pdf. Denecke, K., & Nejdl, W. (2009). How valuable is medical social media data? Content analysis of the medical web. Information Sciences, 179(12), 1870-1880. Derks, D., Bos, A. E. R., & von Grumbkow, J. (2008). Emoticons and online message interpretation. Social Science Computer Review, 26(3), 379-388. Derks, D., Fischer, A. H., & Bos, A. E. R. (2008). The role of emotion in computer-mediated communication: A review. Computers in Human Behavior, 24(3), 766785. Diener, E., & Emmons, R. A. (1984). The independence of positive and negative affect. Journal of Personality and Social Psychology, 47(5), 1105-1117. Ekman, P. (1992). An argument for basic emotions. Cognition and Emotion, 6(3/4), 169-200. Esuli, A., & Sebastiani, F. (2006). SENTIWORDNET: A publicly available lexical resource for opinion mining. Proceedings of Language Resources and Evaluation (LREC) 2006, Retrieved July 28, 2009 from: http://tcc.fbk.eu/projects/ontotext/Publications/LREC2006-esuli-sebastiani.pdf. Fox, E. (2008). Emotion science.Basingstoke: Palgrave Macmillan. Fullwood, C., & Martino, O. I. (2007). Emoticons and impression formation. The Visual in Popular Culture, 19(7), 4-14. Gamon, M. (2004). Sentiment classification on customer feedback data: noisy data, large feature vectors, and the role of linguistic analysis. Proceedings of the 20th international conference on Computational Linguistics, No.841. Gamon, M., Aue, A., Corston-Oliver, S., & Ringger, E. (2005). Pulse: Mining customer opinions from free text (IDA 2005). Lecture Notes in Computer Science, 3646, 121-132. Gill, A. J., Gergle, D., French, R. M., & Oberlander, J. (2008). Emotion rating from short blog texts. In Proceeding of the twenty-sixth annual SIGCHI conference on Human factors in computing systems (pp. 1121-1124). New York, NY: ACM. Grinter, R. E., & Eldridge, M. (2003). Wan2tlk? everyday text messaging. CHI 2003, 441-448. Hancock, J. T., Gee, K., Ciaccio, K., & Lin, J. M.-H. (2008). I'm sad you're sad: Emotional contagion in CMC. Proceedings of the ACM 2008 conference on Computer supported cooperative work, 295-298. Hopkins, D. J., & King, G. (2010). A method of automated nonparametric content analysis for social science. American Journal of Political Science, 54(1), 229-247. Huang, Y.-P., Goh, T., & Liew, C. L. (2007). Hunting suicide notes in web 2.0 - Preliminary findings. In Ninth Ieee International Symposium On Multimedia - Workshops, Proceedings (pp. 517-521). Los Alamitos: IEEE. Huppert, F. A., & Whittington, J. E. (2003). Evidence for the independence of positive and negative well-being: Implications for quality of life assessment. British Journal of Health Psychology, 8(1), 107-122. Kaji, N., & Kitsuregawa, M. (2007). Building lexicon for sentiment analysis from massive collection of HTML documents. In Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (pp. 1075-1083, retrieved July 1028 from: http://www.aclweb.org/anthology/D/D1007/D1007-1115.pdf). Krippendorff, K. (2004). Content analysis: An introduction to its methodology.Thousand Oaks, CA: Sage. Kukich, K. (1992). Techniques for automatically correcting words in text. ACM computing surveys, 24(4), 377-439. Liu, H., Lieberman, H., & Selker, T. (2003). A model of textual affect sensing using real-world knowledge. Proceedings of the 2003 International Conference on Intelligent User Interfaces, IUI 2003, 125-132. Mauss, I. B., & Robinson, M. D. (2009). Measures of emotion: A review. Cognition and Emotion, 23(2), 209-237. Mishne, G. (2005). Experiments with mood classification in Blog posts. Style - the 1st Workshop on Stylistic Analysis Of Text For Information Access, at SIGIR 2005. Mishne, G., & de Rijke, M. (2006). Capturing global mood levels using Blog posts. In Proceedings of the AAAI Spring Symposium on Computational Approaches to Analysing Weblogs (AAAI-CAAW) (pp. 145-152). Menlo Park, CA: AAAI Press. Nardi, B. A. (2005). Beyond bandwidth: Dimensions of connection in interpersonal communication. Computer-Supported Cooperative Work, 14(1), 91-130. Neviarouskaya, A., Prendinger, H., & Ishizuka, M. (2007). Textual affect sensing for sociable and expressive online communication. Lecture Notes in Computer Science, 4738, 218-229. Ng, V., Dasgupta, S., & Arifin, S. M. N. (2006). Examining the role of linguistic knowledge sources in the automatic identification and classification of reviews. Proceedings of the COLING/ACL 2006 Main Conference, 611-618. Pang, B., & Lee, L. (2004). Sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. In Proceedings of ACL 2004 (pp. 271-278). New York: ACL Press. Pang, B., & Lee, L. (2005). Seeing stars: Exploiting class relationships for sentiment categorization with respect to rating scales. Proceedings of the 43rd Annual Meeting of the ACL, 115-124. Pang, B., & Lee, L. (2008). Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 1(1-2), 1-135. Pennebaker, J., Mehl, M., & Niederhoffer, K. (2003). Psychological aspects of natural language use: Our words, our selves. Annual Review of Psychology, 54, 547-577. Pennebaker, J. W., Mayne, T., & Francis, M. E. (1997). Linguistic predictors of adaptive bereavement. Journal of Personality and Social Psychology, 72(4), 863-871. Pollock, J. J., & Zamora, A. (1984). Automatic spelling correction in scientific and scholarly text. Communications of the ACM, 27(4), 358-368. Prabowo, R., & Thelwall, M. (2009). Sentiment analysis: A combined approach. Journal of Informetrics, 3(1), 143-157. Read, J. (2005). Using emoticons to reduce dependency in machine learning techniques for sentiment classification. Proceedings of the ACL 2005 Student Research Workshop, 43-48. Riloff, E., Patwardhan, S., & Wiebe, J. (2006). Feature subsumption for opinion analysis. Proceedings of the Conference on Empirical Methods in Natural Language Processing, 440-448. Riloff, E., & Wiebe, J. (2003). Learning extraction patterns for subjective expressions. Proceedings of the 2003 Conference on Empirical Methods in Natural Language Processing (EMNLP-03), Retrieved April 11, 2010 from: http://www.cs.utah.edu/~riloff/pdfs/emnlp2003.pdf. Russell, J. A. (1979). Affective space is bipolar. Journal of Personality and Social Psychology, 37(3), 345-356. Schapire, R., & Singer, Y. (2000). BoosTexter: A boosting-based system for text categorization. Machine Learning, 39(2/3), 135-168. Short, J. C., & Palmer, T. B. (2008). The application of DICTION to content analysis research in strategic management. Organizational Research Methods, 11(4), 727-752. Snyder, B., & Barzilay, R. (2007). Multiple aspect ranking using the good grief algorithm. Proceedings of NAACL HLT. Stone, P. J., Dunphy, D. C., Smith, M. S., & Ogilvie, D. M. (1966). The general inquirer: A computer approach to content analysis.Cambridge, MA: The MIT Press. Stoppard, J. M., & Gunn Gruchy, C. D. (1993). Gender, context, and expression of positive emotion. Personality and Social Psychology Bulletin, 19(2), 143-150. Strapparava, C., & Mihalcea, R. (2008). Learning to identify emotions in text, Proceedings of the 2008 ACM symposium on Applied computing (pp. 1556-1560). New York, NY: ACM. Strapparava, C., & Valitutti, A. (2004). Wordnet-affect: an affective extension of wordnet. In Proceedings of the 4th International Conference on Language Resources and Evaluation (pp. 1083-1086). Lisbon. Tang, H., Tan, S., & Cheng, X. (2009). A survey on sentiment detection of reviews. Expert Systems with Applications: An International Journal, 36(7), 10760-10773. Thelwall, M. (2009). MySpace comments. Online Information Review, 33(1), 58-76. 34Tf / : D E Q [ \ a l   2 7 G H I J M)","1"$$$/.r.0+000O0O1r1110313334 hojhXq>8?:????AADDDDGGIIKKLLNNFSTS}} hPchXThXhXhXDELMRSXYZcfhimͼ޲ͼޫޚޖޫh=hXAhXhXWhXhX2&& $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2%+23{2& $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2 $$Ifa$gdt|3;=ABEz{ ֽ髖~oV髖o1hZ hX2&& $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2 !"#*,0EFLMTYZabʴʉʴq.h@T8hX2&& $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb22& $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2 $$Ifa$gdt| "()*18:@ADKLNTV^yưqqưưXưưư@.hi\hX2&& $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2yz45ޯpޯޯޯXޯ.h hX2&& $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb22& $$Ifa$gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2 $$Ifa$gdt|2kd_$$Iflr4= "' &7 }0"644 la]pytb2 $$Ifa$gdt|X_fls $$Ifa$gdt| $$Ifa$gdt|XstrʹʤهnX@.hgOhXWhX999- $$Ifa$gdt|gdt|kd_$$Iflr4= "' &7 }0"644 la]pytb2w $Ifgdt| $$Ifa$gdt|A5)) $$Ifa$gdt| $$Ifa$gdt|kd_$$Iflr{  ' &0 644 la]p[5) $$Ifa$gdt|kdK$$Iflr{  ' &0 644 la]p $$Ifa$gdt|Z[abiovw딅o\o\C\o1hBEhXhX  1pBObHq.rKq'$0^`0a$gdt|'5  *!!G""#4$%%% !""^#$$%%%%$a$gdt|$0^`0a$gdt|$T%%%%%8TCt  Q!!! """"C## $m$$$$%%%%%%d&&&&&& ' ' ' '''''''''''տనh:Mh5(CJaJh:Mh5(CJaJmH sH jhvUhvhAnh5(CJaJh5(CJaJh9~h5(6CJ]aJh9~h5(CJaJh5(jh5(0JUhG UhD)hhX[8*[ fweBXFg0: I٪#8gAlGM{&泅,D,Tl'QY,BeAwT3.<! g"3l7²/U F6&\_^Be>vfi7Bb2Gx$̓7BX;*)d+ .K> E,TllrnhnJ-Q yEGM-T@6F60k@6IR5/H #*[ټ Y34 y?-yY* #K}s٣7E^)@ Sءl\k]} AG S(<-ߊW5}䠲Dz32JYGՋGM6˽G\ȖҢ?*jj=t#=lTalYY$_xtk3$Թn~`lSH-_77G5'TV6T#*ˍ\dKx5ҤH…kj?꠲!s?]=̲$g|!PZvHv< l|m;N]xEX=Qʂd< S le5rd[]Ozm}'aά&M6_xЈ/KufYjUS6ch&?$V#^Beɽl>~d;Z[} ΁sx$;Z|kG%F6C[}i[ٸclyXkd{&[,/b~DhiKŵoy+[S[9zlۋ7+m ? /eVzڊZ#ӳ-~ِЖllR2"ۣC(41}-mix##ۣCۄleʖΦyd/l glf@! vEbZH75ɖjzl[Vx#[}5#Me}efi2la lXlbfl4%cܐmd'yjs~_X6WIڜo Sv "p5^Q#C۲lR{:<-N:HD=fe{Pzqڝ?d㭠V 3kmMfkɄlFZlL/lLݎ.z=&z'-Nd/5>&e^ 7#kzW,GBtE|[WWxThk Q華Ϫ&xިڶntB[l~bcXGm[?1-wP .$d{Hh͈D#6cHoDf EAÕ;ЩvKjӭ d&mc fIuuja! qq T!J5ih+,Dy}LTȦĂl2~f!j:K~${![lr-se(tD{gȶA=G9::7LLɶ\>@*Nɂ `Z dk3*ܸZ6mdM$n"(QG[GuB.'h$pַۛl$-zUG!ŭnl׬0f3PFY^[݆ ٮd5:F5nl ]YÅ谚z^669ڜ1KG^g霽#:<R N(~]omȾе?&n W o.dlRd-O'ld;VS5bǩollK츬BCmD ߳V!^ *{'/7߈ emB[t!z@|]EMK6N]E1ٛ7U[ݎ6>PF÷C]I!,'`ZLޙ+f|:S6uchԺѥM]~hsem›iv?,0&oۅhp:]kP2dd߲iMl)8mDK!DԼ [a:#]k+XM^Y݌Bf4(AfuYhKnæfMl.+.DY?y:A(nnخ' d^i\',$zo>dcB[fZZ/ DuY mE`+DC>,5ǚiTPLdd6"=*T'wM`eYㅨEmԚ+fAzcdt?n;=҉($v/DTXfYnn5dd|zB:7)˩R5Tdsw{F͟|7|¸Zu}ǶnI>վoS9ZOӈBR.{n1uۺB֍.} [MW]mFobdc}k3\ޗE3n31ӘHhf5~1#IUǛ/?0+J/D-t{4{{d#k[Ӿ}=lt{Hr".3 MV0jcdggoFKoOlt{49~,dB/\NNʨQL!ӍY(dKSezvmkĬnqP.=TN}0TnbCbr2frAiҡx\)^\qiU[j.8]Z݌Wn;}\%Uv\Hu{A^V;\ ɣ.Fw*yX.-dcBcgdnSl]q<:~Զ|Ge= ٖ腒w{HMG!Jң=E= ٖ6!N)[7E-ul=Cllڣm=tPR?)[qmn0ҁn odS%gdR Y8!NPVC6 d.t{#[1'^/ #ȶDae?x$̸2bE: h&;!G̸2EQȶxP+dSՎ"}fA%BM7=^Q^Fm FH+"9 .<5@y"Q{$[Eruۣm@PG=VR,1hDy"* e dMQ%[vmGj<Gd~MNT} aehB4~ #F6~NuAlGc d{Ig \t =fo]~ӓި@QW+][LuB0RH-|9>@TΒms@˅AҖۮVZ@9f6 Q8d(Rk=(D!Ȃ7 mlklYm(d{f֯g}=TtvH7G! E%.FS@(D!ۣc}wO8lB{Ý'K! Dnn}mA4nm(D!cs*s)D!c(v9>"4ѽRmo@G:_Lmlng6ħB=k=N6Bb֣#yd#İG#[døG?yZB6J;ٮ\xRo QFɦewu<:ٮ??`) u]|l{;ٮoI[ wwul# w@6"t<FĖNR% 7hUY7@66vI۱*٨&ٴ  ;7m%3PZfl4dVjlN] Pyl^xHHtU~wA6 *Ki>e˒-U;DEd{!zl]!/L>fcJyl*_GՋ>+CUn6dyTFgVpb#0чPX5G=m|n6e+/<5G=m|n6d Qe0'/<5Gݗ|+^w&ۙqݗ!&#[.PG^'x&[ȊHG^'%[vK{u|;gEl/Qe?dC{u,v?"fJ qlv?KYd?e} l}VylH\-WىNIe+Blr lx#[}ᾕj ic5:F'$&/C)[lJ2tMS޶MLeE,TVݒVtMlyTə9:6ZI6GiT5샫aN-QCռϬlJ3tl S stntljk4[e3t͑z#[!<˅dz :C&oxnm^E6J4F<{t#Kb,'Wۏd&IIRk6E@6w:٦ U ْGEf>d<1E'$dpR?B6{o޼A'$z3mdS%ţS"6ɹf6?~D`UlAct\5 Ewvύ!MjDiD`BlTM`l(F'='!5x~6 Ï(ųɖm%&ajDMx6dͣ?~L1srHqZŢy| F cɖ V)نt'-%dm$C7ۀkz\Kʽ>@G: z2vi-^z`\{Gɯ]%q?l=Uo@B7[dK2xkR*5N^JKbPVJ~%[˼StL=l.tZ- ,hǩZrt5TkKn6=l βq(*,O%[ey*moɌ~*JbG$11f5v䁸ϕ~*ry(Dgg$[!u~L-V#S73 TԫCo6;nچqye6D%AhmӪcjjwx`٪5\6S-ethRh۠\8GDi=%LCt5+&l8d+D)o+yV/fQV*ƩU''[Eu 6D{V׵ >S$5:*9VFgV7w#yC ʿ:NE"VJ7BT`Ui8Z}>)-US:Sn]f!#3.'"LG RD\,a07zEF<4}Y"ɺ;dky]ǿ"0j#d% ۲l'nS 9![fۏAء3,u6_G d[ș?3 qo]{]Z7=l]FvQ%-S]FYRRX~LyǕ'^ EaB81&*UbW S?T6C\迨Ve1MD,TFrh?&b̲E5o QE/ſӲJ}fե Yjn QV:ib*R몾T*/v:llmk9"*e{ YْrMe昜`KR>L d߫?N6 ZB69,zkGY6wj׫GLv#ۨdt}ܵ#h:qن_u+[} YWGrG:.۰z6+G`现U\-]mVMkGXc7lm٪q-jdSv̞heV2nd9n G%m8-[XoVdq a!6e{h%vm沉"iٷm*_dc&6""dScy'[nt/jG Jwkܖv1#>fe{笫pO] 0㲕׎LxԻfhpqҫlK* U ւ=]ֹΎ9-Zȿ39 YپaγJlml2)GFh38-OiHc.Geg[3vtZ69cydBe*Ȧ ]Z\l5dՅ.-.Vo^t's_䱲*nw1e y e_ي=N6l!{ljhϏlyHśDludCq;;~&N2nyl:{z's]DO,cOٺE#N6l.cGD`ֶhU=(흖-Zǎ5+g$|=ٜdc˦drF=ٜdcn˦dsj$XN6l:&[kd^VdsAY~=mFBؒN6>yY+yܗWޱl-H\+]dz-[>c'qGkC6w߱˲U1KWޱlMn$takL]F^Q˻e2.[}d'H(zOt\:Lea- ߱ldH軶tz7Nƣltyiкvάn62f ;\йqIlUON>rZ@|)Hkuh5L\3Zd?5-|Pi 5 lLP֜j~~\4?@9^ih5ܸWj}xvӤcm&[q.uDr{AZܡZ#5|%߾K\ҨF}\Φ0wdj%J6l mPɺ5aՑO~ls^N]-5ZLkMf-]qe\rY60^_V L=lmt݀^ t3Rw9m7_=%xɦx͆ CֵBְzQ+"{6e^ b_[{MW\}Z|1N˦ve2{p5l=EdK/8-` vHh][K<:[6ˠ6uzfwz\R6_[ȶȫ;F;OWf[7.KMW\J5zׯlsc$]ƺFBa˵dy5~xO{ L~fvE/|lc,[m^j[Jhj*94hQll|zKVzU?CK^-g&e. dWb}lSv^z)L |)U&umCsf56+e+ΟͰ ]^撙Hz[:->oR64_pr'֘a+N6gЯc e+cY˜ FӲˁ2-,6Ij>$miٲiDdOfRZ( FwNӲ 2PYՋM^?2z \]UڭFnab#Lgdd;RP-2…X['Yid_a+N6?U6FvuSgd.R5`U̙DXxDx|PH[7e&?.FV$_xtK'\}{FlRhHG63l:6ٴ ҵe.ueMd}WJɬG: We,lc?qf>Q MWtޜs[i? t=`r#LdrVɖY./Q#FḀxx 5Lvv-ٺeV6ul8ҵzsy]ɦB1 Z& kBFȖ B%[d˚3"[ӌ Uc^y :p)5r?ןnl<̻)R%ۗ挹~6JA[1m"ޔPVW581;]삿-P4g,FTcK!&7-lo|_|m$ۥvR}Pll ]l*o[tzgwLnlaqɉmY(ٖ Re.w_~] 7-~xG~tX',9ɵ~6|A3(S]6=R ?S]'e+lu+Jd/HQ &?L· Ye= 2<}ifxSq+5ֹy-i^ZDrlaf R{>}hS~Do*Ksj|{;?֞b7JXl'Q䱒-jΐl5tk_l&VȍZròUlfQٖTA*Qp &kg?uet.թ,3uFMm% ' RV2*l/:DZ+.UPeX qsFٖۂT+ ׯ_s'Gڲ-^✍lŹ *[KVa3g z3p)H ,_l.1LSsb: QTNG":;AI&"ے?T+V{C$?[Iv7vbeXlY3c:[8fX:.çKSlvhN#[9W.Ef6__. )JTwr3TRo3HOҵL{_gs6ASϬl_*a|zC^Aؚ9ߪm?ڐlJXl΃NGj JP]Gzir/iΰ}ŎMs)>fS߿t#orc%YezSU+Ț˔'YWɦ\\@N]߿wwcpnZVdi'E:]/m\+dmX0lUVn5xB|l.TVen'ن*חŲ=qsG#ۊnod'XΘ/ & ˑm8h.v/-sXmH7Ef~Zj$mNA Ԣ:qS/]mP|ۅl*̶Vt6eٴߔ*[,{EN~hexM <,tx{w6dwW kQ釲ʘ϶"۸l_Ov8;'ڔ=C*] wwa .PЍl ^X3SwMQ;a!6&on{l~)kDFtqBñ\eY^~;槟~qJtvYlKJUM5 !v5IW0buY/mHYlK .~`Q;6ˏhY bqζ8hvU]6I@/|U\~*ݤPaVVXl:t>Sb[z] q9jl:,(*m;7[񷙨bg}5e.t/WzmliOpZֺȦ8m˃I] (0G]t ;۲-}-Ap]hS~lӤӑ-mtwh3 /7\wo߆5Z'_o͠q`o@>Wz"W^ҴXu6۰&kόgM yٌ 6SXŤKw7Jq5d/É9%w6k"[l+onST-4_ŐWM|j>57fPMf9,6iFHbUql+%.dʘl{~_7Q ڜkg}n[JTE 6W5J&jSne{LD*'~VIT=-A;y⃿Tl6?rlͣ/0[9qaE?^voy-oj^#F6y1ٓtdm@+_McPF7D7[fv>n ̐m0nf( ͗ iۓK5WYdq-|3pfwlso;Iw_'-!dnz{f HVFg*a~ݎe4zSWd?kə/bO&$_ ə,-E/5>RbW9Dx-Ld6 ўlZ-ƺȦz$Aə"-jx|J{4PdӚo x?}}:|rx hTZ$L77_c%韊l]`O}t"{߮F9:y9濊OMb_l*o?ܻkrtLrW%G'' Y|ll#ď"A^~|rEEd[u⎒Y0ѭ[ū絤Kݿ8Z&g#sݝYzgS-7lצM`R16j 䫝əAd[u𪘓_~;3ME-j%oIИQ?~N~ae Ew&gf:;N%sϗz$[M߫٦b4&~G7Drt693col^]zZt iQo_ˠ-՟3yLNrt.9s!.o6L**7#ergӑo?"_u>=U,Iwn˶<_x >r޽sgN:te[+3p[5ck:~9_3$ށC-[(_ee{F{NOg$y|Ӳ-w_x~FB6Kyզwljirgِe5^OLN)Eruj痿[f0Z6L'gYQT87$gJ?k/:MEpx(N l,]ztrȶȫlj򝘊Ag{_k?LY dtroB&gl[syٲ1Utti #pȶXN'g>v!-v\ /n[g泱u?M DiphLfnı )į?F69_8lkS#ϛLjkDE^M&VUAcf 5Նh`ͥlj8C|TrbLl eۜ;V,5l̖3D=ѩ,:y,B2ulv r5d)_쑫G-Խِd JR{:3۟m:9enϖqٺreMGO,o~-Mlg 9 dcrf)nʉeR#kiw3R4Z?}',~e\aytoWcl9:1T&'f$g"0òތhekJ#VW`"ow1冿PrΖYfU~B-::5lq#ۋ:03# d[l!bkl]2Dr>~/`"9[?a:9s9Սlu+[#"Vȶf y'#-:xU ٺ5 JF<ݿ7M/3r|l.k'Ox\9zNiӧ$gc=o([[rJٚ[̃6j*_;a=u?oT!z'L|L9zƴF3ў7*lh:]'kqV~'ޗORsW/X 灛-,\2oquQ9Z^)G 7SUvestZʣu[69_VtEVes\_ձďrԿ.]e]3Bsd+,f޵M0Wb"]EbW6G[\Ɲlip#WF];*"ER9f2N˖ÚmeE6W+K娣EY*lmm!,v,lU1*'O9XۑB9a㲱#XUgQ :6|9aqՁlw۲MӸV8 ٜr_"Ȗ|WcBK9:cK'rdLX\ /j}{K' Y.Cfsu,l2rx޺&َΔ&wd{ņYV7w5TL^v/?ÖlQg˸-[|ؒՉӏVvӲUMcK6w;QWWpZS%UrԻmTilg~v:.|XՉSt78-|XD9l^eنOƚlnQ6fkpYR6[xr.]e>(m{;q\[ٺ+9[v-.Fؓى7[;;ed9S:|$ ]eK;ON`O6[]9ZKmنOƢlWr4Jw+k8-y[eslyZ%6Woe.9BhJ..KqB{P4eswG[(bot![U9].W]?ʐ,Gb˩\ᘑU9ot![w'~4(ooYބFn}ƣ!ih>n!AӟotL:v}ɸe[Ƕl.O@dl.O Gm9B(06:wQ|Wr_l#֕dp |ea]6'~'\M .[;<}/mVg_6w[\]<:nq*U]m!T{ ;벭 Xٗ݉umhl)lIU} [U%{gx-lNH9㲥@Nlds/re"?@NlCds[w=лӲBώ8-[Nׇ݉1We W+6x8.˦Q#-qYLsٜm!x˲-]dsw8,[0} 8,[0m!}'$?]D)z-Kk+UVڭ7&l;m -]p[6֓/uz_ܝ5Nv-[h}qЭNȖr̪=j_iX_\X`UԞ-h!˲]nd!H֞-Neod:HldC a\mƪ3K" nY>,[p#Om,/Y>;lh&Rug; ZqZl+"Z֝/h!ӲH" پtg0>ne^ڣ^eֵ۲d1G >6&[:Ųd+xnSr*ٺT_ayٺF,ݭ[RwݬlH6eAZ/ ՝)'"fe}Ie%K7h12@6,pEIuRݺx eن{}ʫ<. z'6dӲe] \ѹaT=IlrSi@9dӲ v ,ׯV}Ymz{ցQĺ< EO=3O?CmzrʐYTrě򜇻Mnl-vfեMoV7ŷ?Mb{˟2x5YK?7@d ]%dSX,ȿq?Yqc} |af'qli""֎ qY:U\lH6rDž)lVq]3ϲMTBelgŒxN+GӸqB-"]ܬܫ:ق?)lVq^>磊:~ilvL6Qƍlu*&f=ߋl7 2>V_I#[u--sPg?]lWپ~oB"eR7#Y%l;֩fd[W*Y'۹D6#<)OrBmU|#[[fodΣF֌OvЦ"GiTrɇVG'~6ll dK nd;v>[m;"β?eAJ kfȶlζS?qјVK~ښȆ-l/B֬m m-:eK#fJ61L\l-sȆfd Ԋekl6sGMdK,IlqW ȶ@3-Chd[v'-Ўfk)mvGf#f;O:ls0hec=IG2^B6Զ>I8l smK>eSBOV9l ӓMLChۈe ӗܵ@j&C[G=mlBVmuC0hȦ lm8ld?6l%]x6 ׏ж eeà& 巏-@6#J /۞OݸVlf`<~˶"9lYVsv;Vsd ,"ȶTI0hu/G-#`<^ 9XDz'l1hu0ndd Zd#ہ{8lhb=@6ʩ#!"Rs C*O2 9lij'wfd Zrو^oVfBfEA6ꩺ2E6sǛqfEFBfA+# &0od{,"5m?h%Ca#-xȶ!d{xxo)d[=dA+ȶ!d{u<p3U[D V8l~A+ȶdN;V8l=."8mffЪ~@Y!ۣ461d-=El+BMmQVu[dEs.x<" >ʖ%#TRMս6Qg[7٪DXl%[h?c )d""VA2ծ&CٖPJ4#gOl&[2:*l Bmd˥l9-9*e{Tݎm dbV$_xtk*ٜغ* ϶olqsTة-" ɖ k*%Cmdl9*e{T lLn+[{Tl+x&[ȖGݒ mdپGl=|<-[%K{)ٖM)>Gl=)|-W/l]ǚ[!-l<7uV=|ENx&[dDzE{ MdsaM<l0jds`.Plrla2h"9$ jdڣJ6ilx|Ȧ6-|nQ+6 ^3uF&h9-U٠dscdܘB꦳%x VjXPYdscnJ̪Vp%Su۝'QlrKՖTlNڷU(GG6֓E,T4ݹN&+?:-NlVb2S;UH6vkG%mdK6%]d1d5O*e,'[gKX,qSCVS)[حmdsbU]mSx)[u+[OK -T+(GR2-'{x# VT9[|A(\vʖ/&F6nZ7룷iV#'Xeu?-¥ladNd_Beu͍0du7 4) 1Uw@1Q#ۇܘ; ȖFE[3*M?c-kWċGg1Uw#[.;ldٓd+f#H*7eÐdSL&ۆF=yb Y GObìx7eÐd͉7ÐՀTjp\H0r&(GE6J0d50ɀ!>mW0d(7{R!m_0d(98UW CVW#E6*rtuF6W+Gũ Ǒ-#Ud YuF6'gOJ0dvCV-ٓ Y5@Uadssτl8l{@98lnNUUG6K* Y)#[׉!+qdsCVF$ٓ Y  F6gԡI6{x&$;lNՕe-#;rH YJN :lϞT)ŐdCV᩺ YHʹǑ婺/G#SuU =lQ>dy!I'lʖ%_Q'ٚgBB6e=m%VfGٜMvOW)|P(ds[?ۣ /fOJԐds[7etfU gd9 r 9+-'VX5GE:dsY6Q./<5G=MYA6eC#[I*-~%r?4)pᚣ&7?llg%9 I_6O*T~_5ddՅ屔=:͓ٓvٜ-d-keeJl'dپGy("˲!+#!K{/[[ר9-[=:ͫYb"sVd!eci~6We%g'[l̞TIGdBը+[(oK7hؑl,CdsVȖ|fdKgЎgsUSpG3UWQge nH~F]-VEnP6 \-UoP~泹+[) 즳-lO7KƂ(OBe@6fODh#ۉlbd y-lϳ}?YuQn l^Mu6+o_!ekFgOn@6fOvc\`%+VCn>[c[lѴlMH6/e+9'lM4zta_duVbN65ŏNԮ7l--#b0~e{}ݳfdkpɪ-ytva_d/ܷV66'[GhֶG'kF6EܟMlqDٚI$c_X<=ysqaRod#؉azlwZێ:plqv gÜl~ؚ͞äȖnU9f>d;0?eW  Tɇnyrijd/v՝äVS#WW6Uw#[au͞WȦ>ņY+Eg&|x6Hc B1ämCel<äm d0)dXxcWm̃e7+B1ͯ B1-ڜ#0&O?wx0hs8h4?=ã_eä/9(F{@6 6e;T1l6'۶xD_ဲ%r_xx--^]xxߚ--~[xݶOoMݐxDl>SDQ,aLlO2{RW3"Qtd9,_MdERju#Dd{>ml+|au(8c4{ˆiFV:''/Md;7K3èlT'4%_FM?"ndleؓ-RǮ-$O5UWI;ʰ?43u<'ȼh"#[er!,s3h\Su~4-٪#:&`%&?eBatvG$X6/ɖdr>[Ȃ(OBhֳ-+X6/ȖL44k3V\yDGWdf|~pȲDFZ|_*`G܎,Md;Ax ||%Z;F66oBXnـ"[ ٲ7BX;ʦ֍[ (%sMd+lEIF 1:7M,&|PL /f[V*Mdcr:lVN6mZؑ&U˲թ /G-Z,v|KBe$ n{vlEĚ>؜l~p k9DF6&aMLe+ن٘M/e˗eN͆mV!~ǯl!czfM-M<ڹxo"%[mei_Gl7<"%YӲ%G9 Md;뿰fltF4̟'DK:-k]p;|&_1U%Qސh6.=lN7MtG I&eTz v8!MdKE\/j㓶l]pB69<-KCVL6nlu0/57<Z11-[8}kpۥq#9DI6֌D ԊvUvS@Nw"3 ~Ad3'֍52&D̗sbxD@d[fݴ_WD~AdӲMΞ\LЗY gWlKcl]K{K۴ζī;e;%R3AO~Iǟj:vDd[3'O֨q)lf YU6QE?mfl̈́uي^ǟl9"2m=8'u+7&[0W ue^1-[i(UAї~hǪlTٖx5~njl[:%dRwdsbb@sÇ/fTd[hng}Գg*vl͡f jZM6lB6qMW?lwbLEբᛧ-5Lil@5dKLUpV6ntaZ<&K1m64{N/ݨ߾snn6|gw|&fK1m9ѹqljK6@5fd (bE7%[X00i֘-ڒ|7ʂ)M?[l"ms5i%7w,3#Fڨե|Il mL^W6^ nt4c!17'74% Rh-RJ~y5*o9m9,,,Mx6V,0d[cF҂l\fwHc p;-6-SLRT٬@Sy725mA7ȶƌl&YAi]7^gLQ sYåƥ.-9sF'lk++u߲lAsFAȶƌlΞkd_ړmR7ȶƾ-Ll61e:YIF qd[kZvPmYْMwعg+!f69vel6%g\>lVDvԺlSuv`MwرMЛϖ|;62^CRMSuE6Ftƃ lm"$u{)۵Gm}e#ll}tv4"2sm=Ɉr49ZHi.'^Yӏ)ٓmaod5Nol-V>NɆ~u8~ u+')DtV͒sl7 S~6y~2(zL[=?4H!e3{r_4&M|ő#[iwvن SKuْs6dOu6;SuEojc4? mh69lvA'vql#Sp~4cct6i};hs@[߶J2*MUhÚ:[io6L>i;pT]d^*8O˖\hEnA;]z]y7O70HWˋ[lK" Rlgu61-n˶׏a-*< ծm_"[ҤH…kfgW]>mȍl+QJ.ۙeI%yٜy25#JAze.,lmu7n-^Pǣea!m =td{}i>l{-5 Rc|gzux@' l5?Б-*ۇ dg{m 5ͧ:[]ed^CMՕ<")&;N=ٺ_-tb,Y?U>X6j.[}ag5d;Tݽ# #[}Tnt3d6݇\gc =rd(U_F ѦpDeɖ͂:l p!dRz%["lgn4OBelǛ=iEl>ɖyE2 Od;q@guۙX<: F #۵ =ludYHNW\6̚llۭ!NueS ^؅u3udJ d =lbqV-y4e#UKԒIf%9e#V݂d!)$G_T#.dcdd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@7⿐:g#MOw"O-/IOJSt,͓#O#k&'iNglxN*yqL$9UO1IN@džku6Ik|_d2 Ueo,P,yj%'BH䩹er(ĩU\Snyj#ֻxȖɯ,|h!3wjeg&3G$9R6yjzɑS@'\SSsj=FTHdl~v.3wheg*SZie$SSHS#eZr,n-`s;M<;3Dvd'H#q$GSP6W\39QNNlw"[f=oݧN>MxfO"j3ArH)<6J+cyj^&')ԩE'8u_d+lqN𳲤xy$Gɦ.%FaE\\VRlww"[.Z7 OdO."~Fvcyjkd'$eP~ed3yj*ne.ɩԩMϽ{De&ZO?/+ͮdt~"_<4IM~e2"Qr&O}]6ṩN'gYs&JS]e12Þ]Ld 2"~e]=xR%OeoW'G&Aʭm5#]yIv!d3IΥ=?|v"݋l+udcoMh%"M{؅S%GʽYrv6P6f([ #D-ەVz/[h&YodKvrH{eo7wMDOdc]dKfC`$[fVD~'2:}ie$[C|E,&pɖv)Bȶβe]eB}~%GjV%GJlK}RPr2$1 #m"2-4v볱l&o2#d[TP6:[e*[X#^L#IRlf[ 9i,Ԗ-Hv4+F1 Odh([Rld$H¬ )f CvꚄ"2M5)eOQܬT9l 'CٌP;`"[c玲efXLe3vL6f,[e$[}1ͬT!yOٲXDh_ Kud[ e3vSf,[f(D6qM^-oPm<2 v-nSl'C٪ lKdž}XlgHٖ'y&-)Oγf05Vl b6SWfX)dnlyMg069v~5M&l+&1[l!Fg&62Y6o49VAy>V&S]qq3^d3KξnO%YgpshѱeGdqM>N3d;f9{ׇllAf/@09 d{1s_|o<2؟&UF"7D3NM616j&ƙg?`ǓC{ɳVΓbVirl;OL7yRΓR6MlO*u{&ўYlqrl{l17ԕyM{:;Ĥ+³07-49B6Kayؚ^wpzp)ٌ`%srup wIdd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@(w{ NLsdT=OvbBS# ȶE# ȶsdT6OvVΑQ?dۉ[٪d![dV.Jf-K5@Y -m'lm' -m' -m' -m' -m'ZIrGCيUTA$KQ<]먗A+lYJegYrW8YpJTA+l\Xٸ5N r߸b41Epl{ߕFRčqKT/G!e*ET&FR6+lDddձR(KlwEN1\ )d,[Eh#y-8 @ּ䗂lw{ߕV6.TRXv#}NLTܒRʖ k\6ޜvqe/K!۝>wEVzYoez"hƩd TõjŅ{YA{y}J+[dˇE acέl+)BʏM@P]Ʋ%_$U=l''SL7־l{ߕll,:%[V?ndۣokm':جl9-RgjESR5 ƔlI5sR9# [aW3m+JFR2hB[&GFRv}{y}JOP4%۩l嘂-he e@-o~}7fPlKv/]Ȗd;d:ȶde [0#h>d[җJ mlլlY3ŨS mllEBH8leôlUrd[&̥*#"yZ&[5]A^^}2-m0ٖl ecsqB\M mQljdF&W/@w-5rF!kV(.ri$׍&%)oy}Wl<IJCv5JJ\Ɗir  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  O>oJ)x c X2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  d¹`' xʲK^dt#S i R&7{\1U~d{=T >GG4LFV;EcR֓ Ӳ]XufPNO8)Iيm59dյ|+? enkd2YϬ5i?` N/t5.f/Y"“hYl\pq(8)&jvPe>"._R_d%:Lg*ĺPy"Gy3Av4^!ekdK![$)Oٲ$2 xٲMg:eg2eSdQKmM rwȖl3lᅩąk4w7VdK'FDw^٦TU_\r9DWۉRZGr܋P~6Og"䤄͑Mdg>lۏAɐcG(U O^XE9JWKNT,s :D Xg RX\:f}:}Mo*Fex]Pyd'9~llSuZ{\ ^1d'[҄7FӸ.}"[u!aY&e X4[`7L6S٪Udd8ONƲq;Sdˇa36ZHUD;/ um@9Hg)^,V /;` 'ћ-"TvhنRWTW~يUd?/˨-)s|Iʖ܃\ wp'2.MtVdå)/Hd:Ere+V+D ycVoWVƹd+ [ΫF&e.u8Kٲ`rvˊߦBc.7eB&u)Z s*G9Qg8%ܯ7va6o$~y#Mv0&~":8ē\הoe6qIb)qvùYP)ٲQWr"y{6޽{w# pBMJ|*l1` ۯ?fK2aHb(ۻq>_Ą+΋ldcO& elm,&*j/2,IYIV2!E_&dD:BdDCgm&;?vlc, dKƖ4FW\K@:YE>KU6Zd =lQѦⲝKيUnl8);sDQ<5+MehՄ\vZXn,V6Ȳ d7P|sDԝA`r>؉W)t:uӿv;m|#;>d:kWd?5D]GDҒmMgcb6٪lbvlט˛%\ h6f}d˞JQ&g*,ƲM,!h">+2ÁQQ^lsɖ@SĀlO%cV޿EVq=D5]eS @GlbL1(Td3m_LhS%򏺝 n(|6^VJ\q>U6^eg-ngeK>5m [3-P[AF"_X.M23٦"[|ltVTq炗vNYT6+!L zWlbN5WG1% HrX~R:,Q@TWȖv3M3Y$[YolrGK255m*-5^=!SFE dsTddSݧO Gl56zicD􋰽Rf"QOrlKv-k~6qlJҖ-GptTB?6fFOu,>bkdK:Oab">'kc~6rf2&F'OMldĢ1 nr>~HV r*l{}l |mnNf,C٦F&.FWY["'OSNBHS{#[*kd;!nFv3@]pMm(tQql<B6TG> [feId  dܟM&r@ =lS;OtEP;E9=ɁxSiͩ-./j)Nv*$׺|lu UFNntj T6ȥkI2.[=?Cr0&eA(=X9?΢"ⵜAŹY7μ$6T:˓n{6AX^mjT9+,oMɦ/b\\L5gY-<0%^~w,w}.٦ʯL>-ЍM=tC M7˦ ]x^r"? a^ [6UkF D0eZ6lwL>(Klc󲥁iv1'IV p~xMV˷}0٦!M+[aAiI'[LUwZ [g5 ?숭^Edy$g㟰Ҳ=l}\Vm}ΕR.Rnd+ݧן0';Γ V1T˙77T>N$wLŊ%&Ǩ|mUO:m'=ujWTqI] .|ېglm>Y۲٧ddqÙlQV+ }zܪl5d3.ap[m|@6pk`Sa;@6p\Ҫl!enl;p̒esndn eȶؕ-ˊm,p 2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  d*tsAлdpv%brA*ZW`A2qЂl#\ۖ%%y":l{/u7-B շ`["fM܁dzLt* zM۲1:ʶc&v٪iٮaԶlỴM&dKKz~] [fU ;/u7ǒmX%-by) B6Mcёl<&C6M>O'YVoٮl܄TTܬq+[ٻǒ-,vM"ָ-]]vVF$oZ6֯TOmt^5GRV"DZdK}m~SlVq2md㭃6یO%[I [@y>$?MMUx zg l(/yhӿvN|߈ƾ]"&~,[Ȗ т߄lȑ OOx'}*[Ȗblϑ*8CE>O^}<oEe x *[FEjVH "'V'Vw,[F лd˧&_~![)ƢƝ*y"i^BvMBy>O_~,dMAu,[̲l'.[n)DwiۛTțNW_D'l+R0%ե*bz˗ F!ȓTm6/JdW+N֍l!luJ2m<ք#V+=O&ZE,dk3?8l6+?~σ=W[:m[.) 'ȶ/l*nd¿AUY-*Դ=e* *ۿ el-֕-ie{/`]<.Դp˲߇OD?7d; .rn!dѝd+]d v@6=>O2mmBdF 鲧l R؋'6$|y"5>sOf)g28 ul"˽D Ow%)[peEc[21[BU>ONo]-4<9b^Dֺ'u>Iv$lJHt` gc|ys:;D^'OڑM WMN s1-|Ӈ۞sd ˆOb1gOzOvKA ^<}n9-²MLؽmNpj;\ƚ)FV.ۭjb'j) ǒmj✇J6n>f7p.I'[{J![FV.\}n4k)* ]$9lS[ӗ")٭@-KmsL]yK+ˋ8 gONH~}],[VB&w1jM264afXO_wRSP]g@.݊x[(R97l;@E>7ɧLt%'- wd݊x[(Γ,UL"lpn|gEs.V(|j'G+5|Gagn|G@f;>Iќ3Xl)FQ d[$bQ{-gٞ .ޒ+'AJ!'Cc4Y!ۺHc9!ӑ \ KlOG>RTZ\d{:Ll:3FfQ}0dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@dd@6@l Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁlʖ?z:@e.dȀl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl Ȁl 0WTIENDB`nDd.   C AFig 2 my1041 feature reduction properly ico mispel punc- FINISHED2bXb l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l Kdiza_z7b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l*w7`V-[]sndq^5G}ӳ\"4NZ:˕-O;`y,[Slnlwʖxn L2eU%hȶ'"TKsHTo혅6"eS аHT5-(do+0 KMWBsl e47a([|Qȶ9(*9C!Fa]sَȶ9(*9S!Fa,*e#h;,U웱Q"es.l}l%ʦJx6"6eS%d n"6eSG"6e%jnw$hC,Q<3KżQ"6eS%ge'P⣒-хZfXlOϢkXuW -:.@MCE~TQXvKB قK![p -˦"k)[zȋTY8!+dcFζ4!W-9U}ȶ4/[pe9Ertڱ.ݻ"Y;pۥD8U) ٖdKmlM3-#ۂـlld[.ېM"d[0-"rA6cӒl ٖ ].blLCVgCe~ْlǬlbJ)\.=T;! ؘl'5Rl c["۲ؘlgcڭxƶd+1@e~!ؖlMma l b l b e"\[-dWeeق>&g˲Ud,-`0ˆeK`lX/-L̆e=Ea"ϣmQlX؇†e mKbE*gmIlYOh[[ͣmYlY:ŖemalY/_ ےزll cò}Vu6˃9ٰlm cò}c’ذlGO۲زl<e fcޗ/vel=aFXەAmeaOwSGi#, V6,[*P76,[maUh#, VFa!lXʓJ6Y v_S62زli#,-v6"زl^h#,-V߻6زl]h#,FaF6ٍl~dvv$mw#h#=FI6ofWFx/6[ٗldNv&mw7h#Y؝ld6'mCh#=FMQ6obFxx26[ةlNe*S!xFo/]X[4Ng5)Sg6­%m:?;FQh#Do5?{F]KwaZZ!pLZٵl6o+ Qb߲NQ !cطlU!>;tb粩S5  ߡȅ;{h#6/DZwي6B䲑|ʲ[ؽlzu?!rʹ}h\l2*d8ʹ>|Xϧm V:|el *-ӕ/_6n>A65^ǣq]MNd #nl#ٮlIuq_g_C "Vlq6dS$N!HM6aُ9} SoE+si|؄mZulkO _搳`ò fwp}^fòmmΦT ". C 撳gvpsIjlX6]2o:X6,KNjE/݈}atw ~6f!aٜV(Gٰl.u6LylX6a"gufò9u9kX\Ɲ 攳9^llX6:>[Qg6,SËrԙ fwUdrԙ 6i{dTD}QW6,SvѸgSU^_còvTlrԕm~is(^tQRt o/QxrԽe^̋TE f^0u){[F9Vd @~52gζ[F9Vdf~ ͛ζBWvQ:LrԍȖh|UB&gDMv(G]نl[i%+,yqZϦ"(GXl0_VƥcSVf MS: |nxKTc]+ /8޿] پe9 QV.Z^}vUʖi.lBeklZfRg մ<[l xWl+WiUQuPlm+]'Vua մZVF=Dꎵ-}r"t)Gقl%KZdk~(ҖwG9d3%g![a(cw@9&d0$۪EhBO9jd3%gVF,\v~J!/^,f ̬0j[-ӶvAoҦg;=6Zj-u*[-У(Gقl~y%Jʍ6ٺlk-w-IrԚ-fJNϖa[lCQk "1ij۳lSH9jd9 0 ~8&C${E rԚ fJbvUaubۣlk ;Q[ ~ʝ;UFYl6TCOl\Zl@6Sr{rKgۢ]ysG-UQK#^]Nٞlklk#7PڰLɩVeZlmᗿ/_P{ jAUz9DBdY o~dkoo[fMRPڱ"۽앭a=gkhj!Pڱي\V,t0ȖXuST܂S; UCQ+oɩn[ڽ@9jd+lZV霭6 hڰLIj^g?{ rԆɖ.Ulmz)G-؀l-Q6Ϧus9ާG9:dKe3U9-Q &Y6m[Ygs-q)GوlMVlSJm[.l@6ucT#Av![~u?ۈ5-Ȧshl[jKf! 槎)8l.Dp#(i~15lFQdg# qܨ6R"?F2~I(GMSlOh:[r4vFp`+U6^9[T D9Z&̑KoDӳƂK<[Esި qs۳*fؖk醁W|ԲwlBr=yQيTr Av,R/G,[{ĤU5P/G,[V&#[vB #b2j=Z9\ق4ˢ 5'[rtOT -nb=E c A YOp֙)9f;{cË?xqLH|eo+=/vHWoqx[q?Y4IIaٜ-I3 n x8<-m6 ]d))v,:كlu?kBg_Ϳ"C,d8 Rfsy}R6fwv 7՜퇧@:?itvvpX ?Ehng:m\R6:ϭ}VrdLpT{Ul卸D][L=sd;ꬶ![y#nC ˉe/ޗ/6qu}ŐԢjӵ-s 1giR6_ȴچlOj zՏR0bO>ٞ:%'r*DچlOXZ=BHO>55UDYS=c;jOos h2G-#[ VϤPlrjIVL16tuKbYZ]S5Ϲ~oچlm؍ZB=;\mCz{\v:.\e]޸/FZ5[O<Ϥ)dz3P gK1Vd=ZlkP-7.oۙl#BKO-iN'* YVRo!vvydsO'[Ly/.oڝlΣVU ړ t^d@?y%jt5}@6}JC1Ż@6ɣR 6Z٤q::jɢ[.EtgP6Qq-[:E{q!i+nGTS}¢@6˄*\|V%ҍنXlVR2e+ VlGӫSCA*UOYΟf3jfktg:"x Z-`Xi>"Hfnչj]?eԪ]Z,Nh1rskWQ7&DX-D6K:Gꕵk$5/Mׯn1j)okՏ?gwYQ9[ZL:6~۳;,[˨U}нi]-IֆtqԪvq. (Rׯ;eײglM_hǕZ. ERGr6'XA@F$qQSD9۟.oٹl>?mDl?e=Or?v5rH͕,] o!^ G֜r:bVrQ1e-Kԗ _+"l-i! "VT+mE4l! "}sFW=\g#ߕGu"CJ):'\l-oplQ+:[3bW?n⒳+ْNm:b[6r^l-?-jtFg s!߭,k/ijHؙu "yO-j6#-A~](lMAlA;߱,`o;/zKM<1j>zZ3Kt4LϮʒY3BD~咳KgLRζ"9_o";=/Dw cW9K] 88>=yfkʊDO }Y@?o}1r_qK̳Ytdk}-KfEO 㒳EMfṣ+[X p^[3Lf@f cZ2Bh79yT]qHyڛZ{g|Uf:3j=I_i =y KAz֕=T8*g+{o{Yl.u6]i^|woY3[]o[9g}qB={ֲl˖W7Z]ܻe _t~[f=}S#n1s\g&[xqK~%qoe{@ˌ>yg:3s66ٚ"%8UvْٞSOY5C` Tq*_k VdӭH/Yd}iyx!L/0h2oC_ݒ_l~z|ۿgכ=zDf<[ofw_`#g݈l?Ow?~Y<K7/]-8Beu&pZ-oEy<. yw_]u V;{w؇ֲ*C~ҞqF+U6=9p*)-:y>-oK-~m? Bsm-g+xo-L4V/"“1Բݟd zr`|7}^;y9j{&d  oɢdlwZ#]Xx&gZH"e.#*d;W^l=9[c!/Eω? uK9['Uf9Q Fٺٚy)z[ 0l{Yې-)NZR} l}~2 `~0kb3,R:[mF_-|7~oD*͊-ȖWazs BQa 'Eŋ2͎-_+[O?[}y\d?#Č 3w.lB*F7gL)겝)h0j]WOYl)[oZ0,JC }fʲeCjv%)9 F?g+"{g3]TV>d3Y!=he匾dDdqL'/s4ȖF,@GR̮"I%)9㳖vKV "#*uCFB|/SЯL6aMr6=uU8ҍnL!%lϖA1h|XQ;>y oYxɂ,8eGue eGy. ȖVD-ȡ?~6l*oQRmH4 -v*$N#Fuj7Ž͔J) 3F:v|%S3KlyVClm7棣Ck$%vtۑ-w.emlJ.;YoXTZ_Xjo@6Srz,m&=R:2jNRU܌l"N2S)Tl}IH#1W(͍6/}co~WIA}s3żZVoFh/|; p\\w86O}\qFBl@6a={7XtEbG2g;wg9Qžh$pPҙn&g+~KNEG㺯vtm~":TaFs;M*\(D?wqy^VYw~< .fi$S$ْ"]6=ľw #۩"ǝVd+:ۥv\OUQ:lA}Zp<$Dtn(ko[*([ԕ)\0YHxCي&Îz6ul oJ^ :+[Y'9esؗ89|gBgh2d’oFB8Qv1s3g\a'ѳ0/][+lA^\ +^[x#Szh6 eG|yuKJ)[{A0km$8E&8E^ʞ򐳵\k͹T{-,˦ZPxZ n1WZFd+ZZv|\$uf/5kW`S})za^Vd+nm]OsC=-hS#!t jMaR<"( nu{']{YW "U K7jzӰ>Be ^٬ (mZjO1L[4OtiM6̓z=LbW*gQ+< ܻ7>]ĥ$l̤\V^F]6Ͳ5>|J9eݰ~1i~}֧ei>]bǫ[l~qg=16lVEc*1ϡnDHP{LjYN~ْl56l7SIsqyԢ h+[Ei6k~/LeG˖r &-lEM^QO4U.SJ&:BkΖIvBe+?U]+3 x6վ DU>w_THpZ;5圳mɦ]k[~C6Tg4sQ<  jQoe"/h4h`c=믳lfl4i}v1o)^~Q[aɗN9LU]cj˖M BUU} swʽgulW ZS|{lxښ3MT~=sn9l>?usIG6Q%ӏ|>]ah˜6n}1ZJQyv)Od˲%Avh{nTO)[30`r̮"gc1 9TqIa}26ee⎶"9mزl.1a_q8LoRƆEvuIю=gL%u`pj@1朳ͳ[̖es24}#gSr1`&usl[l*\2w)OJ[YO摍͕\"oQ%.:[Q®ѨΆlopn4]F] *Eg3c :)~r}ŗ_l;UWQǧ:ij@6W?~̛ՄR4˗Y6 (a2j?=/BdiAb:l?a.ʶݘލDųQ̮2ld tt73N Lۚ_uu#dsƌ=>fMXfٜ1lH/Ueܓx&T9OoK6ݶ3=z hG6Ulv!B3>UJ Ջhcތ@-_'>"idMϮRs1썔.eTr\SI1Yd9]U4/Noa"fu9P-ߦ> 9SM`LN8y:i8^w;2+O&O ѨvLms#!o&=?eӍdDds>wk$؞}jvB76T{FBrY厐/`{nL 9Әj85twi/#nCSvLڻ4Lk3v͖>@ BFBd4.ӊ3Ng5Y1ds'je 6U43N5Fz 8޶PDŃ*o-Gg\FXf/OuudT{FBYfm `r&Upml`B~m `ϲA nlq>o٦bP+EwH*y d5 4TwlH(jC(MGkp[Y=y^4lQdFBC lBnmJFBof1B0}AIyj$ -JS M bE!3lHMF4Z Gm.&aLhpS|NEdfp/y&.#7BlӘc62mᰙôr٦fhnɢc÷BlSo$MͅP" ܃GñᛛlsP5RvkeJd2-NCl*m+GmFBKi/ m*mۚl3a ??[']'Bl3Q4>qp2lsamSiT%FXl]OX%͇ ngN·4f:"6|[!6#پ\ݗ16'qof6+Q_&6|S!6+qofUiTd7g-;2u6+9MlC(@djEOEp:%c算m&-;2ry6+fK.GC͊g;wne|f%lSji.8eoUv|dKUֳUl"8eEȦfW] _p[eW?']m!.S62;_Jb"Ny*eh ?O{tRЮ]lņ/22h[-D6NE>||l\6>f5o!{tuXQG6JrB(657)`d _P9ݢ=Ca6_nI6LE(=seDXlBKf7o1!I.۞e3ldQoLE1zhdDXlV[MdžTBD#9ma+$ m !fZW*ikp0!QU&Je*mJ6ngJ6CZ6V*[5TJQ,MgDXlʢmɦ*mvs `B-Md lU룛b蠤X-ܮJtaceΠlTu#z=5V)hXUɦJⷡ#ZŪJ9Ƿ;&ۜOBc,)ܻʪFVZmVLEH<-$SgDXlg7K!'$znT4( ː65يr4%lO-P,H]LjFԡU䏬 v7*[X{{6Eﱃ )g_\R6>yy./R۳lslGF$M.*HMbX!WMk߉+el"Rkm!âg#=¨O/,B(HMO#UGrk eUsl"tgcLTо<:h|ZURnM5 lZ4Eiuo3;N~ jr߿%j"Rz<[NtCl˲oc>T }++DXlC5/9S)=¨zs?TD-٥iNp̯JfDZ޾FقGdӝjl}lR@_AGNH쫂JD؂l,eTt=tzI^ld@F٢e6|9H `^?ww]-󠴿0nn8x,/2nfQHF_c&KEEh-k׫Mf[ z+΄Dy}ԥM5ʖ4|ZzZQcXe8Qt]N;O=t!&e` TEީ$ecvոd& mytҒF DXlAJ*b_/4ܴ=ib{z}6 Mx-M٢Ɨ.[]˽8o^QR=ɭ uADSlBKa5Ʀ?,C|VSx-Sw~r?>:ekPM%>ψo :S+BXSՃTTיh _]&m@6[vgض_GZ+j g5[LuMֵ>B+j  ӻu~4&vfnP~mO[r:$@Lln)[z̿ Ǧdcֵ_]z{9 M']#Qx7E7+6e%mrq*m8m^q=~vbdf 9N/MLiĒF'@V#5d;ZghUZ/2|=#N)/Eh]v'{W|,J}]xHřC֟cB.vGS]}yoRvV; }],>S^f{2i/Hmba dk.ԍzd;;V%lrx&EE/߁窚IpݲN--UOשwnQ&wgįy1/1!;07$[̷.ёLgxWCWfde\e{" i+H/>X7m-o-VzuH,[lUfA[Aj EtfIly_3hy6\=7Z djO}ԡslmb4ΡbyӒlmlU/] i3ys~[l홚g: d{mWi)H>ȳfm\k>_"U{Ӑ,F '&5 ҩ-pQcWʖ6rd+FSA:P*Fx bʿuN`[/{2x6)SA:Pl:ZkUtݼ̿VRYDNٞ ɇ˪>S%z%S3Ֆ1p v/L+r Ar) rIɇԭ\k`"ړ'_EȖyy NK\"يZ?u«2LW.C6C6yKlS S>S/6(r;Vd;o߭)+d7P|/UusIR2tdD]Gh"۵]6nVZK&mS]U5;G$V4y}2/[lUj~Nx߮|j2S;ru-%J0O=߽h%fk\2d .-6P|UWJ5'.B[p '[Q^!]˸vje+ қCTkgKK -&C^l齢>6,rB޴ΡU:يo:@o[Wq]m*[\j?3+w, uG2Pt@5|q+![dU]l{8+/u!po,A`ee8P|r .oYl[lE4P|ܖ.(-qsưRA%'ѷAyWME[ o ۢyZ-d-jODyޜow -d[4-{W}k-d[4me#A˶agmѴTTj㠋vٖM[󲨥5dDm@eӺC=A-M>Kۋlt]%>(~;OePl&ǫj ۲\ Z3ȶp:`=ꔯn#0{#[?Lydٜ:Æds{1:dٜmlc(V6mlc(V6mlc(VAJe6oؐl_lGۜdP*e,>[aKmb(~4aCk-eܒl_ly+xr#Ck-:urnIM ůQгk-ɶ5Q- ůQ6_v~ædP<- ůQ6Ov~ædPJeޔl[_lDy m C+-Uʞp6l`(~]-_㩱Uv-ڔl_lvB߲ldPe@ oYMDYPet _}`ٴFmɶZ*exqfCj7`>jX eTبf aCo/ݜUϦ*45IljXPe fX٧t7j~(~+3RMPl:V([P-Vc_~-٬- r㾕m(~ _[.[Uݶd[PeR2,:kQ?5ʦ:nw˘oy H7&ڇ([+_FTܸ5%<%hxK*قڇW)[0Mڋ Q]ydC-zH؍lvnLůRyS噛-7ʇ)[%%Ke~]yʇ [^MVEe7hQ|(~ ĖyζoA.[a:6жMvm-99;$V>ºlUQrX@ڇW*[cW<3v0sc=t,AG[[-[P*e{ؕj#sɶU+߷R6?Z8[ h[Pe{ܕP#m-mCkqWc%{Qt.֠_llwGE9dKW]ly{ATXuUůQ]J;Ǘ۔mCkalw.exygCMFϢkFl% kSe,=ߗN#[fYYvR#6̙p..ad{XŨ&'-Wu-O"%wWytˡZ-oDcL89]*6.ۭ\IFlǶ|52|ˡ\:q͝.VΙppwذٚ5d nGN'mz[E߇({mO2jyv  |*6.[zy8d6V oS`yyf@y.\mK N19]†Ev[r*>S|(-~p\jkqpL_>%lXlYc%Z%!9= Qft~#c'-b(>O!9~=fNV..!Ͳ8o[.9Zy뱌 Muǔ.Edu6ۑDh{.SűG0aT{Om/Oʼn7G"6,\.[~낣w'>-b(><8hXD†vvdK3kƒl _ޖp艼ַc k!ۭS.[}g:RL%|}[P|uO_e$YlJ]6_Y[>V%|{[PZ}}2,@]lf#wԍB݈ݡ&acOO],HXٚe wq:zu9.őc k![3M˦Wlʖ߈ғŁN'O;X;|=f XuE7d-0]!v)O88qz,"aRe6eo2OI\.CS",A6]V׫e;%XnO|^>8!.AP)޻J5x8z E꼫?KfW>G[䀰 ׎Xl]LoTr !|O{)5=lTuP }olͮh##^ۯ-<:[wz&Hx; b|FXumW7.'7"uv!a30ZTbe-W3`V㭗 3P ŇK%j1<=cP|!폍ZaD}C3aIneǚيxo ag%Yl(R73P;[šac9%Ylzl+]Y˘Um`faZѴy 靛7jD3R͈ש1q}35e3,]Z[ƒ_Üyp&lYŨ8ņI [SȖX׌l=+m)d M]u!Mvee 2lnhg.gl=L$A6u9 vu)/lL#!szuA64dsu@? vszUA6$3'ٶAmwlmwٶA? hnT6r!϶ѰlL$ LζqUȶͰݛYlQ[Ml[϶#Sg[2g[`ds#澫[YllyT9ۚ+mFc:f3g͈h>k 5Fcx6('ceK\V1hG96W3zزFlMV/ZV)ʓl >V)՚gKWFXlVWm2϶6:efիfJM}d{ca]Lg,vddQ]aKiهl)d5 ,vRddQCv"͎[lNz4e{Wa'KSȶlj#EPxk=m9;͂RM=^[~Gۻ ./:xˀ~{>G# km1Fe@maOc@:[žd{Gma-+FUPu6d{Cma-ȷUM669[E)&>ZbIȓUm2ƨylwn/n#Um2ƨ6lw/r͈**cT"FVqY 69[E%.b*gdIGybGVS}:[I%^> grmȴTg B݄}Tis6km<ۡ k]g.w!~>JmYdӉ"&i!t?`.dQQIc@~ٌ~YVSmg͔flԺWo)ƨZazپXD6ǫ6oij`ԴcTRk#L.[TWD}Mfu/6ƽv6HqplD؇l1*agT%teZ6v"^)ՀcG)[hD؇l;1.Y*Yy-eKb!>JilѽQ<|m0iV dF!>JLqzg?1{U6i#L5Y5!>Jtq@/AVͪlǫήS$Wk4i"C=}7 ;NZ9Ʀl['2g%ZC=?[E6t9[uhNdcQE1[: 7"^7D9[w@3D؉l1*1cWXCd/)&AxWGl"Dtl"D}/: #wj@@34vrlYaph4vgӲ]Inԁ&NdqGf΁&Sguaۘ`:;mahm6Mv@9o4.m9&*D؉l{1i ȶFҶ{m1FUclm(zGw#_7yii nh{m1FyζY4}Mv~Zmg7D؋l1z\y6Q&ndsǣlm[0^:4"ۮcdl{+4ݣ{maϲZG:n6*KSQ2wPlK6?/.ۮ>d+7z)ۨ Ȧz=Q74/ [Alm;Ӑ-tRt66 HUJc n uv.[VmD@DXl)T mnuF4Z-_#_]^~W_]}tb[3]:t4/7,!/U nOZ3:F/[_-7Ӑm1F<~_B6u_l $ ]^:Vlc[ vg]A!ۮcx- ~;ohG/wsYx1f&7d;tVlSM˖zS4%:ƨmem|@DXllnZ\vƦl{Vl/4M˖UygiJܹl1ngJ@D؄l--5Eꥆl p39J@DXl'-^uZ˦ neG-=¨)ۮ>l(ʜ2~Y/25 t?kMh&d3qEF{QS6> 9&e3u4dVg{=&dG5e#Â.'Dؐl-=mlTpMnSS6bl{l-7zIg.O/nZCrN7x!ْd;vF VoH̬.UlNW߂lѵ!۩[6bl^b4~{%}a &e;wF 3f->`La]lU/gŗnو1att]޷8TdѲE=6T M]޶ކBfVkn:#!=7:Pv,V ׋~FP͎ lAd~=6lo ;iy /uF7e%Tߤ:Dudz!I.oZl~+-SOvGcNMlAhGC-h u63}'53Uukv1Fo%D7nF1z/Tu Suj- 䉪4jDk}c^06ZT( Zňe#l!d#,g#l}yȶ vN-\-l}ِ-g#쬟 r6bllcFcFVg#-g#쭟7'.d# ;xbN6>{@/و1z+;w/و1z+;w/و1z+;w/و1z+l d# 3x'bL6> ;w7x#;w7x#;w7x#;w3x'bM6>{7l d# 7و1z#bM6bd#}M6bd#}M6bdt{7;xbN6>l d#} ;و1z;و1z;و1z;%xv'1Fc6 ؟l} d1'ao@F@6cc6 ؟lFǻ؟l=v{`4G/bc.v(a@Fǻ@6c. ءl} d1v(1Fb&v(1Fb^e#]R6kأl=FǛ@6c& أl} d1(ao@=Fћإl}=FћاlFGو1zav)1Fa إl}d1%VW)aaZw>Vl cy*;?[Bl}˖x:?r"/+^A~bϻ5i+]c/[J[Cnѷd[-3Tar"uF{Xlyz,dUzv{2`V-}[l]j_g#ݰ&ass^F6be;d-۹zGKXlQ%gkZJܹnو1z ͿE,1*閍~NȶV/2*v0ꑍ~Y/MȖdGF[Xllk`#U=¨G6>†d3#d[[MWњFݲc6 jZ&yf>j>n/]bWxJ6beAOx1aC-¨W6ˆ)XlspLd#_ /wгjF]&הaYlV^̔G-¨O#ޙYl_<=[Tw/zd{=e.h 5km.}||63-*&Vݺ==}~ԤQjnCsӧO67'1UZla&g5:3OEf_6dE׆l'Kٚ1Fl"lLZK}KlMl#r66&[l-۹[%w_tm0e4dX֌1"gahV H?|K_;qI [9`-[3#ygfV,8$Xlz缾; H?>fUI?}M˦'DGlͰx/˖y?g.Ajۼ_6_S\2%]CQ|e P/[*H|_𾓲^jvU3(>y)dE%LͼFQZDYlD݊Wzeyg˖iYz*YcTLi LDXbT1 U,[_dW*#-qbhc-Z-c6v+[a:l>ssLƖ_|)d#r6v+.2ƚže+2}HʦbSaaDz]cg}>~e .U  cU+@ W\el1;W'f`E2?g$ӳ_قb٭l`EV-#rr*p@U6٩lERNNe dc aR)+# ?&d^ks?caSf a2v)Jأl&c{+t#e3Yނ@J_poR"c{+dz}FVdlW02V XH@OrbU] gd2) cv'[=l_OAN{Ez e}|6Ն\@cٛlfSzjwҗNٙlq}&NHI65qH/KaKdƛU9X%[#c1o,C^;eW=`cT;}=@6]جc*Lg󽏗NٓlK19H$CƖEK+/꿗NّlO+.Ul/#ٞV\p<˖})y)1lc؏lK8}ܯAE}meqlcٍlmkd /Am{uq6ֺ[GI|awv"[91HhL;$8eLWyQpcŦjլFVFەzی>+OFpcӲacF'NlZ/#ΣZx ە-.0=Wq}paEsGr6?خl;;3*H𩔍!R+[41jG=ە-6S}nj1j7J=ە->.ƨ mk6,xN2?ٰl}dvve-tUtر]٢*Daveo6˨MS1Djve&;+[leÊfM$m+6,v1j0?,ذl69T1DjvemQۧ1veƭclWf-u`tf[Ǩ?ٮlVSUGە*g"ƨJ6[˗A&1*c-fMQ[\Iec-f >||@KWԉi#ٽl}Tתbm +tel5v/1F%Ӈ>be lcII)9[e!ƨD ~RgulͅJٚ^9bJ"Oy#V9'g{1*@g5{f (j ۬%gl6S؇:6lz󌟰&%VWd'\ar:%PW\Xl؀lj֜ Ud-CAVg%j#[&+KMoz3~šXlYbd Ռ&lelM/[-eߪW gf5~5㼐vXm2gc]et`#л;ͩB">5aÃ혗A6gQv#VeEt2 mΰ3ogeeK"J(옅UYcJfd\d+kH\E}ҦYljudl|Ѳ% vcTNC+bU],e3#볱a-36eSFeG6gG]OMH5e$[lŚ>6&dF{ls} ,XlU2e lԧ؎l#=V9>>_)++6#U#!'-ۘljl`)ۜ1F|TLMV{ V7JMz«_x]Yc+Ҷ~0QXPߵV6L'Ҷ sc.۬1F%-3{I5jx-^KeEBm TDW}jFͻQ!T6 *MKleui*m[ͻ9dKkt%Mf:aՓ{,V/4'TU0d[d-Û_]-v6al>ٖdSEhQS Ct6c]j#g[Z-9U}ȶ4V/UWF{slKecked1WSLXS%l3'ԑma gKUz+_'ԑma gK-ueUtCȶEV4C֞'ԑma g b l b l b l b l b l b l b l b l b l b l b l b l b l b l b s#>cul_a(]} 6X&R[j>?<p6Zd>f5+K+K=uVwR7ĻyD wN8P uz>ސJ8 %Wǘ7鏱`%ǜy~ԽUPcp~:a#eXsXOlo!>(?Q\Vy2Z+ܺ`}u;[&ldKfpSӇ@kd3(. 3Z&d zBu,Xl1[aO_p0vaC >&ՆF_ȦsxUmap}Hpc~,o6a#e갛9,ҁ꒧|,NkpLX 5cJdSyCxV}E-}twl*3̏w">o],,N*9̜|:\N}C'XlMߺC>:~"]VeFy˄ oN|@; o WX&i1efCH˦oݱx]$)~[w]B=ͲJ(op!]ªJ~5?ps!e3%߇0>&>חYB%ueQ}o]fl2aejbSUX%{n۲欟XtQw"}l'%Ϻ=aKX=(3; o-LX*2DŽںC6\ԬD6u,8;Ȗe.9ll<]e;n[l|f̱eomNɖiֲ3[˦$fm.[l'nvݶ-[ѺoDfn!n_l5G˄sp=c9.D@ݵope6:,XlY(Ywo'E6c+[rpMn)[l-CilNw=fv-:dpl%E:xV65Z^6Bk+E7deKPd8/r2^lC0d&w&me0 p"M\,Z6+elv8- Ma2abdpztYyfYz&|(HuXl:2?(琰"%\ozś-fYrQ,µl3$lp\-fZx .g˄l]-fBnWc6!!2Or>F6}| b#UZ`}K1F,Y;;  +!wxѲY𒷕/zu4fwx1YRӕNb뽫TۄSlv{WzS, ]Z]e0+~W>u _ؕT-wsIlgmû0; 7;`N\62uTʹ2a%ۆUn_f2,N]hvX6NVmx'ev\/d#s#^m>X'촕`xG~r8VWw)( "}G`pVOvX0dM~2d2P`/ӂ5KluPɨK8+  =7#zZ26X`pX/ӂfn~hÛjtR24tXfc;5Op1bpuL2Ylr @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @wk\K٦E{ûVMEl/dߺ_Z96-o/dgْQΗMųl(dk<-Bvγl(Bdk 6 I4M lSl T 66ME)Ip$ޥ)[e>;?UBg\M=dlw RO˒=JK("g.TR*]п>"dSȖ+vS*yl5.^IV.{UxӲ_ClwN![lR*]wvTY'QIiMj̯vɨ˃lHYkgQe e3B;vW(u~]^lP["ˍ lO˦MGOjX*6l&%ѩ.[nKtB锋d~il5펑-6E-E=4G]M7lOVlw펕lQ)?,[l NXslW5P?yqx2-oBU^5O˚dR\(B]c*nyw@:͉ܮkPUԯᆲ)A/݂ӳl^9<TzPωպ2iP#dSv3M~,QS)UەElPᆲ)A0Eʐkz~-_R&Ѫ#3]h2V,nǛ)rݩ˖ezcfI.٦--eM׮ 2#%3e)d;5GW%[f?lfM [dcd ADyE櫬--e]l_ݩvQ2vqx=`d; 1UZ9$*زZ\MLd+E;v![ ۩&YFJoݩ˖z5ِdS---ٔOgdsedӡE8 \@;u٢N²Psf ݩ˖ e-qkf)d+feJb-EI3~6+A;Mn]dӐuɖKv1(|C-.d˳-s9eD}4A;lfT=R+dWf#KT.RdB"F1BGN%[;uMI[~DZ-=~ 8HNz–k|K}K1_VTJlŝ'\JR&z:|w}=ReŵMYqu6TJRx\1^9}6-kv"7EV'|ʊKJ8_Il*kk8EB?ޗ^0CBdk\^K"/s2R1ϧ2l9k-߂'|AyVй9*[+l5- 2?ҝ?vx'ow7|Y,:͆Tx5v.yIJڿf{J6&;&g{L"?{:&vߞ --m)&_.ۧ:;8הsd*~SWPO-jӓWRͫ ?>=˖t'YPr향mgekmIU5TMӹx.6F]? * Jlޘkق<3;)~~?Nz Gl-zNɞG_.eZ.sczeF65IS[QWe]9,n?ϑ,e;<}TI=9 6}x}4)7-o.^@C%KMsooޤl^˰.J u6J)Y%&oҍ;eըwMa3g%Y?Á-p$>wj=uާOO}Allsd t(Ȧ9'kFv0>yT]TveSiFEmhZ>>~A~]ekZ)Ul O==<.Up8-o^s#d:̾7h9?]sT{ ;+ֲw(W]$(=ɦeuP'e_-F:(B#`/lTydbeݢ9q= d˒v[eF,ɨοϲv-7p Te-q-۟l\1dSlO\e+[:6pr_*[HzJ.g3BUo*dB6_@۫lmu61CΦ:?d_rqŌxG,Z*,4lCȢ5϶Oލ Tl-]Q.[|Km#r򦝃l R_O7Wl fO.XTȖN [M?_-M%tzIlMdӹᏇLw*.1u.[MֱLm&̸w-˞j56TZy㛑-tFuAkdv'[kԇփ6ZQqyz8V ul)i;.=G}d v'[kЦߩ?5cz}&]FUt~韝9N׶y껥z|ݟlQK7V[~MRf_вW[قkD?9v'Pif%}{ζeF|M'enc`fgT4na֜Mψ??;'2Bd2kGkdEjRX#.Q _Qlmk}PRUZ2lV1˦ qɹ',lkn:x2ZO^l϶+glTig7V1R#'[e@E բm2LgA^5-[^MirCsg۱>[xQ3Gl*Tgk]효:[wdk]yº.sz^>,s)Lv*-*l\쪛+Oz Ĥg=Aʓ/%V,MZԍLpgxrH.[ղ%^#(spOeh;/HlL{ҩۺn!7MZV ]ߥl.M_鑋D.#-4sbN.[e+t~|3ݍJ٦tZf6t1v˖nzX~d{AȓMln/a=FK<+ejq;ZaRm+Kfp[Vbg5eSiϰu#[V1?ߝlm+mz*_!۔*̒@ɩëjɬ[VvMObt]v:j&XN{Wlow'[[G{')Sը7d Xp=:l])_2̞:Luڟu׮|*St {ZwS'*(j }ئlWZMKj[y5uUȇs ^+z]ɦCgTw[4^"sIމA,Z6;Q\!6F׺Ie`wÇ (=1l.`dEtdS6l#/`dE9.l0Ȕ~0 l_b 6AQ$#R6 /`dE:"i )eHދAq:L*UV0|#zGȖw ȁl b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l bHo/]]^yKeڻ16emd/ʖ\Ve$gx86QlSFvSlOl+X蚧8Ot٤lIMie Y^Nγlg6)[Z>.[t~SШL,[pI3gٮ>&e˟PK~0lN/[0i,eć٤l~ke{6nRҎCɖDѦly|O\ؤl\5ʖ_*n$[8ZE:ʇd-e$[6eʖԲ&8;OŧIa%ޯ~l%[:lI.1_>;O}/v{_zwq.8l%VSߋ=[eK+}V!.s-eSO(C٢l=F7-'%-Sȉ*-,ES![t&ejׇ\`t7il-Q2nJ_"[{V5; O}/v~lW.h:eU'r--,Q?ρ#S)[G/ bl?snƏtN@eocdk]S~D5>>CMC3)s͒qϨvՃI)E]UGS6:')s/X1V>6zTr ٲ/|idky0M'-{룝'BuOe&xlltʦ|_xAc!NvlF6!-t,/M;!>t $ϲWکd+#5!#M_0၎O-t3f ^D0%tS]F?eN+ 5uB,fWSⓚwe*_6EO>oTd90b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b ^ޏDIENDB`$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5L 55T5T55#vL #v#vT#v#v:Vl t0^65? 55T55ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o$$If!vh5 55T5T55t#v #v#vT#v#vt:Vl t 6 065 55T55e4ytH o'DdA.~  C ZABComparison of data sets for SMObbUo^2Ԯnl1Θn)o^2ԮnlPNG  IHDRg`k;sRGBgAMA a cHRMz&u0`:pQ<PLTEٟ pHYs\F\FCAtIME1Xh$tEXtSoftwareTIFF Image Printer 7.0f;Z)IDATx;$Iz zSէP/:%ё^R%#%Tlɏ"!:MMB}Wk`RHgQ0RvSYu "wg:f1k#|.um[ڥ{ڿtގ0puf"]7F |}N cF4[_$q c@ HÁVTCyl-oX@8_ t`|KBVoM~:o1졈9"'ˑZ%n@9/h;o1쾴-ms8]7F |dTǍ:o1Rh¶]7F pV6ϻ@%m9&ҹ6]9N,tzcYp1l[l*FO+V~w,G1Hf6_~_*Kȃ]1wr1iv`Kݖ yǶ6czr:96~Ƹt:uYp#e hc :nՋ9[t5&*/F6;a hE7F:M`1Vl#%ywx.O0\_ {bD w]7FR'I](yXwݶvQΞ I -8HRQ[{(DgFFlӎ{Y1 $a&F!\fgw:on[evumnrŰma(A ]7F!M1ސ7êNz`zz Q7`co"q[8um{VEt!I*2]7F'X+6Ũ$H~/6a7Qu ŝTܖAOdHfU7Rv RH)v`hoCDyŮl#/&VKwqRfo-a{~F_LՀ[6{\(i`Es$nD.ov }[ڟ I*u`~1It{ z6b:Du[3P%A Il#1(szmfHi &';G:HRaH mU1>nĠ(9\1\$aX6bp۪1i Il#1,sn9,le}t`a17ߘ۪1e)jA ]7FbXL8Yܢ$a6bwܚC arKt,t`Q1 :n&݂/0d(I*w`Q1q6ߘ꽱7msycTnPT6bJ-z"{--w/aHTɡm`r;4$nĸƻ&NeիsjTnn6bwܚ4xRo] #1um$$q˪Ȝ l 'cpcr#4$Fn&Q(xUmYq>&n"nK6 #t v` N1%29u:}LNv I}c]7FSLo"c19: 1um$xo̿~}oKv#*7T4wm$x4qd*X[Օ۱61$m.= q݆[#m/= vMmXycCA4bѥ?ƩPYT"m/= G&9uVXGw@fp)VN mc.H`KO-7zɲn{rP-TnV1$hnZ'm$$snܺ >|:5&l#/HЮh|0Omb4Hhmh6b_y5lAcưD9\u*7zM9H)4a:HҳMy/'=mcp(Xs LN I6X l#1QLԎ[sk[poR(&n{rx}) IC+m$&Iqx>~RLN6C+m$Iqcys~`rH9VHLVv^@fAj*&JҶó:,dtc%T1c6d]0B1VHLSH1*(J`b)_nlC{a3HҠ+m$&JDLN6kHJ`b"uZ&dlAX l#1]L?QrA:VHLSD)`lé' 190$c%t1r? vQ H +m$)x?> d܌"HrD)&J-> i{Tnf63D%|}Fq$i63D鸥٦{19$s0abtܲ(Wr3 I5VHOKIdl AIFbD9\}ҙO&'VhFb(o0%g܌4H@6sŔQm+Qu+m$fMDlj`rfIA+m$fqKtLN6 ‚X l#1[L[x ?|Vޘa IC*6Dڌۺσ[&XFb9\c^Nml$ Fb9\cӘl$ sFb9\]L݆N$0+m$拉q+n<2LNg It/&B-_U ? Y1>+m$D|FP b6𘈕6bކ/4Y8H6b"pLNd I7VHõ>QYLbr&K,6b"mȄTnz,$iJ` Q16ߨiO%UIFFBTL[&td[c+m$DD︝^N($H06b"!-u%HЉ8ܓ~F|(du]?mD6ߨGR7LNV暂$ X l#"FEo_)~d2W$i`-:&'+ lBl Lar*W$iHa b^Q7Q$mtĶ;n0ԌTnj\!l#C$wHZbr$i`m7/Q*\e!mDQ6ߨ-zS t%BFx_w-KLNZ$6԰qCwrr/r7HR6:s:+o]Z0QFfG:[yH`k629\kg_uVޔE D\sm$(DZ.Ҷ }` R1Qs2Mm7ɦ-m:H€m$HD:jrNHhwԭ np j` b1Q7&L~䏺UdrAFXL7ajrrnB?HhqeYț 6bvܦ$D%m2£G?HhPu:a2gԭdn$9\z{8u7\5G# Hh̘pzs/AF\LԎduђC 6&m$m4DIΠ\AF^LD)3qpdJq F^LĎ`6h-A[r$9\o~P5%}^ /j~F^L7(o^,(`b"Nh4Lڻ6D|c>amI^!^I` b"vܲٗ 66X 0 FD7X61yѥm$$uxXkʍ~'+ l!SLa[$E[rL1s@%ι-Ǐ l!SLL: 1ЖGѐ)&bǍ0i[AI` b"nA ^O_hK%HhHFl~]$z}r~un=a bu܈JEIiG7іMѐ,&bW0?t!B[7AFC9\iʺ6g]і-&Z&anck.hK&Hh-kJzۂiHF~;/+tD[l#"[LĎ-<ݓ&4$m4q&LNz_&[6D븉5HP.&ZǍ&aі,| 6DJ'n{6 mBUѐ/&ZCt%||%=ZP6"D|瘨۰ԭ*Hhm bo+Ж, 6DA$c)YP6" D&sۚ-SP(&ZWj\Vޖ)Hh(-8a27OrU/uN ` bp%~XknK$m4Õ hw/u*HhF nⶤԾ\HFn.+a]UP)&ZǍ&IjmI$0` buh4U^MUP*&RǍ&ɂYڶ_AFCH9\i৳_k]Ql#TL7D% 6JD|& EMu.u+Hhi Z&U\y:ےEba b"u(aPs1gAFCH7J&ƶǹ-YP+&RWJjkkK%H(in>5m__WJς$b1r$~[´ I` b"p%4ں0M*KP,&RǍ&im\W[ҷ lZL z&$m4TqFǯ^zI8պڒ~-(`b"u%l aݶn=(l#ZL8P[lEmI$r1Qa_W϶[Wͳl#\L7ęlbMmI$r16I~{%W˻ l\L7Kdմ%}[P6"D $O>Ϩ-_P/&JMݗ֒6 l^L0Ioo FDX׺i7]{ u:&$m4ԋp& mdicES8a(0 m5dQڄx_l׋(&M owUmxW?a(&M& :uh[C?mB7EWAFCʎߊlTio{w]זJBB|Qm$t) ~*m*LR/W+ m61HhS3g/P\'\Hƛ|AFC^]Y5nζa$ tmEn5\Io2;:mDt츱_[錷;%7Կ| 6Zs$(ۊn ,&00AFC`ú]ӄه' &u['mBl{Vl#ULnOҝK"L*!?%?D>~>2HhhS|y=3mW ַIm 6) 6z4[͙"?O%jڒ榆<\P6"zM9Վu׷6M lWL nL4zu.nkm$~I` bJGmA z.^]> nju FDӶ^&Qﷵm? l{MgдoD%_9- $m4m1@6ծJ$;\P6"66&}6O73I`ۈ:@;Mm\㨮۔[Ou[R O$mDiֶ&%LnڠBx:%^.(`[= KobZowYۆmx.`IA lږKUm]eu{^Mkےfp\Lm}bb|K6?md[4kۥ6󪙄g lՕOHݶL~As엾5`жȼmm5ulm۷?x\B_:i{29Or0ImG: (6HۢmԼm$|g$o!A6-8fsPowI%t@+2mV{x.y 74u lnՃM8DJ_G[HnhAֵ-TK8臻跉#6g50lm[V $٥1(TJoߣm%NMoK>6Noum I`[l./sd(VO-AƷmX8M&IxWlvB ֵ-om݃+D!\zm]$El# (VOɱ&Aۖum{hm.pfW"\=ǰ ؖnlM`1ۄ׽%Ϯ" ?. 8M&UOr?^I`[}7ҶZ9[Ͱ z¶KƶK& %k] ~IVo[mlVRZ6a&M69Ig}$msz6.pg0Lɳ'>V1|^PwL=PJ^6nP=%1mlw $¦AI$wLZ]&^6ѷwWu^ey^P䲡uc Q8MTwunE l J*XR  l~:Qu[y@=HJxm[ $ZQ|̈́jz3q1/;r~ĝm@@w[n|^Aƾ-Lm6xCDSmM1߃$-fǨ|GMp& tǿ]~K{@۪ RZVTI2ιW8M H !!bn[K7q?CۄaKmStvl[U%/;Em&\⦚"5ۦ8m/T7[rVi#߲m̶mwRuhWر;ns9xOso`w\&Imc3ayɰmv,l*|v3 a'm;$.mITeA3yɷa VpMYSDU&;k|&K bv;g*2gxfHyɷa?|Dzb[6Q͕ n¶5eWc(P?(X`='9|s܆m6ft^ζ,j0LT1{4k JބmV͝t Y[DA^$>&K|I^on۹y/m(L"J[8/9}VEM-9ڶ9=z擬nB7r oo^r]m#-3 TK%s%W&vݺͻd mt:uVT"8Vi>}C/M5O|}gXXrs9رm,[CʬIAǹWMvǷd mVܶ(h|E%*2wӉn;Kn[2 KnN:;-n39m $Z6Qn&dZYH-iSvmwA_;/:mCaC<| wi8sc۲v;mU:& o;m*֫,_o!PZzXmE6Xm~4UMȵιu~&%~5|'7bsk&xpo[6AD4o9GPWYW %F-{,k-~6tA?hƭۚN[OmΑ%FlHV%7'*fDaoUUЫd 7|/X X6QL~.ny sǫy7lfmlm0piOV~WnV6$l#mN 8lÛu[Y64fNQr?(YMtR֪I` ̷42goz%D5$a%l͆{2S7~%B! ۨ,a(L2gLVT7k{u;OHUl#m0LicTf1 szzuCQY6ѾRaC;M[ IPY6Qdjnq2uh||ݶ^1: >iAlir_='4nۖ`elIqC3˼[H[ð2 &Dӝ^ܐm4M hد{ous6I/5o#l#m00(lҰx$lmL@Qic4c~Knۥ! h,d L2N]"UR~$Kn[viH6 &4Y%6c^K@Iw%N:ö 0 IsjHp : IFc!DKf՝6I1/%@nCX6Q$LoYRC0l#m|EI6F3[uR $PGF^Qx,m$lmݠ$V4(| z IFc)aί(:%@ڀm$M0(V^dm$lmL@9ո̺zI7n ok! h,f LοwKnX̶D4q11O-Ym$M&IۼuN#6^  IFc1KܪErqXړ%@H,f(L r6F5pO?6&W|Iv*²),!Po]^Æ$lm0IZVmwF{Fb9 I]}C;k+,mwF IFc9ali{N;kw"6x}H,g aJ>^*Y{O46 6 3v=P)̼dyhܐm4m.LrFIgec2hĂ 7E9lIsq2V6 6& #'ac}Ӑm4M)l @b^sۼOiH6K6&i:mêki`+,m'y[6K6ĭ oiu6ƒMI)VIH$Kpmy[y IFcI2n$nGچcu[Y7vm6nCXҶ*Lj U}j ,!PݶH,jK<-u‹֛wfҶ}mu((HL*Yo-%# Ҷ~g3GYSj?!YoɁݺ@s IFciۚ8{&$zi^-7%=:am ^u[JiR[nJu! h,m[UM Jnao`MIm IFci۪mF͗wsSoGtzw'6KVm3ZHxs*[mJ(6KÓM9]m'zMh)w$lmM9]USn䃳lJUP3y(g ^ Z6^wRA;XڶjnYT@{XlJm4$am6,wzp%÷^SnQH,m[m7 ,w^'5$am6u^iԳUo_1ƶה:P,6Kܚ$U¬C7 ]FbqO>$n7(=VS뼭†$lmyWmqQJf'5 [o3VSn[$am+'˅ݑt<@JmՔ&N H,nݫ KJzmv^'7$am 7AeYP ٓ6C4%N$nH6K ##zmv^'"H\m%'ݻÑpqbmFYiJАm4!z#ٕـ߫x6BCXڶTxA{bݖ۹\@z%榅ݶ9l#mx^΂ohgJVmFu hٔ|9/B%i'/;f3czW%=-4%}NDiH6&O?QP̌Tmu H IFcqr蘆Mҫ6~y[iH,nj=5k6xSn! h`R6xz &dPnJzm5$a lSrԗs^`tSn(` lrWkc6~馤 @ f)qVz!6~ J!=fn3ה8o+! ha/{Rj;oo+ 6%%ͺ4Lm暒v@!l#m$-9I8.m&mƚ'7$a ?lxaZRIŎ;PSu $)GV3F8cz)x[a)oV]` Ol ˮ^*zMTf)oPVm$n:gg %B)7%}MАm*|V%A& l\f4-Ȑ;4SlJzHial+RHԊ@NF5Ԛ&Ral-(̓'gBy&~j@O}6^$A,Dz 4%RӼfFB=s{x Jp]j0xH &gZ|{mS 6G_9xAIV.bl`ruDMev66>ee7x-=ےӍ2*+?mBacl-(_e~%D6Y)4%충NSm$< JmhfjdM)i ņ$l?9 8Ymh+:Ry$SRMv䛒>/m}\9|G$n;'K궟M'˛z~mU^H$Ήmr?>>m;Kf+>HM~ؤ~u\UFfp-($/37٦gݶkEBD}aV%im6h0;{f:odSҳ@}sת52l#lý3`6,m8dJlJz([! h$ۯvlnG,B6 (vgj])Yִ*$ll㝝3`厶mV$3PTSҷiuvm$w_9 8 J)ۊv` 1IN:=L&ՔVԫ(Bm$?% ^P2$.d Lp&ՔVFC(m3[a.' 5*@-7%}KThH66ixAiLȹU4u; fS%%=՛D<(YJ'Y= I4% `6f=gs‘}#amϙ1%wy[Y(V! h$ܧqP2 "xC821L:zSҷn[~iH6e1y/I$vn `FnJzm+§L~ӶDYL>M,(%~''Ӻch|"Ԧ j5$a£S[a4 >|N80kJ2]s}:Z ѿ@FYLΩ0^PMe}:F%ʃ$jH6e1E;;8(#Qjf3Кm̓T3/l#Qw_9YZAbiҚu۲Mאm4XKrkv-(-[ wN_aN;ZSҿn[ыH6"e1}sj;fb=G]81'5%K.m$bjᘓm0@{f `flԔVږk5m$m>M%I0hj&m̼+JSҳ@Ua<1l#QS-^;Qس?ZjnN1)_VV1l#QSgvNm 6ԛ{ps'ChJmc>l3wim~ UnH&F1MIm!M3l#ɫJETImL!kAhJچd ^zVL^Mf*m4?#lJ z;`۽s!_Uh1Iؔ,PZ6Ssn;1zIu%=30W~mmjN(m$X1}s[BeЪ*K%s)W-e+K4^ed˱rm6 cs}/ASүn[}Y& !l#mAɨX? :癪$g+W 6m.=%YvE3MCZMIM3}0S̓4vi<->ǦXOU3?hs٦O 6Br6sC۸Awӯ?a6cj5ה(okջifm$( ~ݳ!a0(2Hm)M&<,{V FGxA~[w4x(oَtSҗ@pd}4>K k{۲a-MmN6%=ۚÑqƑ6nPrҨR,wr֛'ݶ&~4^?WpR Yiax(}IMkڟl6s0l3%(FRFFnf_:nJ^yVn8Q܍b$,֨7Y Wmm$<. ^έMgؔn[`+Y?IO:6sVS0nNm۰)y Td6u1_95so&t& Wb8hl)GC J9&n=9LOIAҔnթ`9=8 J7_&a;~^tsSZmV"wm$bwvNoނ7rc$:GǠOܪe:7%4pd[ '=|Ͷy7Qr6(q;*Y(I'\dxnJ^g ' F)*gd{{3 óIDMɫٞh붟9E8 JpuM4ūУ4%2ok$_Hy7Qꏉ&2X~>!i4%*YH4Ŕ[drA)?FR|x3)jM+͌0h)–NPmA)?FR'KF&r0l#vg67i[K %Arь Y‘D[LWq3^PZmL4 \Ɠ/v8 GV6m17Qr*T.nb)ϥpJVz ܽIO:_m-2^PI6HP06Uqˮ37S;6m1{;?(M}I']m܆#+`h:n`[޷m8h黯&S1ns\8uܮö}Y\MEHcqPyj:8șy5ؖu9^`GB8Y/NؑCqvtpғWmْYIXft)]mral#%9Y}t9n;>n*a[@N"wm$ζ7Q4{__4|^k-9@8)Ɯtj?s~t}ω@|Bs_mw`k'6'JeӟTD ~`[|װԽ6bp8te$??=/Yq >TĹ&6jYZ 'tήxCb F\L۹Me # p]O7˅#+`s1-6fb$~Y JnS;6bqdWti=esѸ~Tℿ pR_ ĹY"]lQ}8ȹ=s0H~,8S Xնͅ EyWq r Fb%@e[ROp{j^P8]Hq+?XLEHÝ.v|d y'xE+Mm-6bʮ{mmy$WqK Gpғm[mٶ6FRb r󒯺VM GV6|\.T]b$^ >Rqhm?`K18Qڡիv5⒛|vMVĥq2!Y㒲n nKp@-m$.h VɛA#e$%_&6mFRLfKfw?2݁lh\ܠ-9V> lrITmyp ڶnUvb$#)GE/,:nI׵hF6 u[wkޫ6K[b/ ޶~_=~{j\%Ko^VT=n6Y]Zr^xDw+m$NUnqK 5ÊFJafn9md% wMtC>؊z^ m V%K[<ؚl,m^r^5tܸ[Zxnl#mUWG/9H&Kb+-`ff.EJKv^hl[x'gF{۪BNK2Y6.m`Hu/(;/yюtVɷ 5l#m9r%%vMLV#p[鳚6Bݖq[-n[פY$f_rD&j=q݊lnK/;/Yk3y*mUl2aV6>?GDv^Ŏy [6 PDd sNHN m mSKÕ?B*+6qmjZ6YܯV] 붩 YmQV6"oC']6Y[6jUxrllb+`m~{Ʉ"luUll#mܺS2 4KqUVll#-&㏷cM`|p+[omVz^[GU۵/t~ۼzk"n-͗M%XtfuJ6b5l·S2Hz^7U o+[om;<}ztKoD;氿[ۈlpHOKkb>Wll#++}m:oτ%51܇[ۈG;D3d;z]HvJDTu51gxl#/owv,OPDˏDlaor,Q5q *)NDH&JQَK vJ۫[I`ḷ`>=DL!.a{aF)ȜJNDMp-<seώ'=P<(Y7ʆ6M굁m$ts%Cѧ8 Ӈ~{S8/eC¶ N:`3:?6o҇~{S,`-*O)tUūc }7uXCMr9nh$);Xa(YkH¶IN:`ۋ7v.6֘ˇAQ8/T-bw`'};s3Y ڛd :nUCMr9⽝,#*?}0h %긥ILq9 vv.GB?G m$tc% UFlh+:uC6m$Ŕxޢr8sq/Į6bl{$v%5H+ 5$aA1DɷG@)B(qdж()\oShKtMi[{K IFcPL4nGzF6lQ,Õ[jH6bzj2 \oèFсXڝkH67lmb>'ɸ$KN鱳#`6y1Jf%'H9n! hܰm̡ _`XNn,$lq$=n*Qאm4n׶jeh-<7ؐm4n׶-(|Y4ڛbvL˶HZ]۪6o} Jے7ؐm4n6z"ܦ9^clH6k[U9}.|.}97q+]CYn! hܮmUKo CM[Lֶ[lH67k[ׅ쁗icGmFfm#)¢2mu\m7ِm4nֶ:^Xso8*i[t IFfmk^+@[ul͆$lq5Cu (oW&;p IFfmSHއ5l}.6bwvf^xM܊fe'dn! h)Ε|nS6q+ \O?"v IFcM$i)Hzg?!v IFc\Lvp Tlˏ5֭6$aۈx?Cbnb<M 6k4zXUh[~?őiڐm,hSlnǶ@̑4hXo! ζ%{ܴb۴mf!K._ˏ_ҦclCŭmQ?W1rMĨ I^ۏn7ېmgۚq]ʼnm{y*< + ~т7ېmmy^4jj]Ub,?ľ_|'ېmmi>zh}f)]m!K~W\#ېmmUdp$e4X)[(IFtEǟ6J6$*rϢƶʺmf۷{Cݽ/^m}nH̶gmQc[pGp-,^o ZW%v+ضK؟A=n= N1DI6"(-n+x7ܐ\m@@XjVsՃMl[fBd e?d#rCrm(b ؇*N1e]NKCFe^rSx!i6Z n`doQ4YyFkmJ wͯUmLDImyM&xrCru8۪ɦZ)xz1UOCIۄ}nHζћ}m_ 6 GZ7Kn6QmNil0Rq{Fo8_2<|6Q7$:ۆ_ͻfgKZSa^r[ ݐ\mó7]cQ¶e?(, xM7$Whۮƶ}=_l+x]P_oS0/oHжAk[.e%mGӐ%m`ɍ7$Wh[#n[-ɷmm%a%u[YAɛܐ mt?k]zGd u,>ޭ7$e[guۆ ^1yQr󒛺m;Ln!2ےQSoxŔ>QrQ4%YB=V~%o!2cm?jԲKm%p%L~n6[}/d6wk\KCre%mǪС-vLnڶ|ũqHͶnH=̶mq-$7g6yIփkiH˶>ٶ}ho3oζ6o+[[J!F m;]٦ʾedbzO6{7gXkSR%o^OCr}ܸmm46j&ԃԅ|6ۜ%ܭN=lkwֶm,I:s4$Wg۱h6p&K`s.km=4@pMNwR+4^55$f[sQ̓\m&̂D#uae?6=ʿߺmVJd ,L+m5ZٖuVĻ[t­r08׃jH̶Gei?b4y̯a`փjH̶ wIi[)YuoBN\WCr]qhmK;Yn6۴%gCLVmlcn6ۜF:ྩՐ\mշ|t_6Ѷ~`~M}ZaP׃+kH¶j߇:@RW:7f9CMn5rWՐ\mU-xW-!m#%' J!Bc_ΦTL>T(}?놜y`|e ɵVeJݤ:4Jnc RVXm{BbvK?˒ )YB) Vm$ɗ8c3CVMX[P.&_&J,oB_7$eKvK[`<''6-yA3Km$)suGfi3f<_[P.&_&Jm=b>OUiw%a<(Ŷq1lQ7+K2vm'mosF8/!i|$l#1SLwGuFS[ߦEXy RV\mdy-9:"au C8/Td Mgɓ6&feAɕf~}[ ֌կ4hL<ڒ6Q3v:FNɩjlH-lwjbcdJ]oNRSqzD!Y/]2.YmE͹DlS6d վx;vilbrَH6h/^jU?ll5ܗ}F'K~yk%63ŔyV+mx:'žՄʂ-m;JvnGuE~*M$K$5$a.6syy)m3 Sgyl[u6!vmWp\d5wkG9印.Kn,KZV)qgޏ6 *-|4+(.۪*s|-*&YG:ӾԂ|(Wlv{Vy='?6]6,ﳲ ɪlv{Vixn[F=ئŜm{=:6:l[麂벭P{n[lcSɉQ+!>._WPrej۞Vڙm3",!rֺ%̶Xmsۢ66饩[xܠ8Z ;/9y9hlKB~ۖ_1m6 n9h4FN Y񪂒+M5׶_1? !Tp]"ϠxqZڷ6r9*k_oUp *d 6cU%f[pٶ*zzJ3xh]4lSg l[ MGn\X& Uk J6V iUmH\lGf3/9L]Hxl[R`p?T~{ܯdMAIFc۪Z]l+< eӉ J6ږUCue)#Lmfm 9%5%a mEm \oe&2+ J6T#u0QRfmw6ZQPֶZHp+󤏘j׀;(Yl#maYOAF:߶h=A5?׳(os~:kSf6JlKT Om {>6gWm*>LF@86qcc|%WfJmcmư / .d 6ll=Auٖٱ-ܲ=JJ&Ba&~'(.R;}ӹe{C$ 䘺b%:,\MPr])«g̎_aΏPv8^^&fSu¶x5e[lǶ+z;7rey6LlHn&(."K]e/`'l‡Nؖ&( Hls[E<~;6mj6q_,`,h<Vfގm?ܳqH-^b?(^Zx!l }A1]Dd^¶\p8^T瞍жM~.8Y dUIV1lcH #Oж9o{q+y*e^`x>n _&,!Hm,L.(:#њ|;e KVi)b ;Jٲ~靓59ޔz-٫&,n׶|1O A&?r ǶK혋v"bYɱݽI۾j'g&W*%5Y,"[6lIVXLlօmfmS;Xu[|Vm{[=IO:۴'vl~EFhz+5~V;X&YBj/(dS`Ӷm/}Za1 vɍvlu΋l/L{ss`Ӷ*//g,phLIF&YBamJΗxZmYmQbL lx2ͱ:K6w+7 >[xuV]T>|bl>Atb$)?I`kV ^t4`70[rpoD:X7sa靷Wf[[&-z~$<J[cJ>2,($ϑ|ruӬ̶L1Of1lFlV2s+m; 2 AT}1J2mM&Y[ ?Zv٦t~b6,ĉmI<6e^,{v%g]cM5ƶHOJouZiksFBmŚOl4 K*(iv&"YqgFB:4f68qiۤ%Vڴ` KMn1κlJT͵mɑ$|9K6@؍BeYL`])uSRa2Fb?.RX$epj?[ '=l5/^zTM.uoK5LPiM,:wkUS5k]M*YJηpqVfjKE'ۛ;Lf3xȷ#5/^i/\m*-uvĔ_s e205-c*ԣZW.02;]tz]Uyə0y:[cW*:jNm:Fbۖ|zOeI`tyfX" ݂m$t_/pu/_\nuikT?)Sךfe~t u~߉6 %a b*;鶕]?L3g %WeTb*c5]s^ קK%KrKq;F(NJ܂RFdnY]dv5V٤|-yən$%a J1r-s6wb$$'Fن8m%\HUۂY` J1%O|DݮlP<鿝\ʿ?R 85CG!u؂RFRL??ܢ6]nyA[€THk" J6bƍFRdfkO'Zyɱ}S6$a |m"Pd Ri.m$(ŔfGΞLasoXaD(AH%fem$HtJ2R|̔>LI%K5r$ͮBAIFTL#Hn1CXE&6|#3/PO(|F%I[(w5mڇ7/"FȝM4r{a=&CݶZk&_IKU$ȝR.MMXw^?n'J"Wmem$H4(S&wG3cۗY_TyJr&](( H/&o!1Τ F#1u@/O#px:򅅂MR1 ^D1nq3S/kY?ԧ[-D+^i}en6bJQ29n斡ny^v)&n2W em$h?^ wmEtl 8wv*&ŢF2tΫ`FTLY?բT [&O67N`G'*sHӑ, S҉T-| m?7F3sf6"?]&ږ J64z{ntDCcE/͟Y#<Ы\z8d>?e6bݹ&24% ?<>hwX?u:rf H6Nz9cz%I,\?v^ce'=g,wݶ"Ԯevy'AöXfW=\y%&4붝  ʶ/^RO$ݡOGDPSmkt*޷"#4EMI.MP9rO͹n#~@ܭR|%AL9I`m3mu!J8sSyWؓ36#qL(t ǿ0w9g۲3Zm"fONMI__?V gʺ'gۄ!zͪ$(K`k'}-4ulsRb$mݾNגneG)Ԭڜ% dtٶ*Omvݪ!%Fbnl;o/Q :9aG#_ܭx#ďI`mQuch)ivgI}vӯǯ^4?sm~17%y0Λ(sv}[m?]Kbԧ#q4Pb^X_ʶ]nOI1S?eخu#y}[mofD4RVїA9l{5RA6δ-u]O7 eJh_?H%byI>k}> rJq J"/#s9H*agm[JJv~%w6s뒗d(Fg e9lݶD%l㬶Ѱؔ"A[I' }t< Fljs4dCk-UKƉ6~CoskEwقܭ\oȝ[p:wmf[pmlkj#enk:_LFncMoh3h腗g3+%YX6ۤg2$k]N$ Fdx <_^`l2۠d5[mg[,O#ܧQ,ZUm 9tI3NP"H=LvinFJ8f[eV(DȶqWۤ#KbB*{`~N[:UCfzyzn=jOG>vA*BI&UT?I)(\(?[{jm&xzk?y \ƪ6AumliIȶqaCpD#6݊q c۹h@m~8wki}'PLϿˠ:6q$TCpplى `pl;Ւ$rV̎ “գ$(8ԒvJM JOOVXmOKNm$p'x熼[4KFq#»W/Y꯴4J붭iq$7}9Oi[Bp`daV-Hc[>g6 JWoJj7Irӗ~f(+6m&7N1rMH&p@~wzU~$Fl8 JnoQm 0Mnn#tޒgPN$g5CYF#wJpS!s9g۸I? юb)ES 7Jx*ܙ%W9OP(ajV^-neЀ\|Nhvc۞l6*׷Ym; h {mǶ3qo4ug Jgs9dlnlo]*UHm&ǶitMpٰn꜏rd_ZmOL+<o;*UL[qpK_UDO,wq[3~V2S f(U m$Ax7wvQp;u^r4 |I`omŊԆRm9;3J*`[E>9<̼mܕoJwIb[=m0ڐ[Bd((߃u iáՂRztֶz`sT*19%(Te>2nf6GAI֫ p9؁m{Rl-N7*^!7m&ǶK3 ӆS[u,wSYs9WۚxHS4e!7mf;cmIpleHmmnV)44pһ/knm;J`[S]rNE2F䒽1rQH]InNdYPZmX\m%Z68:>6S1_e':wU6.u9ؾmVi!y7?>oʐalZsdKudW^f[2'feĶ>7"m+I%LGi$,Y6mSxdq1NFnĒ3fj ͗t̎&s_NO"gkm.ø ?]u6H|\O[x߆I3ۦGQ/(mCa Q1u牒dLemTNI"J[QJ+V?=Cv0Sx#n) ͗RcѺ-f{͠Sù7ݺSr/ƶMa.QI웹[6mփ9:Dɹ8 OI'~Af҄&L*޴Lf=y+l#!(?!?6[*MINX _%,3t J|FBPLI1o޺xr-ݶtoG37SS&LNOI|㰍ηnd gfKl-x06mP1,xZ¹c J/6ԟ4p~6>lhT m'S-Xcu[0I۽ʦF,%/7HS?_OLCnsbxl;KS{zgI{>}g^>Y[VVPelۙ3!2"Cn3EY#Ot"b9nTfᵚmY;ҍW znܥкl jvҧ.&΄Bj[<䖾ORNp'u-|PpcLǶEmU1,%`bcS&4PIgI,Lvu[3;*|lyN׶_ J\mԀabcQJtnoը[KH Jpwhە,wWe[d1z66Mv*_۲g,kMu%g'(9x6I8I@;Hԥ/iTLgIXC,ڭV^|ի)ؖGֺ$a厷4'ƞ]>YLYIjhCneuLo;Yj@uo§}$P~,92|FakgAi۶6۶1ULr~!iDϒL? cnFoym/{joLZ Jm벭>Sę~,H28X~6l7ZP=r6j܏ =|)q#,w^l۱? i53 W,mS܆<TX+/Q q;%Face _YO =0vwwxO^Vpu%] { JFb8F$ !*^?87-ȗ~#(n5|_J6{Q񍅠(` HLg%#IwJsn`VNגOOcRʝ{̍zR/da ~1qfHM$iJR%7m])p\n@_m,(a ~1q$}DCncm[݀@Oˆ .rݖGMH,u3n HmHҞ\0VhUue$K`miKT\zj|A8HiGSZ` rK,jF=$c۹B^ٝƪGlmEJ/i<(9X2vSm(9A0䖰fkضԚoc}*~*YƋM` n1qhICnڶaƧ$6ׯU環R%ue#nCWe[;+ِm>b!j2NM0mT,"yU2.M4VK){^AG`ʟn[}p(}75 uwIߔya068jO$9: k.j}K/moᙒ+}ZMt'?}riA4O+W-mQy9NNkeUcĆlw&%O2?]dO83K ţ˒gm hBA'wlnLℜ.bb~#b')̄_KB25 ftVgɌm$1BZb;h|JM۪|I?D6_k@m].ۚ{ri<4X2댇l>|x|F8+붗2N25oR7s6bGD&4L U$eEW2Q2F z~˶(.~q1O ? m7;eJwҽ`s߅ĸ8ljQ.|۶JrV~6G}2y\wlk[˟v\LLu"Iw-ӧ۶J7AOmDWTOCpCX벭.fQHrSvljsã'7`n˹߅벭^л"Mn7ì/ OIK#Of%H0)ie[vx 6 rxmkңAOo]66uٖ_=˨=~g퇻ɟxmSϔ\wSA҃mE~fh-~˶aϨg}b0%2*R.3r'@~9lP۶6BQ1m$}p)Oqll=NҶ̓#V~U'jwfj|XLn[-xߐmwƶdsDj4Wlgonx"lKBb/菿%pȭ;v>s”d RWH&mm,nlՂ$bV`!LF6$VIf@cێjդ q}'w.ϯZWf  2(AGJ*w^3ݶU^$R犣fK&mMq A1m &(!-eltEj k*WBc\sц6gl3S9f"btD;Yݱ&eVůUF{hܱnwx',^ĔWFlmk|Rs+4d 7NyDԻHim(zCn/I ;SsATwJHke}jœL;DݶD.Ū;65[Δdžf]>2=]|vHʽS/nsGSFG uܒ1#6>ld2VC)yF!uf+L$l#+q2+^(LNF-֨r9mU%ne9fjml#+f!ӟO>o2%}:Uo(@N~?b݁m$z4^m3<<6g)V`}VUlA}ΜYN&FWL#8 jܾ~oLIi ) iwKM$PF܈]ٖN-[ڎ)micJruXjMK+ JoI.PeRJ!6=L2Jr'-6j%[:%9pcm{v`fbsoDN#̶6ͣ[LokTc aY-ZN'i{sM3ܹFm۶6\SLcΗY>x|^O?n[:T?433AL}`Eӧ7fl ˲ti<&Soܒ+ۊf(}I^Z&J:+;Eu4ff6 P[L5\q8,/uʈeOyA.S$67r.۲6|km h֗Zuzy[:4I5]XpYUGm۶26\nkQ^nI#1%J_ܜ{mi֨!cp-ukNJmS.ے6}5-r2YؕT%;6ߔt7Qn`%ɶuWiq3Z?Dͺ kRNQ݇f2[jbREF5Hmv,t]ED>d7-dBChhb0yQl ]mGӞi|to{-.cdFEn6; JU%1O}Gmv,t]6u&'N-|Q#~Sfp<. %lc-uֵٶӴm4L4L'J{]b '<d}Z9핶~uԶO{.DL* Րn$z+Jip e {;Ltg9\oF-qml?23VNC޷r?~#4(cĊeZ*8S1~ ϶\Ƕѷg:`Ulfn$!ENI>lݶ)mʪcz^,w=å*~põ*FuЕnii5&J? S_&}=\fy-/mKO~@0 m+}-&궍Q0H63CnE-L=zy}짃z;].K )&mk}DU44j7 ifSCnijTq)^~{\s3鶴-/*% ^<| ̶r44j7&e+vҩ!&i:YsQ?H:q?d'$܊>"i5m[mMkim y%2r;)Y_VhY*wڶ,mw72veFݶ|f<Đ[~zJ}7EHrXnK'9A ѨiTL{3|] s7:Ý9$8&8Lm[j8(v2PնѼht%w*g HpwrӉ D(ERc/p:抴ϑ+*GN~Oҧ?8{^F" H₪߶(2gȭ2X, V vd OX=ۖLv`*Ͱ \Hz63Y츥0ua\/^|bǓ^[youZ(WiF8JhH؍"ܺ` <1q?sC~\f붉}i*T^Xm;xK=3ꥅVHk<}50FMD!Za+ ep̮hl2Āl#/ ')Cn FlS8[mJ d2?.ܪlkg%+Jl z 6\8?q.\7J䟹k4 Rː6r돱ٞw u[< z͖y۲@\md~ΝNpm [cJ2Ai _C7o[E? W5.QsU ۬35춥Vuk!ATH qa;4AmUܺ-lcvJ;{AaH7j0prd,\/z =#pyۺzn-Xmu(M5aސ1IGHܠmQݶ>Yl'{l [v&? Q 6EVɏo ^yۦR^f[Xer= 3d<铘E+y~h1 JgmZLnS-Vev&uR 9$Hp@Fhx,&ap ;IJß\܌>yizŜmǪFv$/!7;qa6v\ qw|F(Rc]=VFMGݝmJGsf ޞe~[P:g)^.RŎjKG3tRMgܽ* ?EwAt{/^m84h f9 Cnn.{͏*G= 6PwMK[.6GL$iWoQ2yI .кl{;͔r\O 7cړnpe%J!IvKݲ4d2\2c~G@i[pշ&+|oև2'qٖؕ+I-u̖&|7w}Gn}yK3}cRAIZmmIY@k$]Of`m*;Cm52闉ݶٖ|:mo;]|:?dӧ-}sTn}o>dz:`6Dh/qd}:c[qίvv4KP%z 0.KnIJ"̬ӚM͗ZH 04=AIzF ulHŽ#3w9t'[EԒZx!] FYs^r#1cl6gwz&THi${Py6>jmI mi&P /]Qo~/=Rb!wh'sf_T]߶n $2miNh/չѶ4"wWi[s6KfAZ&DfFz'8#vJHy{{J6'Iu[m CҺ8CYSak;"7-;C˝'1F:B.[%mͦ4ɬؑ(1_/ar]4r])wd=$I]PJﶭҶ/Cۚ=lHD/'uLLd nmfmk.&(R\62UqN%O}$&G }}'sO}/S>2olЂ2SVh[0k[V_bIk`q0=ܦ^YΪg+HtVh[ַ-jl[ȶK1If$ 6BKckt3J{>ʂRn mm լW~[LI%mͪmfWTgt23Rg[ n-&Ɍ$%z4_8pK\wXtRUζlh[+J9d.uFՉճesW+!()5ڹ:bmʶ&m1nmC(M |yN&APR~K):m{DMv'%ʩ$WԙM՜hd/(\mpT46u\cPɉ4?66YMEP8 k-Ʀ ۤGx2{_MUӖ}"3% Jy[m-I`4ۆ@b&'fg?a iD&:_僒rVf[6.Q6mϒ^fmZR~DNnbeVaWDKzKyzy9܈\V /(2QƁj9)Ou&_By7\o?L"XP*m[muCr߶᧰*&m%Y-:V~DTr6lKF I d'0l&i>FڧvIAIcmܐ<66\*ÊIv" ۝k'Չ~3Yug)Z#sdg6NC %{67Ȍ{ʶ"~e[G%ܭ˶\Z˚ߕEηL'hEtv:GR ϔ>[,wJ,Wfۖk[[i6ӖL`wwK uh$eۆl۶Nf/(WWLWUڜ, &0I<9|̜ (uVַUdA{w(6FRLnd?0.'gŗ|XH8&j8 Jզȶu\z_l;&mn&T2zZޏn|%s&8̨YmqՔAn;=>K3O-<'O~:uJ6n,wy[QnTnsn5l~ڶ7T궣9J;3ukPɦ۞)6r}EqI6l~|X}md6}w,m1-LR v)gS&bmf|L.QEVۙ"bׄm -LQuVe[ve;㧼_j|m| XPr'>GyZmT붭67%նwn;jb(WwLl4lX-#IЊIê;Yʶl {׾a }cf7IjO#9 7&o-(U]*8{6<xj[N73+m 7ծpsIF#ݶuw{ ?RQEn/V2lPаc OSՁAm[muSscڶ_ȘXNp^͚- f_x%U'˶s%_Vd7_=/)/)-;m0˝ru6nJ٬jĺLf{7n- h>LRmkRu+mԔ6&-mWf[<nߜﵖdʶfr s٬jy@Fj%*]G#{uۛ}NPRcVpl j 3ura%)򭑩aDt&y]u~dԋݶlnU{m9%hyHԓx8ͨDrX❰iu g[6mW^q[G_fʶfr3%LtM 2`/ZPLn{H-(ktVh[ѳ-kvJd^0dέx| WԷq4pZV0ɸnʗ {,7U/m iol{ Hd>#>9ܖUe ~.lS*u TclMļG ǔ_iA۶Fz3Lv9ee\(8gwlm`9y'>T9hd{uv:{жԦ* I* 7ZQ >Dw/ L'Dm5H *{܌'JZMZ]i KSyZ`KbmKus%>/|d 8tQȷTvo{Hwww u~=ׄM\KNr9ؾmEe\ɠ$D07i>lD(~mnC"7ûiI6剒,;ʇ*۟Ym%ܒB-6)=t9^ &wX tm봭Kք"p_Tv_ULÁ' X_PptIo9[ps|T4¿+OhLzۊ68ʶV1<=s#~A/ce=y|,ܰ0Ij!5%W^Z;\¶JZ*(i0Qvjg*Mĺ- 7rwԈ7g >ikN:g-Ǻ+UL|_XutJS+H^r[ u9صm$)iۦ(iSCl[ے~7 *{pgFomg$5*'b'(ʣi_zW6Zԛq9/{?nIf(xb!ns FB/QL_P udCx_hMSm9-I6zֳkFnYPXOXv9l#ULY/wcfw㴚-KW? @n*ErԫU o(bu۳)` =zb[CPJ쥧+9F}c8jӖktue3l#WL=ڹ # ~fMU'Sʡn3{iI-R"nKv-W S$o 6Fycfp0J#qP n3{ii0Q de^q_ݮc[cS6Mddfz~,qȃqVUFٰXjp&yT ~$]zb:ghrC|d#dIǃll'ZQi^v)FBΛSM*hp<2ǕIޫmb KOY_wۦ&H{H_LeÌc,(i3$KJ_HhSH. l]jb;#"n׿9{l ik2l#a6ݶ;-HG^{VVd $6l#b6ݶ~?\z}mRS,xr8ٺep&n6lh[Q|}7m;wG3'Jz:|x|fb,BB]=~_6lsmUz2[6K~n%P$j m$m$9o2RY+KZ6u#NW0DI馬>!JN* mO~jgf\<:@p{:gQTUAII%m$mӝvۣNWT@SAD'ّ$j 5 ck'oԶXۖ:nSq`ǹjY(ȫjzAm$ mKl:n4(‘fMp$ĤJ m$t3%OWw޿CpN\",ð\.nZ6lTn`>ۨ')8L\0YFB3{keiŜ{Y$ϔ}6wj5e R*^&^G %LնbQr (#|ӧT1a3Ťz66ȅx2{%"jyIl0 %wvֽKFx(n+l z6$9,](zIL*擜 JҖ(%L0)a]?Zm&x/#˃;X%KIo+`azo"v$ `*ͤZ6]&L(/Hcr!%U-A F_Oz‹EJpaF_:Lܞ݉&a\m&+tmfojqڄխ%\kFcۺ&m WіњuPJ$Ƕuo=;.pn@[yvaFcέ܁a}Q*.m$/l#l2H썟JJM HhLo݋rr%\lFw۶Wi!Ic#I޹MHhSrscDmK>}jMP l#aFf)L8a2l#m/6RIj1[l#m.ob%AU2I` m{7sC"BƤ_>Ɋg>?Ha a m_7K䒜!eT9sfKm$<_nl7eT J^aݶ/RFKd%Om%W'Kl ﶥIeai[Kח-ݶŋ`~'HjadC J6V,\'Is> &(y}aF{۾[4fMj𽽫5[Hxo[be=\0LHxo["j˻8 J^_ݶ{k&$1;` mK*v]?-dl/LHn[\ٶC6]dWgpI` ئIw%[nkvڒJ6M&Y5Rm$`&_wm[TY)6Mt_5GsbRɣRm$|m$Nzamaۅ&(ymI` m~m4hm y@,&q0mdO[Y\ڒJ6M-wҢm0[gI%a ئˋjҲfjMk˖H6]mmçm.BZn3FCewwwG `$)LU!6M侲ZmP&Aݶĥm/mri%\2F"nJ 8( H6?}˵AIRmK\zئM5̖Z%I޻*Fi"ܹXc*$l#۴mr^Um|[lN~el&l۹ܾ~kJ* H6mRosGGlӿYd WTm6k 6V,l[:1J,$lӥJstEI%azމ)-L'pғqo[,$lĽmdjRIH6Mfp0 l#4Y6(IL*ytpKlyL,T-6M&{3 Ĥ.&6M׶דTm$-p0m,&ytTm&q0m8elǹmmZuKOqn[zJ6M綝e.&J6M]גTm8,&q0mz5` ئ{AkI* H6}vNp-I%a ئ{\%[l#qo9+ ⟋lq:}` ئ{.A,OvEXˮm$`> vNPAE\\sZa[v,`%(ܗElQҽKC)KAm|?l˼䌥+oq6}` 'mm@ulKP2 } ^H6?9'K( `ɰlsW&[6O.A<XvdF%(Y} N* H6?$K(p@0f=a2l#<•lFy%(2v󯶝0mrNP׸l#<%$nswi`\%$ 6S:AɄ剒lK&,gDݥmrZIW[N H6_%0 nm d lzz3[ '= ۀA.$VolKz6ɢl#ۼ%ϔ\zm$`\%"H6odp:-S6[Jme7Fy%(IVolsf.ɒl#òf}ۂl# J l#M\,U6_ q7Lfql#MINWoo l# J\Snm&KH63=n՛` 1uP2ƫ%'m$`RޤbdFyL=/9w[ srVcN:6`:k)mnrӭE~nH:wH6}i[č=Mܢc~_;;m$`T-䡴-$Ѿ-l#|Z!S"l#| J&Ҷtz{{mKD;H6"NJTE}ی]z3)ё-3~6gr0lz7}\m^C$a 5p3[}L?FyM,󙚪޲#]lD5Qއz` 5Ԡd z nm^Z­` 78[45˄ l#F"(2@79l#&PWo,m6obJM2 nm4`ߤ^iRm$`H%/^N; H6e=$LVomlP )٭vzwH6ϑ J=]Mm#6yN")0ar~KO<' s<#.= l\6Nzӻ F~Y&/= lَ/av mK-6wIލiw` ;~-qm$`%#nlW^j!qm$`Wz8CԹFyO][z` =K>j[P H6u*W^jFyO? }6{zl#ۼGVt` =u[7t` =x{N mޣ]E}Bl#ۼG$Ln}bl#ۼGnqvm.6{. >/5:nl~۱-7FyJx+."7FyB5里Fyxۻ粒zFyvL.mޣo{I9]l#ۼGn{TH6QYSl#ۼ\T;nlm?NT;nln܀&S6{6 T;nln9FH6m;gרv`uI4dJtŽFԺl` 2A{[ c AL`ٵ_2hslf0iμYaW~0;OC9琼##qeWDmO$Aza[m ;a[ؾM4gqö X ʱ-l{).: ۂ"z4'n+ =qö X87l ^sۂ#r b[b4;j m/F=6zl ^%#'U۷oba[.oyBNܰ-l{ap?o :qö Pl[18mA`+6y-l{eު~{?tۂW&lyNr~ ʱ-l{ilŽm/Mn='a[҄=f~?7l ^$mMb[?3{xۂ&ONb mM\?-8!ou~o{xۂ'dw=:{tmA`ۋr=:/{tmA`۫ö yxϫ<ۂW'dwy-l{uBpݷO!-l{ym?u{pmA`0Tݷ_Om/O[)a[ ྸپAmA`>+p7}Ǜ-l{}"VK+p_wOܰ-l{W7l ހr ۂ7 xܛۂw /WsmA`;^7m@= pxۂ p9qö -xxiwNܰ-l{ W̌rsmA`[w`osq67­~ }&]읕b6mǒ7W6N}p7=zateTM]3P%y4kL:<3k?A F'noo&%;~>l{CV=Ve?~? A=; wo_om۹ zzm;\׿[WcX=EuJm9گ9=ۘs o e׽f6}xKh[{m^lcdڦOmu`̱iXz_u6o?̯ζ}2k@*C6E[/m`t}۲Ol o[6c%y[(y̼$`[(mnzl%I6{ok]/1&/Kl{ osƶ`!I#9}[m_+l{ *m\ @޷mɱ1mW?vuTY{z$K"}UX`$K"$~% 6md?q(I0Dl[{MԸQ`l/ܶ'X6]$K"~6lDn[Nb$Eb BI ^s(I0RDgP`l@GJO<+TvLQ~6lCc* `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`m `[+mA`)El6Ov0?gUٕe]gzn 9ۊ,;]`|oo[l84ھG9l6k[mV `z0 `=k/c6ky;?Tm6X韧V3ǀmֶylTn@ `ڱhۆ6Xlmնvm& Vv{omG= Vmvf 􏺽Bm!kmn/Bj >l;vm4` ۶fGl"m4` `|68\sU{a|ߤ ,> / M@\YIJ ]ƶ/=dIX%؆m M?cѳouUm`<ڱaۈ؆mf &m-ۮ9[cMZw̒=N*6lklւpm7|/6l3mw^mf6ikvb &m-`cv7d3бa m ^m+ ZZ۶c Z6]&^moo>[cvSZv*[hdlöě֖]+!oWfu loZ[vY>.kڙo횋*25SPp6.S6Gy7t;WtWQO;:-~lS) kKUtJy},z +*R\GVÄBl;&wL۴ںsEwuE+\s,vwW2ӏ۔$_6sUyLS+3{IPukjUfmk_EGh?}B9-ʚf֝+(}WtWu8,zt<+0s6;(IWbmٚ:1`{=i%'JY8ֶaBT6զ{>.]]] aٺ >=.~lS%j{Gwo55=2ֵR`=BP׫w ^ncne m]}]`#Cs=+0fMJl:ζb6!Mm!-f\Nܶ!rqŌV&3D67>bA]?+Cs+w۔$_ɶ>Dp0mU]ئښr`boZL;gz.`mP.c[[tWuheMJllu~9llGO'%ٖuEٶ Sl;zC9Cm2;YƼ+0mt% B]3ujt5zϦwf1;vo[_cv޶I%+Q(ֶ/D\]]ƕ@PƉk./qMJl;gV.1E(J_ٶ:6-W񶝳_1'+Q(:r_&~^j:4_"ƶ*ۚGŏmJwP/d[mFW˅m3MΈtnwzMH7֖m٢EUfXe ׶bV̶T`[דBL}{L~hǞ+hݭwm;u cKn_fquEwuU WtW9ۖ֡eֻ]5i$+ٶ{~۲Xb[աж=Rmml/]l-mmY m;`[ qJm'mm|mŶej 5llEQh[aE>u(4 6vۖ'Km+lV۱BvlSAg[7U|mh='ٮ췼mkJAm'g>,8>k=ض'VQKmEm5zR3mvT[hmۮ (*sVvLrs+mkozmwԶ ^ڶض6ݰv}mâc:@h[ѝ`a}&Ķ(I+v7mض'V&mǶ $ٶ|ds@k! ֌-^w[,[Mvl]>,zԺ66s`mcɶfn)l+.JJƬŶ>6ȶdmzkg[smkozm)rmUi;zbm9ƶu7=j.a`Ru]_̩s wOrm1TGAmЍíXvv*XonmB_tWqo|]]a/.Ǯ.3plG7y$v-d6XO.a&ߙG`ڶ%4/z:ƶl+pm.0/֟+cs[ئt% UVU/hwvzǘi| i`%Gbٛ/M%kj_L+\ɾ.1Ewu\ۍT?*~lSE6^&6תګ%zUa123z̶ FOuڹ|.#plEwuS؋P2*6;(Iںi)lvĶm:8Jֶ6ͽO(l5^j52lȂɮ+duܬ*uptŏmJwP/iQi{blsby@h+F= jm;='\#m.#R0uEwuqEwuqc$ Ve]R̮VdR/bVC:xj1ƭm튾zwPpC _V&wEwuexE?-C]Wt_1?)AImzh3#| zg[)͒ (ի^2^ە>+֝+pm.C]ais2&6;(Iܯ*ʝmm!Um4⊠2r-%J9*[W7|fv2+durĶqc$ V{l|l)~7Z؟e,->R2ljmB2]BMWi.1æFŏmJwP/hۥ_߇`g~ulvYEҢm+lK)Zʰl 89\}]avE?G"C (~Ng% 5>gζ~ؖ~HYZxQBEE8\HYt%us2]m7ŏlJwPh[ޫSxؖR/lzѾ%FEwuItEpu6^pLuٔ$_Ѷw'}JEW;6CV Qϱ'-+z) 犞GuJ^ljb6:.~dSE dl+]E~ѵNrƋ=8N(rTKli~^F/.n] |#ǩG6;(IW8Y:cewR/#ۊQB瑗EDbk{UVZ۠+z=N$d=;'Ǎ) 9~G8P\]5.ѰWy6ǣ֡۠+i/nJc$KUQ`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`lKNfX/";|]W@I-9(I0%m/JpڶFc*(I0%g޶qQӶض Jmə>.Jpڶm$ؖy۪mێƶUP`lKμmqQ!(I0%g޶kLVaۇ$ؖY5m+CP`lKάm5m˱CP`lKάm劶]2l$ے3k[3!aq5} JmgFI-9(I0%gdl{$ےmόc[rQ`lK=3JmgFI-9(I0%۞% ƶ`3$ؖmܴXӾ<8󠿉lm휝LAI-9mgkYvU}Umh#5yf$x- icFm$ےv̵J7U?12hZyqXjbc*(I0%-3\c7w[myso*A⦱UVKlb*(I0%v֜7gkYfli?t]9[^nm$ےm;ۡW&lks'ndm?v'gG1mmu}VAI-9޶ZӺp5ƶܾzd>cYm,5ίl-c[r5)7 2 ګKk9KmE'VvضJm֚ܘ2aۡU?k\?}Vʎƶ3'(HQ`lK-qɺ7?+6v|u c[r_;s3i[m5){]\ax*ZVCI-9C۬Wwm3K *}ϊ^gm;m $ےէCa:ӶmK϶=Uζ}v]߽Xۮ[¶5P`lKj|rwvG#۪}^ޏwfN:_?TimwR c[r:6Vmue&)h;[:_im{>c(I0%'gSک;s¶1k{%j):&צ{VAI-9mz9i۶:mg;@4|kG^ٳS/]W@I-9ζk3cCl_N1iz=Yk93P`lK΍mcwf&COAI-97msu[l$ےsc[`[޻LR3P`lKضKme϶3}Jm֚}d[m] ے$ؖ)ێl hi(I0%[cfmsQUwttjr>4ض"JməضgSP`lKض Eo[uNێX % ƶ8iJonֶ=J=kFܘ%nM0mE$ےv?9=N>Vk7vqnBA/m$ؖmfƺǶ]AG7olGI-9l/=Χ{l;0J(I0%9 v٦_~]29y̕~Hܽ[m=$ے3-NV ld޴ ΕvJ>03(I0%mչo~oƶ8f:F=l[% ƶ/s_Ul;ِmnRY_77RGkܬc[r:۶޶,ٶ®aCgq6뷭c[r:27{kvm:Jm%:MJmeձM?CMPͬN36m3w쭻oBݭߛw VAI-9ζv=Ҍ ѤnsgcleKejGͺۥND(tʴmY}8c*(I0%g[ ~qkGlCغѤfm˺:ceݜ FI-9l0iʶRfL gk93mx`z(I0%fg9߱m8w^ծ+uM8yŻ`j(I0%gl[=U7{]@I-9޶wV 3ո9i7ttvĮQQ?t'5[l[% ƶx۪/^ƶ?l+G~ڸf>i_dUP`lKuC _uV"uּ]qsf8s![8|v_% ƶxkomk_ζg$@漮nyqm$ؖn5з!~Jdӛ§9oZμ'ear0;l[% ƶt;:VfF?!ҟ0Ji8*gX)Dm$ےvݑZ>Vv {bl5SQw'G;dVAI3ɭ {~#{gט5eg{ɲVնUP`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$"P`l@I % 6$3m3'c^ (I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$c@Jm(I0D$Od[m` J]lgga5'l4KzdTI2D_Uۄ`DYDR`l{@m~նh_l`Odl[V lt }%-1|¶>l[ȳvFl[v}&3m[I=V\``[JyIhQU}!7Q`Bbl+ƶuEؖemEdD6c%_?EdD6c9>l@2/m:ҶXNi, ޭW8AQ}}%%˼$Z86]ONP("AmnEorϛm`4{US.l\܌LwAJ>MUm|EBa.c*7y7He%˼$Z8vm=ֿ6"3CsL%ʝm]_w*=UU?Vvs5vM9mɿF\uj3#il?z|/G~ko`eo0W5SbvR%˼$Z8FhltvkDs_9ٌmn"+n~:N߶6]e_~:fbf0G:S٩t ng۠>me滯A8Zgn1EbJ2/m:ζmmئ\t\Rs+YlWo[+ybV f]/l Ag[ms5^_{߹ |϶A}jwf zٿ,ot),hѦl;ڦkV54_Ug3ۙ+ O^G۾ӷt9\7c֬lsVR^Lnm;[QW{ S'.M`ﳿmnl; .y~ %˼$Ziۊɏ7͜mfol*\W/ȶ}oU[m_~6%};d繒wu`[9h J o"(fTǩ:ۆvսot\?h*M-minʶس|;]պ_p:VToշ]N698gҷesox\:-l¶"=8]zy[:ۆo]+r?xtR jq7_ж#]ʶƶL|U;d[7YXg[;(\˰um8c3T7:\û~k[ng\j/W̽1mac*6Ҝvp'Z\׳2}w>fmk3u+we{Qپpl2qlײm{lm7k3pk[fg\8Sou sζ? m*ɶzzҝ q殉%K |Tol2q9퇶߬m(3GU{6]ζa}%]cȝ#ˮoǩ|=爼ͬ1U m;]nl+{cBl6u;=۪n9 -^:ۺ]g?[ Etz{ gk[aXN {hžm?r/stMn_-Vrf;t YWVPl;{ElӲl+]k_/ضm?zt5ͺ]m.}b[o龥mXsvl[O·=丵7vͤʹkU&ϮoMzY}wm v# sԶ]m; .Kl۸GݎMgJmV:MvhF򝊸j/7<6ֶ\?ffڶ]wͤmnԼlZ`mN6wo25UK/LYMNLw\oyt^9GasmJd{iMomCLUnƶt>M. *♂㪽m5z vl_+D^hq66}UkfGIW-Znl9m׆y}k_5wwl70~4!}[i0`/m4k@i¥򹯺gm6V_dSmtɝΌ6\:u_2:6 m lr:,}N37m=G;j%7m[̲Zi,xݑm[[Օ퉏̎M-vos׸'l3*v߰F/bB6ӓ#-W86}NrUjfζbgێ=gMN6ͼ5~t2oymmLU_{aҶb`[W<33r't;:F_}jaz;XN6#T}hm;g}ڔm bo[i2r' iCXec!}p}a{mI LwځWo_8NgʑmGoԍm}T8yl|?:Fo#nԸ?"ls(hM%ѢMOVe~1ظ[ܥظ}=hmm]Q;B0fw:i{fvj\w/}Po۶o&lkmf yWٳj/g:Fl+mh_c\ 5o[ofLvm[bLIܱܭm {l޶][ɶ|]f7n7ozlmS.,ȶxUl:2<-\nosֶScd25gy`@awc[˙Mf Žl7=%6moڵMж}cmCDm?4mmՔmj¶.=|N1j Nc۪xroζ"αmOE4_ζz~Uxr66ѭ8Z5ض]7 `;잷~mǦhZi.`[xG 곳_g^1-\ax\1 o?NL-/g[ozp0ؚv;y[϶mgiڶ+C߶(.̶mr~c[uE拷ȶzz#vsw1]J|-.b4vc=wИ koO־љ,޵O1MpvjVn+j<zcYۺ]ޔwֶ~w|͙|nWϿ}~6AtiEvsJzu8v6)5 {iێ6knŏwh lG=ec4#|LN~Vi\6lmsܻǻt|m*?t?v (oӵMnĔmq]sݶA۷o qێ:ᶶnm gsl3s-7jm>۶"aӸ>SɶP6785|%k?hѦm?omNxۤ'nCnN^ƶr;nSC޶ˬm~b/^RlKg3Ƕ])YD6=շm}/<`mm;CE,bmt[=6_nI0ͽu HE:oj#6X͉V|[1*-hd;m}Em^g&m;5b۠>hR~,cuMڏ$Ziy[cvӶ3 ğYqvq?a[lۍlN2Cm+6 mfu%k?hѦymNIlʶ*La6so_l!NW3&5jyۚVTͮܶmGN2؂ԝvs4Om@grzz5c%mƷݳ 5eۡ|۶|g~3(5 R>mx\F3e}i?Ķ@lz8ܩf3}IN-|0mU;{Ǵ_̐m~l\m/6j7|5?'9S ߶m?}YX۶zmU'3--Ǯ7_ɼ$Z1ǻcBlflZ}[):lmrm4kv-޳[sgmQZKE^4m`[{59Ĕm݂M0䪶Uݥgyuْ۪޳Srg7^ئY 0D6==v)lfms:΢mr VΖml,ΎM-yIhӋ%LVvgm][ZB;lW=og[;u,$oP f^-[S~bTmmc16فpk[5]b۲m skP&K mcS f^-$Gmo{$6e_ m OrPneWI\&Q f^-\>LٖOfOlwe7}۰쑴߭h[\Ƀ4W풥U;6m%ѢM/Zm?OVI|lw!MZѶug/Im.]fwljaKE^Mmi=MPv"vә]Ŷ㚶q3OyȖ}mmC`cD6hǶUYfMWZ&]ٶZp>GAPmA gYoBxs;6]%ѢM/Zm]]sҶm;S|zԶ®j~\ͶI2_e~ȶn]%ѢMݞmK3 8|E2n," mn|=AG Zw{PζƟ*w)TwޢdD6ڶKoDGmWm˦m+m=:NV}]\Aao斷=Nl_2/mZf۩/Sm+m[q˔mJOz۪C|mE϶{cۼm.!&m*m[ ʭGg۱V msnm+^:jo[uЯ?MYb?mQm˧l3- mtwemkX9qжS_w{Ә~ݱo/^J5c⥞M`[Giwk٬m)I6uM.&lk͌m-l{uKmVx۲MDmug$ llLf~=Bl;m;޶\2vhv2:0؎mkOGc\R&#Ko>@WCv}C,hѦex1l3 ws!SbҶyo< l3GweVmǶ۝m$F}m&#zmwחֵ;k4]%fk'iNE"r͸ulzm5s٭=Ӭ3ӟeV.uKi=vn->3aڶ}㎜(IԶ=~<6QmNv?F¶)tv4Azl۶k됛nm۶3736=v͑Lf;}v$Oj[mF[ l;ꆳM*m[5޶lN4M6]ff6:emf3l늗ζll/>rflk.fG]̔p/ڴ6ygnmnS eZ|nmPcm;Z><^d*F mgKd.msS 9tdK]3io0ֽO\CmbҶzim{sٝ]}*yh mE.?ض. vmݙ-_l^nmܠx mkv?|h2k[##-m%ָyo+=cQ|>ǭreK7ej7>Kf[#.>S?Zmb;%p4}*[ҵ$m[mx޶(q}ƭï2׳rmxE5F?Ll3]Ám]m;:ۖll+|fmzk7n>0ڶpFU.SEy-܊RQVsڶeꞒئwmw$j[uSzlyʶᵹ͠-w>U.7n?.tS=~`_ րc拗ж[9ׯ9Mov$j[ojNmy'V|^>3t_rj϶cȷڌmLZIvv/mnEk w@o6}lΖg$jveuU|(6ƶTr>]\Nnqp=ު?ދNbojFKi[VkJKԷ}%Vvh:J5I?[Jwln?T k}|ɹo6G_"K*}!ж t4ڧvʓ/m]"T~oֶ7w$~Zz+Wmm;믹$?m3_qֶem m>MOZfɼU]ϻoMƟ9GKi[/7%(W57ݲdyl3ӿey2|2xҿ_Pͷ%+&ٚ3fuN^׽eN* ?7qnYI t9_[G*EГ ?|[fVߛdE;ńݐF̎U=mCZ{QؒK%G(Yv$ѢMcl `[mA`lk>cFb͛f>Y=!dّD6mf l[7tjv3&ֻ?eG-:Gk$޼mO)Yv$ѢMclOe={mMcWGɲ#mz-ێ͇?mwPHE^˶C_n&rSJIhkV$E6ԛ$Og[.qZѳql FI_Ƕ*vrm3$c[mKp$ul;ԓGn¶`$ul:l;<~Kl FI_Ҷsv9۶"P`lɶ"P`lȶv` $!ۦ7c?md7^omxL7U}1;N+o5Ӷ{g~>WEgś]elnӖ2cm%ɴm?Ĭ|8mg&bg}MnZ>˳~lsXuk>'?h*Owmg7/3ٙ銦J]-mkaʶ(ͥmZб]1U=6߾=0Dv,Mm/$%֖vG}'jnpc۶nLqp8۝QfӛG$Yl3OO}ɿoG{yvV[MmCmjmr\?;7rˍi8v}Nnbo]zvX1bmzOv2Ǘ39L7ݔmGa:ݤgֶN[90Io~1}ۦq\cݩ[x.y=|;Jl׸A fǶ^=s3H⢏^esVֵSS!MZm5'p5ʹmwo #{Kb;m_JmĬmS۶kvho>g؝N(Iضol4m;nMZ7L ug3ǹ\}/ئ. vk۶Xc[/}npc߯V#;5wuSWo?mypMZTvzp⇹\ewt{;%Ҏ>YAР;s،mI+vj緶[&kV<`.6ol۸wL#br;hۨ7}=~|`4Cso[9ءl;4=΋%q+l;6=*%mKw7"ݵc۾[3aE]mΧ6M`6zOo5cgږ^ml(ڞ鄽mf[9m۾_E_$a[c: =.#zi+gmɽg'۷_ݵcgt1c[жCV ?mcBMZOlfm}w{m^,رqUv"myweҦoNm|۲{Ŕmn~ۺuZ؏F7g[ڦݛ-MvlϞOYxͬmf`gi7x+mch[csmoyʦQO~`[/zc[P'_mݝ4m[ֳmm6xa0roal[l;tUζ}s /^B7, ^m;75J鶝ߋ.6&6ts0l ݺ۲+KmIv {os3>vwFI6cmy73ƨ{ oM^mӟ68?5/󫷭{S.A'nJLQgmgm+m|lt9}0--l(I3V.Ͳo|`۾krIdf0+^yй]~ٚ_mݫ[ۺ2ζ};7% ~"ƶv >LZζ4mۼ#ζ}g[l;޷m?o[ˑmUfl{$J鶅_̗vMɥe 4C*0~6&YiMbca>+}k[$9l3Kc[/y۶MY״-yOVa[iʬ=gyny϶7n%_rgl{ϵ7$mqrՍ}{~ ˤmU7)G#1aMb7e(d^&ۢy ێ<ĶmZƶъRcqIцƶ+J%6-?FrZ'oJ>6}/E{اrض-zfac(m/r,]-g:L+KBֵ,mWnlGo#-KHlLڶwx;$Yl;5Sk8=7iv-l&3{LulzFr~:fzμo.s$6mk.d=k$y&ۊ ̉.ڶ~\0 t=6MZPm[x$fkmIvM6/?mƎֶm=^<ܤVeyfq.Ŝu϶vYƉţۦol3{[ی$^% ~*4Ƕg[koG>Dm;Rr{.t0mzecƶab7eͲL7=h{r X?;J,&WKԧFmm7i妱oŹ ]H/Ia̟mo>vٽm3حmeIMh` @>i[ % ~~mWٶ'rrGNԶ-Gg7we[Hߌmn%i.v-ggIڿe$bqƶ\?~}۶*Tvޑ6ܴC#Hkٴ3lMxҶqb3b2hRw3 m3l6I<+JLmnm]|#o]+mՈܼS[.#l;FmkgNwLضIlr%Ipt8Wr݄mNP4͝V{!ؖlV-\gpYz.`7NζwL6Nlz&Zl@!Yw[?+J$yٖl+\#oK=#KoEjxVfV u8`muWSu;CI64mmtKֻ$b>l;-]DzJ`;rmĦ^B 'lvro?mlT=`5i`h!.|{3Ny܃ Wl;{Jض^$vc5g۵6=tlsnzR$Il/{m`exեok.WP+{վ޶=휙ǻM}Ϥ]Ym~'v(IVfV;mFsL{7ϪXu775,u>جF^b7Y.ζernl[d=}vqN=mpGY>u ~jå"s# m+\gzeƶ4vYw4}JC\϶:{ӶiO1>yW!"|~foA9w|Vs6CلY`aJS˱#ӺnbY.j'fK#6sMlD<^Q`lv% 6mQ`lv% 6/{>"cX̂Tv% 6vggK$3Q`l $0cXJEI ,Ƴ$¯((I0~}v>6Jm`)W0cXJZ"EI ,<=d GI ,fwcXcX%t`% 6tKڦr1ێ 9/G︨\lȶ(_:v-ۙyEbC_j6)M@lϘVEbCHpOސIۀ\T.6eہda fl4hlCQ2Ŧl;V ?3r),۰Eb3eD+8mg Ko\T.6i[ۀ\T.6c[kܵ Ko\T.6k[3mS.* Kw\T.6d[]ۀ\T.6c[2-Xr%%︨\l6:ums.*- 5a Ŧl˫ L-6MVHFO 'թml@# m$`,qQؐm\IprdMٶ+iw-7.*4c[Qvt$+.*mOOݸmi[tzy/l;S7W 4 zo5sԍmqЋK~ܶnԳnDlmyp$wEhov }S-_c*~`[DV~Xmo~Xmm!ٞ$٦tmu*m Vh=|`pupGG۶a[N݂m=j[ۀs϶H"`ٖ6&볭3o۪Cs=/\e ==ln ێ QmGWtmM-Wm`pmwJapFFd6!mHJ0rx"l.mǪOV>mt c pmWE0%lNm==㦈zMۀn[P>==>E` $mQ%lm1DJJO"^(ܶƲ>Gw`[t:6[£;0 ض+w`mw`km6`۾;h#؀my7{w` ҷ5yl6m6`6 }p ؆ 8l{uz}rco0oзG؀mlC$p ؆m6`ַG؀mקؚ8l{nad`-v: n`׏ m+קopp-vΔ)A6`=-1 Xg#=8XgS䷁u6`[ZvȐـmy@rol$#L܀m`[Ll 6[-'m`-VAolƶm`c,R䷁ec6۱ m` ن6̆lCa2̆lːvِmol6䷁ed.[ .[-G~XeK! )ېVٔmol6䷁U6e[6ɶlC~d[! l-ې6ٖmol6䷁M6f" m`نE6f"[ m`ن6lC~cka-`VNzo( m`ن6lC~Xcs0Xc{EoKl6䷁-g۳ m`نd` # AېؠmoKl6䷁%h h[6mC~a! Eېvؤm(LVؤm) m`Mچ66mC~`! lM2䷁i۴ m`چ6FmC~X`! ,Qېتmol6l̳Uې٪mol6䷁y6k8 olֶm`چ60vmC~f! L]P L] ䷁a6l0 m` ۆ60̆mCa20̆m+'ng$[ m`-ۆ60˖mC~e˶0eӶEol6䷁Q6m( m`Mۆd`m# mې&ٶmoͶ@-oƶpOT1}aېܶ,(UR\$- k-h/VCCaېYmQ@{.~䷁A<-ml+gRvqB6䷁AVb[dD[XmtDYdDlѱ[mA~c%џҶ# oMm`ؖҫ3M rSmA~sےڶ<2I۰0Jl}J#m&Wi)n[|kP;?)ېXmkm Hϱ m` mJ*v%Ӷa-` m{m_:&-E~b-Hm`6䷁)b۱k۠(y6䷁)a[Fgچ60Eۊ\6Xp6&Cn۾zfږ! ̰JKn؆60:lKUlC~bMڶaF[fXm4e66&3xn[x?F[r䷁`[60ĺl2! mT6 `X|6䷁k3mCa20? HԷmmj;ڶj yې&ݶ{$m.mV\Wn[mC~` C3p mP Lm9m߱mgF`a`CvP(5\ `+v-oxo1 ]IQ m yWp ܏Z-۶ɢ]{kB;@["nہegz0 wΤO;S7h\1>4L-o۶".("$[OI[mA~,ѧPVkiǶ[-oYm$[nH[)maкm y&m 7A~, 6b[9i j"F-x ,*l#Y ط[-Ҭ2R|m[oҶ! f%`ȱ.ךi`ib[Mq$=ڂ6XFz&5۰XF-N\۷˲ۊg[~{ ,m0-!U"oz)˲Jbn m,+-dIݗE"0,l+}vmm `Q/C^~(LKmH"_Fv:o۶>nh!(jrn[qWűmd maPsmO{Z]>Xo޶l'2l h(L mQP#J{{k$P6Xm*HɟP;q}rxn[Ri6'TvmC~,綥uFƀԶaK%:,rxn[Vii{6Vt(i/ +dۖ}YE7L{Y*&70綑1#΢%=gC!G7<>זEMoFfq ѹ! wۚH̷潉ZHtnoݶvv;#:v%ܐmPل%  6Ѵ5`)6`LsC~,lk# mK7A~,l܊B~,lܐ 꽒;7Bl\s[`!o[{MA~,lon6X-֜M*ڹ! a n vnoe؀mҶjRlaݗU'r+o"lҶMsC~,l׭:7"lCi[s,@*Թ! aK۲}_&%Xm=!ZV# `kuno%Xmն$ tno%؜mb+o =ۄ0,l䷁~6hH6X FWo^ PhȲR䷁~hȲR&l/wsKـm_|YrsC~g6=ـm?vJـmR&l6ƼMsːـm̾9N>䷁f6`_{o.+E~fJـmy m nun9@3oOF6lm" 48JـmܾF6lc۪FNunol6~vcY)@/;om 6ѷMwnx emU3':7䷁^6`T6g@+vy[1m c&" ]ɷQ&:7l۾x mGq# t۾2MA~dѾm"s6m ?x;~6l7M  N6`Mf"n" ۪m*s_V6lo|39.+E~hdw% m T~{7 >6a۟NG! 7=$-+E~cUCB޲R䷁>6aߕ[% mMm7 @ro]l¶si!!{Y)@뷍4 o ٝ@۰-<ٝ@-Qے| (L؆mAhH\V 4{jmmg)0$duno]l䡈IC <mmF#$"|ֲR䷁&6`['P,4E~a8@-:`Z5V66lh%g와6l u%g,+E~a6@BR䷁6`[SADiB;7䷁a[Gq# ~hKsX'5Zy6b[x$Ic I" Hn:<A0?ܐZ؈mсX(I:7䷁6`]ms=֟ vRsC~`37U"M.+[@-}M`Yi6Vl#}St󨛖A6lՒQo lɶs|]bV]`{lʶ?+RZ"ؔm4|az`NDUVlcä߹A(miο6wr&L6`[|c[Yiq/+͂]"?lKwV,aLs Db ؖO,.tna&; lxwD7"έgO ؀m=_C*'.+-{|m@ :tgDd+N'$܀-ӚTtڹ׏!R`>-^dԝ>NI`ٙxPs.+MHlقmӻ|+pwfe۞zRYs˂Gm@ ض; A-#A؀mtn\$*xlL}oـm>m׶Xb uY)60 vM\zYiv ۆ󶪷зU6`=QjN%"T)m@-VN揢K w^l}W2n|zj( d2ղRm@-VI^,2nڹm@Flɠ5A_;olf}۵o#ɬ*ۀ"-k~u{eoClk%6O$;dݷacl+'m)9Mjڹ %l~uGc٠2mQge:XuiR"MB ol9MFc;E# Z F"H4=CzlY{#$B"sG s+.[-[D9G8f gr]pq`6[-a 2EIoEocJ\`[: \unGElllST*$}GyQ8 lG锔I<(Cd T7 wy:m-/;I|. $`&??Ke뭺rv]`OJ7( zvmKhd.m-Ei2vm+55nz H9l62q#k݄˷_\/m!-y=#ޕ?ՋrlضLtdt(d0 ۖUȗ`;$CL5&9lضu< Sڭ!fe"y9t7сoA`lٶ eJrdg˶eMdd_22qI/Hyl[C6(yq%P)7 Ϧmp?n$#a+$Rn@MvHJeUd0MۖdF̫nm_4ȳik(ޠ䙟v䗪)eӶѣn* J^nf Ym[|FbNٰe۶]'nb}[% Hm:jzUCI$ٶmE{"[Uv HqۢMj(E3qے`ݠ-JImY'M &/lۊ/mE믒>KۤŶ3AI.z ɋۤŶĭpE&{!C݀Im݉[q-Y %T{l@Gqߖup$vZƶNn[0AI` ܖjƓ-~vϗG&)7tm[a=]:GpTYn-:r5M6o[o}F&{jW&gdz&p-ћMnPIwV}PmkömmCImRmI{3n'? ;U%nv4hi_}p7}_Zti&jƽvԀm/9Ub7 t_l]8+L. J{oS݀4V3m~fs +P8DN6bՇTG|<8a$CI n }mŶ|.2p3mס$J M6c[1ŗq{T'Id ȷnYͳے~<o"Zl9m(=-D{61zLݖ?Nߪîp/I䊑BmHqaݼ68Fi"!lG4m)p:,ߪ|Z3Di*"7a,J=iD;ftL 7+y[P2BM 䛤MӶ_2|AIkQQF͌ɠcmᦸYoڛmHMۄQz1׶$-mI{[[z)4Gb&C)yS#˷hiTK ɷ@v[i 7ŝLgSN%C3"y&C鶅W1^><#B~mo)h` J=fx/bi}rb(vJ"6Swm"(=^["NV q3_OK{6}3hHRe=2N;a:+ 'Ml*^jf{R[/l;(MRZNͤ7h[IqEoŬ_ (Md&SI KqtwXi?+ h&Z*dNMܶh~+fwK 7ImsЭ/HI[AG{v( N2bKuؤmwyǯ^L3n@ZCalӶكkV=B֎{2۴LܮLtv.}m6fgvpSJY¥Fmcy}Fk:e~(Mgɗ̥VmcnQ‹e\,َLQ,Cc<# ljFvJ4Y%+l6V; ~+g&˲LErC[۵Qc Jk> &ǢIVl6V;K]s㾭3n@r,U\Ъ آ7q2ɟL=vW$1zwVl6)܉-Jڮ{֯7>@*kܙmٶq;H3޷v-ܤX|^m˶Oȹ c/ǿrJ"&r L6m8rҹBVh(M`ʙM6ߊC@/p#\(MgɊgmho )n@a8f۶R H+6:?q+jѯ\ H #IgVl޶%ol7mLg(d!Hnpx N&۲^<~(M@ҙ-6Hq'neַ- ,^|}ZbXşէ_~30Bi7VAJn?.(Ma7&nlQ _oF覸ĶgFJb7E.L`{ 0,oN:CIna0DltSܼdb[Իo%rɕ0Nmx7p;̾S+HMc"llQ›&0:Di4 I*\Xuۚ)nQ7oeq/w8 C5.L`[M[v>C h C5.L`[Km]Lhn@SH:5\؅t ~sTEr,`i`[Kf&gm[%To%d0XErn`ەĭm`#u~q1Au^&WArf< V$_u:4[̣njۢ}Uޝ!g ]ɵm`8(HP| ui(a&WlZz l$mVW)nQ7G/#{CI&31Z\cx SGǿJ4B-&J ICI&_S`3< xaO`VH5XHnq'0aR4@ƽ(1|bSتH9qm<-zՀ @ gGaϬvV`q lsn cJ"Va}&.h35>?[,ނm|"JZ(٦ܲ-'lo R,m|b~6:{kݗqw&6>)X5Iokird}0eY\uۦw)@*G݀ ެV$؛ 27f/ ~/lΫ?S)W6AߠDp0JVk&V$X[u&H{^&pDi2tvJ'7=Y7 I(⢰mI(t7$xɋ,W$7rM^Cv`do\k3gaQprtu"@%i1mSn"v5$ȸE?,@Jpz[769ꆳE #L*JKק["spd imۦ,cf6Z4L2H璞i3E4VIQ7-Ja܃$}7́Kn`$R%0 s[(Iʶ~|##*8P\cx MR}2Sm0,;}&ms"nS`ۍA> +-XM2ŕѾͅ;n`4!R܌#J?gb0~ yw[͉;n`41uU<dݢk+Gw>wp" MV#(Z g?|wm| +7vyЋm0 [7ڷ9Rܶ`ݠ[0I(پ͙B` [kB϶Dx,Jnos"Fs` G<-wqhJErU7Ycȶ\8mosk i'n^\6) E{}_l6g*klm¶[D>&&}oosm ienm0~ԪyKp#S`anLƶe3^qm:<18l8` I(W6rsG0 o%.8d[:ԵxMֱ3L̾I;L+o};LO'aQ8\_ o%SsW?qt i>Hگl/.xIW\cz l6GIW`xۄn:ߞTswVkLB7&Bw{{tފVRn.V$^uDOdi&ն5W6RM"L%FRnNV$7R6gUa? \rs"ll" m3d9av_36RnnV$ט]uۄ}m۠0Y&LR߷MDъkmmEtPJJ4 ߍ/']ʑc<[H9[HRatۢuV:V6,L؟W&5F'nۖ5 jKa C0Ym`E);L>~rgZ7I%omY]uC=lo[im`Ny9ZH&Bڮs[˳ۢ*wT]ےDRvӷdNwmgd3Q- &Cۖ6?Bmwdї!{$YF]!FEs>@l]Jl˫5cM bl12+(e=.dװ. />μW$_[x6X밍('p`6ö2$4DV1['8jC8\`0]`6öcW_}y?*LI޳o-SnLA\`phc_zugPmY۷ sv#6eC%OEئ8]\cܶ-nl~^B>XQ*)Y~%t8;F⚱4+k Ym:l>w-k[o=rYv)CKr˃CxxErPōmS[-ݷ y})`zB cn|$jȳ a߱0SI>8ns"m%o4 h0 X287;ns"m uۚ|nә`?3Oy24f-zEmmd$ f( N_`*-mӶ-<"BE@\&lm03{6Ngfڦij * Mn6(L&Jpz\$^TҶ@mP |fm>:BxQ\ch 2uAa2'fꖵH6la0\'層?$FVxn[HCmp ޾Y09fn^0gX#7mYs!mZȶj%d0Iߗ!4MZT$nTԶ}0YҬV{im:;x~zm ifՍfgr{BUۚwEL$'}.r-"C~ßw 9;h%.?qJ<<'f '/VKVIfMԟ$/?ƒվ$c&Vݬ6$Տ7:APkdTwAIύ0p[.)Ӣ{n99"@.e-EU@Gc.zlsiW$/ܽۺHܚ$KzU`{ێUSh ⨇j[]P'^zgL܄7m>ML:F~kzmk$dnAiRn9s\~U$m[FC,k['&(}8xesRn3n" j(˯ݶCU]J2"\S׶&N_Q 3]B*v=-wl,YErCws~&js@lW붭>S$MJu&MK;I+6jRFIͶ5qč;4vG*=wۭk!RA_um)h}5#ڭAP%7 #ndo'$r#+hܥT,fCڭc5&mUٿ*r#y&-PHl5|Zl_|muX^lnpm+c^pw@Y_Pi :N+LYt'ђ6N OVзQێuުίZM)>c1ͅdB*] Uv5sP>xo~b(ڛd?Fr}̐vTѰXxk }# hY۪8IEi%ώ;2Ql+tZl#? Vaz7ٳ7HrBǒ{U 9Zp m =l$q0* 5S*sk YmUadĶ*N/LR4t5Ķ]~dD~ar_dǒ:*3kՅ'nk-_ڶ*N{s$nH܆8@G=,]u>2"ɻ"4N+L<ǒ:SXv l &nćd`:<,|2Kٶ쪛ٖޛI7MZHN~`-mllK F$n";}7ZH^~U7F;8IoV=&y~}*s;/o*˂c׶xo6'&OI>Km*)朩e$mȞmNe6@m٢mKo̚mdݯBm%G- tKGdm;[wlKQ8#Jo<1;/ }n4V~vlڶm[.@g˯<Hh jqҠF9dU7v;ul#lbME<(KtϭdmvVm[pmEsulgҶlKts۱PS9妴tF(_JщY6 VZOIc܄?E)wkKܓUĶf`~ |.nQ*P@ebZyb`U7M8m2zʭ.旌|,C5>y1ˬm:Ⱥ$I4d5c'Pio6=\ $zXh[#ɫm6o6-D8Ir}jHZ pjV2E6IEV6-ĝ.VwN )vk*Y6IEV6-^pMKb0`N]2{IE&nM ?0ɧ[tǶ/E3$7ئi'yp%Fz".tku#k1 Īئ3T&hX2t[8zYQ{6-O0IJ|akIM%C?X  w?aOpz'}HB?X` lD0OII[. ҒrX :Э2W m:#0I;dѓr/1sh/ ئH'e %nRnk-]8$i" vact$xu:I$tܸtٖ'^q`/ ۴=HSq4:gIJƒlK-И$%mSaJ\bhZ@ئQs' )!R֔rcpi[ZBʖ v0Z@ئ zNjc3!g۴ܺ]DxBoljh+ } lFbq>x?|p1SaGKo<;[I[MdI&(Mג}[Th9iHW6}['sm)jA"oGMQvl3m$U7M߿ud"L}I>mRndYwAm7­$މlHD>lUd"L?,{gKcIټ9,%N`FqO:Hm=)UϾ^7 !,%m[W0I0kzsf1@D2.yF{֒zW6Dhchxy/ss)v1_2x^i$-ئ~7&j#tӘrk!xW̮v $%o6Љ['84i>6/_]cʭ. cn Hh'Sd YJh]utB'nu.IMCcʭh6Hx5 ݶօ 6(ZW6Ts*N}Δ[36MEէE/טr NymHZI69Lږr98A]𬂬ݖTAr Y5}`&mcnnHoӗrkAn?*ٗDrV%9,l$`FB!GkK7ǒ;&2ld6MoO=q6m)vT{V?FXH$ I&n+8i&G:M:)1V,mtmiwiD?[Iٳ6d3nHZHt7 2X"Zi2g?r3sW۔um D$\Sn§A56=NmTN Wg|Hˌ&)۴6\ݑ5)7MImz@xgnM;ĭ=XgnJS6ˮ 7,hӒo lOdݪm}=}EF́4߼dꔛm=I5/3i܁3umIX|f%ӷU)7zGz}q-`@6d8'ɯrS7:b% e*쁤u2mH$-^p˶X8 YODj ҅< g(Љ2貭`(Tum Й?xr קذ3-@ (:q|z>e[-* $ \mҧw0>fwHyu:}3˶dzmD-!Uݼ-=ϿqmH=>fGl^ \m i- @?θmq|[2wlIٶHh8Om/holSu$gwǙKV- }p,j+$ {I9] IY$$9o# x}Tyd3`]e y}%PHyydyۮŹ|Zo+im_׷ 6Yydyۮ7I|oɌmYm~$w\d> 'dWu&nRN%3x%?h o)2%3tɶ `QDmo.cn=W]IIn~66lN5ޒ`m"TWV @@z~6ºyfqlv($N_Mo%B6 4 <)l7a[~ }й?IXB$up ~#o YORaq؇oc[e/>:t! @ۚʜYZL U$_}I_@UE}&q 1Ma(igd߮(GJI>dd}:Zem|H: mИyJ[fw)$$ ~΃SmMǩ+PMn$͋;5ȶxPҎm׉ q{RLYN? [d0h?*G=|.Y@p# @nm졤ۮP$뼉{ͳ H)YPkuڦ)0=t$ @Go7ՀI8xvbF"/I=[d[>w(iɶkarH[P}muk?~>6wؖrH: 'ύeɶka2 ཰m#8ߘr9xܺ%$'8ze[:sck6I8mH1Hxӵʶ"x3·[@ҝ$@m {;[A]5m8-vf>+j}IW{8ў[Y)%l-e[šƖmĭ Tq,y ߶aH+f+"g22-cOo*'՟8/@ҡ$aB7lc 9 Y}Ս +Wۦ3PK,3,Je[a-ۮ+J0 dKmB#/-0F0G<BIn6ok0 P-}B%܇s6zpη̏RIMsXyf[1ךmmar&! ޕlF^ou~+|#Yim+9wCv `SIG7lKoXX-L$q٣-&:1V*k6Zaa;@ҭ$խ(lm߽l;)e/4ECEoˉ۹u͸fA;r(ρ)[I -˶|a|*mQ-*2Vͤ_퓮O?%Y£eN@X` ~ٖf[;q$TnVhdo+_u^4vl[8Uf]±$@Ӣn~Vw{%,k%mBZ\ _ּ/};?a`сkIEm"љZ-J+J0 I17!vNmٮ-܇q-՞ujсsIP7l'5W M64̿wKh"+]<m#@ҹ$eO]8Y-j5^Bmn7tHĶግ&b-U#󥃆%(O]G^gɧ5MtӾm~ν^5nӝ#@ɸ.ydJD9ݨn>V5^֮(m$U\ha̟/T.'.=-=d$2]<ClNZ}yas ȧH=J,>,I؉@[85 :L)POH^8x΅@Em{~Nzo9}52u%73Պ%qQ죲*B"dЭK65ѫɉ[MmGٵ-j6y[RkG_X^L":"SN ${@cAu{zsHVק[vCw¹&Wa}[LĶda YxVKչe IN @ _?>~nx:̶0:tN Ta[šiם[KF;h:*`~ Lvwקm7a5{KmKpgJh]dc&Cn[M*|ۺŒʶZmVl $*mB(=Em 묲Ȇoe^>M6: $q6!z}ۚT F[2ybI6 tU\>YɶI:)L:NYdç1fQFA|ӊۑ w6lt}K6C4*~f?qX[,9l%1:9, $;͑+` $u#g Œ\ gmuD&O%?ڨoק8[23׬9`[3qgoޯu7:o& 2ߛέ mtv0ŵ|/sPl;qzk00pkC##" uZڨs1'P͈if=\9zA-;|8;YQJ$㠧"/g9=Ҽe}M:P~WU=>W<,I忨@;y p< SqXRQd|0. $x[<01cM.[0IX"9[DR޽RVK۹%<  Um<('lk I$PS?HDQ=mC JGh#m7e&{8a[PmE"Dԫo6/durhUX"6ؖVIkmčI=N")-|ԧ9swXꞹǿ CΡ;CIlK*ۤ_9Wlm:Nx z4g1.iFHEVCCIl+ۤ'"$/(އFY$ө|b7ތ6T"bFзSf|PTۢzs >4F͊$Fr (ַu:P8F&LA\b|7qwֶo3.N,}eNӤ\~s*2oM 4Clk6=.ށr Re~y*FWկpH(] v!Fl;FM<؄)|4p>GMo;7@ ȏ•H}8l!*g}.[hYQ6ٲ!_J ^;7d:U.yC¿qi(mٶ%E;q+8'uY|}i?鋓t:HU'{nR}w=: %-kl.,GɲH==M%ȲP_;k;7亄cYԼn %a[ċ7;ёn¶ΛUGm {V?Le.7m3>AKYQ:gߎNWmhd fxUd `fԄCe;CIlK^ >|omIs>%j?N>R[Q %}X܌4(ޜU09<&7>'bwo[lq l:BIKd7Gؕ Y;$Marxjw?_*kv3PR$KܖZ~|_n$Y+X߹5UG/=jF+l8X2Ŷ(-V4K:K}7yٚ͞W귭1o%``:?ꓫKHߦ:I~RÞW'r5WfRot<|m%[R+ҷ-$u9׋}9un_Ko[4CI \yb[f۷-Dk{pe;olK8j]{Ϭo4K'|b~D.9,ͼMCi Em:]Mn٭KsGNt T}b¼&[zm􋳶iHumsfoOSz+1o[08%M%{뷭~c3nm2uqns.m~ӀrڡdG |뎘1/a[ )\MnQL%:ѷwi-6-de oOY48mt OMwdչuisxQkXWgVTNp\GjSGՅ&اmЬېDȉ6wmZjpX3e#Il6gG7vǧ.Aw?a&5I.=%):S6.*mVmKi*OֳQd&!T`k۱-뽭Ftٛb+|طuSfkxaA}U{EE\-kՉnPhu u}:vuӐٝ4ɍ]0Jj* e$mM[(u eKRdZm-`4Qկm?M[-&i[.Xޮ_䡄 Zo;oiF/s& jd0*2\DmnIħ+ZĞ[ !k(uz|-Qbm X2̽k&xI(5h(ZşUG궥ߪ4ķS3 O#_WrKrҌ?L+25e_lmsen7O nMd$0| GׯZR# ck)^Tr'+ؖ{a݄m?= #njw7Ⱥ4nS炒5z44-`{& m +Ju;?3q;MH.ƒ<[ѝIvwWtm&` %gp7m\/&iM䴍OǤʂ'\C`ֽdKs SNoltUZe9dmwnX2 ~utӈxky,[S6lrn2e[ݾK(PZKe;_)v/(JpM(ؖݝI$⺉m!jrGxvu;0dC0$uo6IZuKd{}brk7"7{uɜd543n^G19o=?^'vI隨~_)1?=oΘflmq{g>LPOÌZK| r0J ػO.amsޱmacܥmt%5t.3p,lm}mNږ7*nu iNZ[:XJ b%S6G)&<[͊_go Ξķ>|.xC~W4O%ܡd5^X]ty l_$'mNGo;ݪjWKHs %;co̜Q|mϊI:Ϟki tdڷu45jkbb)n;;/)#l,uSh֭t2NCɶ+ºᝤykm1Mi $Mf:{|o3'dSjk/W=$<;o;*Dźщ!l_[3icWLyX͊7о-~Z*]K.&uWzOmfImsްN[~{Ti]X_MYħs?^V f!0)Pre܁-mJ;a321 s(Io~[JHҼ@K %*M[#i18;x3oK{wi?Nn Nk%Ao%znĉ:ƶ\E%s.NgmIݝP2~[4mئ(d`U2CIl߹ Ai@^CIolS$\4m0)cJU٢m=m\ 8;9eɡO;Jpy97Lr]CIolҎ fnbѡmӳ7 s;Pr6 JԺ>T6$ӼtӇXr(E_ 8]PraWit%ݸa뒗JnҶJVGM7MN,8ܤm(&k/Lx @);7%m$fIO,7ܤm(&D[ߖp$.NNYj(mmESƭmܸa뒧NWR9f[F$bCIol4چBIC$HB6mkPLrm}/Lz `7ַ9lD$0IxPr7&g@x]r?^ lC`gCp&ۮ%/5Ƕxͨ-$e`^BCIlKی.*$Jc[6DضJ6syaS %- jf|ڂ!xa뒹'ޫQ\6E*3q.Xf(M>zVρ)E۴ $7y9Bm8aS %iv&1/L| `6mC1M4m0ۻSTT۴J6bI^E $7oIO,1ܨm  p$X) , 'L{,-6ۢmIܯK^`(6۪]`;p$%/0ܶɮc{`S-6I+߃&!P%Ym>"{¶$ unegj6&0)j(U- xn[2zBM/6NV_ouʶsmaltrcOz|KIuE#m^0)dw:o[klpfNċ=c+-mN x C듞=-4?zFK`]8aȏ% /m첳'&mxQLOo`}c{fut(oWd[9m%6E>)^ Cɲo{zm)Tn0')a\6Xʍ&#P'mhyQΰQa/IˮQgh=Wۆ;!ڂcntR;LG]ry}z}m r&ӓ bf@ēmɮm.2 T$Pyg[,i)a_RjCm̶a^T6vċ m-H2@ƶmʍ&.y}{m9]0\PALfI nKc&nUn1+E]r}߶qdڮ;K->|fď@uʽѧP־v5i v&1 3LG JՇ+G&*HVTrN[PLbgՏ@\}yCנg[\ ˟ne[.[ }a8mA1< p3$^%Dy(YVDlEp;%a}SQ\%}-a|T!:[ͳ͇B|:Ͻ4cxhUٖ2K˯Gm?Qi6yVHa~Tٖ1 چby>ϾUMG Au(?=/N[3Ya/@ٖ31b[66ȣҷ1$^%DjCIl+#m9wKȶ9@/eIH(%c['؆by>}iH\P??"]c 7LbFxPJzg[HLPL2/eI<ٝAm(m㞜ڕ ۖT@VďNsU-y $faIJ=^l5?X$ev&1 3LW :ξo۪XOޮ۷1$s|-ٖl+Al 0MK.Tٖl;wlLXt IXaUvvlۖgQn0o)gezی?gMi-mp:͹1/@0Wyn[1ϼ?6m0o[4KJg۱%Ǎz6F?G0u36vm^fm0użw͙Pöeʍa `P;/m>Ϻ&0@{k|0v%{a z#m`K>FK7sn7fFúYCIolSa-8M%)9Cɭۖ7$^f %n[%f$aS3NCm秏0$_zlw!^Iz w6XLvmnI|K1ܼm@}#Le @~(ys$~m6njm&3 =ܼmz ܆&.J63=.2c[g˺d:PR#mQeP(,; k׺dǶ,鶠D/g{'XI|KJc[Pvt[TВOQPD$XIM%-ųNn v&DtO'8Lm @|(ml]vp`1Iܝ":ƶm/d?Qi*h 9GaoP { $ȃקoCƣ0)T yc[6cw6$ܴ@VpIM%=-]6! N'(Lq @t(mǂ/g$xzH$AIѡ7mA1$p+8k]r!:ȶ9n;h!!L$n(Lo]P# ؆b-$ʾI QJ |0a[y럪$уn0)-xʾVo/`IPۄQP#ێ7ڂI@Hda $l+_gQ۪M})$lȂ}D4ps @h(mgsN8l3A]')޷$> 袜vI`INƯ66n&vwrJv{( ۀ.vvu{I|K.D 袻/Iop0)$l('mr0)$l'yoyJzgmΒ?LnކawP#ێ'ls9fo0u9Sۀ.$θIK.H_==ƶYۜs~oI<' Lt!B$U 70=ƶNθ盿1Eڝ<8096^vq%'&ؖ7o Ţ)YgGJ6`a9y05ƶ(;/[2p4 z$Ql k[343 gˎO %a%>4̂mqg.9 ɡ?Em3nw[ x(?;?769ǶmY6CRm./ ?%msb!D@j<>ؖVyl89oL^=765Ƕlv)m/U s48ƻ"/ ؖNnd$E&i3pYpvo~?"Py/s 4P}Ѯ{˃}x'&MU $x^p%=-_(0XtޕWUO=ȶlnD6X&!}[9y}:y%=Nsn'Ж* tIoy~ P'ےIPLb/Պt;]mevu;dRi/P?~ۖ^6|NzNK4vm$l5q'm$lE7y3IXCIVACq'Ю o( jPLb~$ ^J¶180Isw- }jPLb~$ ^NJ¶Lbo4 $lA1=apwzvTa %a[`C^DS{{/g` %={b$qp v{9c(mDz{>ιP߆I$t1ȶ8wv%أ&IF)3^m$# 廳J Y/˃? ~ٶ_ֶθ3E7L5GۭR?ƶm/;&Q; *e`g$AK JmicیTm+II'L?Ӫn2gK&5:a,8gW>v,HǼm*0IjdU:U1ol+$9#z0I]tv!lk$ޕQ:9o%=m_kt|>$J^g 85L>Y:m:4olFJ&)WTJv?甇6h$y׵[l7Jס$lЄI2z6io+CI =CgU:ȶ3¶LSvEYPZt B&dz6qv 4lPIR\YYLcqQ{ VO&k1Yˆ]J'uQ?lat_ WaN=~rd$فAwGas?\xhض~z^CέJ6zQb!߶^g1j޵_n:P2 ]Tn67)l0E*W3ii|9m7L'9pQ+n}aj-ȚG S %[vQrazIweGpӡd>q?\xh&0ɜ賻C,;_lRm B\CF? Nox_!c O "$MPr 72,m}/j7W^\1YyvttzmtL^%Ti5P&Yy)mN/LvtvtP%_^- ^7%qs^d]A o6b[,lL/LhA)}nO'O ;q~jL/R|IrvkmĶl5J7LivIDYmEcQvmjNw+(YD_{ӷØ)gۉ9$9SfĶ*$Ế|ۢڶ&<4( } 2|vԉ^>o+`ۣJ:j[7(|۪7aP^-(&Mck[wX햩?5miо %V@8}WofضG/ʳHǔԷ峓2D@k[t}_SV3n'Xk['(yolRsn'~QKf U)S'Y%g=mYW>G`W_}ImˮmoG9h˿U Iiomݠ ؟mE58ϹD[PLb,fbOLʛ}I :Đh v&No%'g-B8s1D[PLb:oUJzO;ǶH%އĝ-Yӂ4x|Ώ[Uh IݻwTCtz*؆IltUUJf4mW$I:Uryr mam}ۺ,]I'68v$`g)ؙ6mUsʽ 2mA1mz~P)OC]TnmQcYz}g\Tnm dAId/gN-ؙ6m]Aɐ? w9l[Q:OzϏNË}-[$?"f~ J>?}:%&j'.*]̶daPLb^moζpa۰3m}۪ͶYGKeMr/AmN lKJˁ2w6l  $ܽ:%<-uT[3e}ۚ*%=-^6XY3uG̶pyWi9Pۮ%-wT[$F|5Ͷ~ Xv˶jv=_eEGoxe[6pRuֆ"rm1lUFv+Bض '(mcaOT\7 ߧ&'bضmn8 J>Ĺd[۶@m[ AOĹxd[-Pۖ9m@r+|-osà'owȶj 9Dm[gq`goWc[= m>ؖ vwWIl?m^ؖ9TJc$lƶjvm I'm`ļmva`9aPcܼm]a\skYPm@rޫ _lahsk Jb[۶GX.wؖ-R)m1l"M g~Kzb[۶H`%?n H9lm`|yoq3öz Omc49xa[TUmel *hr㠤 |moʶx&JX,mgcT4FAȶ”m*ir먔ȶ1p̍349hTl#ls&0>zmlV9qP{Mlx:0gƕm7mƓ6f̶7g9U%X $PVیmTi=IXyftކbwsXmaz*!(mm(&q:`- %]+Hxd:I8E`%gmrxd5EcT9fPҳEnfn}I]>m7N l3v(>!(m;s%+(>1~{Ti>K`dmRa[jz?ID`%?β7D^H9$nyb[d68ٶt*9fPү'> $NQX_ 7_lچI܂?R'JyƮ `%_Βw/5Chݬ As/(ƶPÜh I܂?Ō`g 9ئ2x˝?ChKSAIlSJα $nArޟmP(mAcAIl3:mA1c'(vㄟ?e&mac);'("7l+Lچb 9vPś> ;8ɶ̤m(&q `r mI NPқEn^~G߶mA~!9+%0hp سз\{m(Um7Ml`ۆIs06¶_])m1l0a'~ml0I;(mq'uUBoŰmu|-m&|avIgnL 6gض|l7Nl`ۆ g(V7>vm!;ض2~P҇En6ضAONۀOD^%aܰ DݛɃ]%am&J=JI"OA %=H6vli7m/R6yN@ϝV=, `s$ɬםf-d'-Lm|mCmg(`y"~MmCm|ܰm96'~M, sǏm| l$nqpE63-Qo;&pu{l$nB86izܶ &So8 l -?M3-DMBOOMmQbOOMmSOz4A>~¶E.^@mz:6Ex˽`ؙQ˽`(&q87~gy<`mD'rK¶)PL(Y9gOmSbۙZGn!2i/vkV`[ 2 #[ )TmC1\;4^&֬ȶ6-a:nmm *mr נ$7+ۈmOѡ(gr/Y`9u0|mePL2Y'v` m?d' ag2Dmc%=T?oI3YK]B mD7uM8M'([-ؖ}چI&gNj[6A JrsmXm ]?mz V.`.u8ndFl$N J:Xmmip䃗paGpΐZ64ݠ$/v+}:o;8ʭx۶F7(1s裴?aȻĶ#ms@ߦYf;&mUwYg6G݇MmYm[ l |H'oymmkd?;Mֶc2!jeM9:'C$RmI<7uzAIv?lwۊnsG7vtﶕö J:綅AUGK %@Fw'vjOض]'lD#Az7wo;lDez}=3ޙ8O{I "Բ#wm'%I "Բs=#>?ض~AI`;6҇ ft( ݧ%I)(G&Cu7A${0p~P zX}:[m5>R%˯Ὼ{,fVwz4:SK/=lQ]J}݅ۖcƼ +`%0oaUu`bDa%Yrз1{Dm 3'mmbdnV~^ҷ36=Fm/ fO1oDhIb88dΚY>YofA"dmvnMs8`_o+>Ym"lz43 Jfc6 hfXMmaQ@7a,m(=-Er.AP2c,DIUGZ[Pz1Zl+_)4]66gA3pAP} J=FOH&jX66a>6%ѷ caylG6ܾ^$l&>6%ۄImnJb&Cc[۬-rm¤ 0wfmۄa^N-?_6%1oDhR$m JN[jlQofҪocOm 6%|`k ʧN7(YmmV定 X+&m6%ͥa0S%`D|VoN4MuO=*`ԶŒN3J$62[mPxwww>4-eI>ޑ6Q֒MvQP&ێه~̭yKᬥ]M>O&jeC@lssqOY61oK{-͂m*ss.wJ=P')ba(㠤_ J6%йɸ' n-&mJdo޼2Nf'L`a\qbkIa%wGё>ž[,h[ܜdܓ .q˺7 b=ߩM}ba+۱C/ v&eIxdk=p(i=xA)?Lϼ=ORK~<A[K" @JR۔ô^I> B2=m`wn=zƍ-ئl/$7LAat{:(ئL\c0VQ݂mK\nwPF-( ۖ Ayg_Ja-ضoH0$I.mːoO:OF)LVۖ6$sD-Lp1l[CTf=Zlk%qC'$7'Q6V.1qO[7U*Emˑり٤5Jaar8&횉j(m1rfBTDM-GmVt{d7-Gï<2(`zʯT.mˑ,B!?roS v/?h0۪n朥A⧛wU  JW*öA($mI7oss+aےss[K`ۢ`$JW öEɰd$mYRa۲\Di'aض0>|kKK`¤Di'aض4GL&˯nq_ۖ&E S*O[Yt|~GmF>SWϷ݂ͮmFH[zLE:>Z={{rf%nJϧg%aW7[یb*qo[ی~Km],qmF@PKHn6#%eSLQ)`йT|♵[ -=n(*\ ܒ$3k`!7owJ; 6SXvcN@3k`)֒v&mƈз9N@j`147$6P:~mNa>L  #/*l3Hv~6'0&mmN0鸱9,Uaim`>LLaLG:N'/*l3Ϳ<6G <}xN& ΰ "/_L_ Lݛ7oaۧ&%L|2^nmsj`iҐ PW_ڞȦR 1n\ٙT7G߈o!-fOm7n0嗌ȄI`=0qs靀_݂mnˎ-~2afLaz'$8 7-fdaz4 70 l&n0=v?&S,#LIOɄI`M0qs`t'dj`Mqs`4/6`.8~Cv 7G  $*hӷ}bp[*ȸh0ӿK:mpxYEt S*&=xVJaom7F7mۓ'GmFd7Nv+Z.*mb`SI7 ƻNO'^TZH[Sa N_TZȿ@'$ 'vʻ^jfqc16 &skz,ۧ \c\mo\u. Z.*Mmv@1&Y/*$۸ Jaۊڷ &E a({ؗt&)gEd_J .*Mm2n\o]z]\T$ $ֵ"nEYɴn3}LI,mH1B`{Ϟ`mBL\ QynoO~¼ ~ H1F`JlM ȷ@vZD`{$x~F-}BDJa_AceX'U1]InXv ' k7'P o8&nN ||%n}=7'P L J3^Ɏpk5`0Ip1qsekpƍ#$J92nN7M7P Lp#|&M=_T $>),/EհM J@hnmfA*aۓs^TZ dnI2/\TZ dI[$7'}ŝnI`3m7jKnN` 6:, a0 lâK`] 6, e0 lsvK`KdmI%%6~b@1Lrk7-]1 n6Ȱ9}$AI\T$ӓ> 4&Dm.4uŨ4=m.>*aJ!&nQ hv 9E7,qDmN3J]`0 lsrf @id% lsgg0 lsSAed% ls L`0 ls L@i%ls L`0 ls_nPY6bm,&maf%nS!&nY*L\ȸg- &nY(L\7!Xm$97[Tv6Hj:Q pmΑaf!5w-q] 㷶$ka{$m3@%L½@eZ0 Z%nka T$ka TqmX"L`čw-lsg nѿmegj } gG-`nyn[rEVUOhwwg%/q[mѶ:nQ 0va&nQ[v*mm/m:*af@Jv67~v%~چ}T$=~g6LܬmgmSLl&a4Wi۠(SېqN&a %2*aV@sۢڶ_m7({bmuٖ(|-h^JK|-{M:*;1Lm6Ҷᗊ}I,&a &AfK<-<~VWrf0T26z˶3w%nQ M޶tY%kmG7m Hsm(IƦzn[>gep lѺw~F7$Om׎Lݶ^'yB% &!æ$F8w~[aFlF[kpIw&.m`6yp7۶Qor:۟ԍlq`mڭ5N.۶n'o(L2iyk^nIlnyoۑpn~V'Ӽmp0L`p{rm9'=~^[moǒYa߶eDg/;׶EX``Uɬp{rm;Zn.k~{g'߷ k|Lg܁68k"V@Iy~FviiYޓ`IvKз j<>ڲj[^c-mڭؖORS-}vMз ji[xB)`.Jۊ"8/:>v6J6ZA6-:ob1:*A*l;~F{wDoI``HvZmy(P&Wȷ j``fmIafy[v OǶe~d}t߭5نU*6^ZmWd&nV{&$-!n[Ey[0o-Fv yqHݷnlKWg[E|t#nlKVg&n6mons#8?u)im-mJ Ox>ے7?݄U7LmQJk-Cf|Mb-:϶{أnG;amP'Um`uQ^NﶝC mv ˵N2iu),IV`[Vۖضloۥ-&wj|H'oOB:϶[ .5a lmymY޵|RZ@]b\ǝv g+[q2XߖJۓm[ќmEɰ hdM6pݶ'lݾ KvmaPꛟt @]UQ.p޶CBK0l+rTvsۢ:Ku 3C"T-mGm];mI[ۖԑ']@V5+L)+hۄc[-mKsڎma?տ~j$u0yvs۲u#"Os'z87SuwUU%T@]!Yn0 l7wnnSSڃ\cVb(uT@V- m/wM|&Rg1(u&IyEᗶ&Hl Lܜ a߽9n$݂m~Sr&n>p]ca+¯CEGmK\S’((yC.*7`3"LZXr !9-炋ʣE$ зyF/ RypyoA0@oCo?ݻwyD&a߿@6ryv T@ 8rQyض~6H?nSXv ,m_}%liT׏6.,woa Sx aCRv$$0A69l7 __6l7!߂7Fv0]Xٷ[f~w|#>~;|#00}o`?6>|k @ѷF`!W_ؗ;7ol7ȃK7 ߽ 6 &'1&nIw?`Rmïm7<%<%}v;S['`Љl +t6>srcض,4ۼ">e;\cVh[F&n+#^YEΰma yE}Eΰmarۼ" QIES# EtgK>|m~=d9l}gD쾀m^;ѡ k$Cѱ-Eΰmq0o\srgض8lprgض8 lR.*m6߀m<y l - Y }%}yKDޟm7ۼh@-;`6mۼy l-Izl nmNt yL| ><&}@c҇͛7? l@cJR<y Y߆#`pwww lb[^m b6;% |y -G c]7G b6lѷlzvCcj1`twonusض>U`@ Ntsض:I;ۍ,`&m}ssضB`` l[!ҹݫh ssضJ2m.p{uosضR~  l[+mͤg-]`ۚIqS5Rmk&%NL?=nLrZ`۪ йl[5ɛ7o~jtn. fn.V:7m+[`!ol7T{l- mۜl[=96WmPO~PzI[m$-6 6z l+h-6&nMʾmK`J=foqlA6ڷmem"(=}v#LDPz1ĽaJ=tⶹlAblA6B͛DPz1osm"(=F}9s6z l,6Ŀݝm(Mjзm6M*mS$g-0)[Bd Z۶Ƿ`j[Tm Q m s5oS۶qQ`X:CuZ) mm_nY<ض9ֺCItзIc͛7-ϻ x7m#w*L$:Ϲ*mgRK"m{ޙy 9z_EVnum;,Ud{ BQۍB6YN_k?p4m8߹Iۘz/m[mWz?y#;ZV1b&Y1*Lfx P`ʉlMvVtd/fclvǮ9lfh߳mй#w`[6COvl{{cQXz"sGۊk-am` N݆&a 4{\(g˜`[6Ccn̶bBp[am` 9JٍmzƍyٚA?öm<Ľm@͸oKzL9s EmsҴ܅/k۫m`^kXU͢N^hġU:8l;-oO5ܶ"sݲ?ۚE9}#Xl0)llϯԧ=Wݰ `{~iJ,l؞,Nb׿n?69glO?n>6'M@s9r_?6g0չa3?^s6PdG7lx}@"ڛ+n$J6Dom|L I= U7lx&?|H.mϤLA7jG6x~:W-˃mڡs%E'ѯ ^#/d[?zc_6Cul;f0+~G5mmo[e96}pЉe[i˜`hܾmMo[q(C4K9ֶ 4k?c[ۖRՋ! ^Km`[m3 m/~Iۮgmb·gm~qmOm`[{5v69sVֽ<m o?n{ `nQ]=m,x%)9mu`M{6D|zJo[v|'l]l+ۮ6_ȶw vͧ?'mW`O~ Gm7.*5O6޶ƍtO\ߔm>mް '-,7om׏`nd*~ l}m|1v>}pb쓯`QE{#_?m>mgmۡxo8`씯`Go(u>"5Y~oK_=`|zNz: |>!}vsgH޶7c[Yݰ vg|7uήNm >mypammx `|jۺ3Ϝm~ @=l10|nzrVe󒯎1vl{2B }xu<߶.;5mաˎ!.%mS71+v.2ۯ1 Vԏ%u osިnTΏ*пݨoTΏoh9Vcm F3mEh[{dmm F3mfmبl[7J/¶~lT-c%Zvƶ~lT-c%ZΗS$!l[F`?fQĶ;٨l[!ncۚwlF`?fQ|mcrm(ro? Q9ضoh9f^B۪crm(rMۤبl[7JIm3s4!)ˏ~~D˙h-= crm(rjl+Nˏ~~D˙ζۤبl[7Jl;7oemبl[7J޺q"Dcrm(r&&eҹ1ODcrm(r&zo͏~~D˙ȶTŻ F`?fQLc[7&v޶QQ9ضoh9Svڰm=?6*1-gȃI?lض~~D˙mP[m#c?6*1-g޶,;Wض~~D˙ƶ-sSm[1QQ9ضoh99ꐔǮ~["-crm(r|%uHS] Q9ضoh9tgps!Ѷc?6*1-g"*fm񋲰mm F3mg m;g$s$Wv)!xJ9%Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %ɼ6;$x+ORktI84 tq0}<53U,`ܢYVXei,efiԙ?MTlֵ'$ ۪-,Хϛ.ݖu+H/8.ٖ禑`\*jU -lAf ٶɾ9i,._@ K+Mcq]nn_`@6Ū}ുrՙٛq8r|l]־歉VfWVq5ŕSڕVZySW^aJKmi~.0Ѓm`iM lyժZ6oSSZSAǎlsxkb !2kG/uLh-*kGrs֗wxm]šQQTpXbInѬڮz+m۫itml݂ m?Ze}R٬+(ly~GQVT0$hUeWZr646veW~]a:Km4Dho[h͚=:/McK:V+*مm̸ʕ+/AѵF4]zi,޶EnKi" -wf$0p _[XP.l+޶;ӵfoitml>u++0L.~t~M@{B{Wo֎@=H~` مm΂iؠWX\>u+cZh\ެY{qMU|yHrf=q;5RNe8RHeVq4]1i9[4oalR^>L :ս@١1pfc87V؃mg{FOc^+S64+KtA:u sX-Qi|u6Kvi0CW/0pGcϯ|l.^=jk 7hʺDM1馱\rt#ei|x̴J6L9e6KvAe%i{zd4ZgMպx`[cm;VnVsT+0ocicl̐,Qߋ_?, tP㨬m{Ժx`[e/jƢc̊/i|DfLW`ּ+nKcWQfn_NFh$k%XnD^>Jއiy`UaSZ:4ݐӀa7㚕gre 妱4y}Оit2 i{L͙Fs;#}:46ƶ [$h֞bz*V˶7mlŶWJlnl;a)74}[amyXaRo4ڶ~Yʸit)vin k4fٻv'"@-&F.Q1`UU:XNm~Ke7MږzDo~nN}t!n]I4:@4=^%nu5F4.:ֶ [_dInѬ-l+7ζzlĶNۊoa[^_MriAƖ7?FY?]AN;#nb蝾bF& ^2w~ׯ8]@+'ƕe1`U,-o/4o>m70=vIDATm+(m?nh@mw?ez-ׅidet(|l*^~.ܢY;ٕm]mle}Ƕf޿mv>tcO,[ئOʶ$Mlhit=47][NctE)A,ukbJ^:(7z@[U禱3@Y'{O#+sԶ]&jU\$EvGc[H}#UMmojrYڶ3隺h0?vā=o~_izQ\kl;u [$h֎";dZwKrێ܇mv9ml+'\ö%z& mDld6#}%֟j`mobۿڡmYmImsߏm阍l;l;}ۺ m+7E^l[myHBnylw@Ӈb7ls]lrU2;Ŵ̽D@To~ LWv~QndFc' ~TN/e4>>(@@;ۮn|l&^/f[qĶ~Pm6Nn[^ධ8 Y_kO8mlb.Qn?5,njb̞#=l l+ae!=v :fXi[lb3[m-l+`rq|줁.7; D4 m;mc&m%[VŶ:neqZs~lssۊlűm׸mf'79n{>mmVmq̶LI.rJ`F$f-l{lƶ f@m[v_#EdW7v6̶ئom/Jmbmye˱ۇmB86 ٦_%my#OobQl`>wJl+䶽bۃlf[m]^FPld[s&r,~l3oeǶG6 /7mllaۇww".ale~Y]aCۊܶ~ئ+Tnh}]ɏmFj[%olwdq@so([V287cmͰ([7~re+*ԶF;r~lޓm&W?pmd[ۖm˷cOE֘Łel{lR6h=[Vmee[e#Ml;AeGُm}fGQ6xH>N c#̿Gb_ضw(ʲegpw](#iw6yͫ#&@+]:z[2&u_϶nwپLny_=FzMa"ƶwu=FzYmnob[mJ 6Hml[ 6lc%=6 6Hmoe [X#xؕ&xM vr& &xܻP|oe[mZԶVG77Ķ^lkQ2f/Ƕ>fǶU&=__4,&4r@B?us4ʸi$mN榱6& ͮگ8rQvMc1[g?cyh c7]Ul߹s2ml(^s@vY:&i,eiL*Rc0íA Z/͜fُm|}fm}ml3۶Q[ئ8i< ~]NRn mTM8?٭tXt_L. -3i,[Z[3#HmWA ZܢYˣ/NO:hn]cr4\4}UֻMcqIhgV7(wS:7] [WPUInѬ6[ mnp7Ml{7Ƕ&+¶GrBlLtOcikOѣmQi,2ӽWf nAhgm~蓵Mh T7v`xUz@眇 qKGM 0~m[m_G2sLL|hdl@:f fm`,-Oofi,fv%֔u2+̼\qҏA (_/KhYދ|_gV+7gMvg}wUoFN[4k[֙it2~EE>nd٩WV+}H`k0}4v`mW{7N7/$Ӷ=ؖ4Nc)cE/3/vvufgc87V+DN[4sISLcl++/4h <5/ogfp93oޕV4r@]4m[l"`A$EcY4Vٕ_IѮ0]v]#𖞯v m+mynKqZh43$BՊ "'-nE4o[mW~Arv5i,.aʖ禱T@WvEvŹil1Z` ّm݊i$޶Ʈfe#JoÕrS6vt{Kv AmyJkЎn}kvŝYkl+VTг#jUڕ߮l]`%M޿oW~ WX]@{[^mݸl]g\ZQAώlkVVmmg+9Xy %;cK8v mkmyJk]@]UQ2VT0$hƶ.W1y:׈V6.XypSWnDJk+xi,L m;ʳHFv+啫6ƶz!jM;su]`[iW~vUW 4v7ZkkcXyvI]׶zdTб'*Ur[fVca{5]Yk46Yj[^Z" tmW9s:iQZUC.l[dW*lﷲ-kIv{ܓ櫳4U)<+:ow"}d:WVKޗmGI2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@J"Pd6$3D$ %Ɍm(Ifl@I2c@JBYxQ?V(Ifl )Pضl;˟)1?(Ifl K5gJ>"~Qضlϔܞ"~Qضl+ϔ\"~Qضl@)%Ɍm{ayew$3ۚvƶ$3۪`[J‚ml-%Ɍm{aidƶp߶yedƶp߶i-%Ɍm{>J} $3l(Ifl T`NPضS$m/`ۧ@I2c^Odƶm%Ɍm{>J} $3l(Ifl #Olk>|w6Em(Ifl #*{sɔmef8__f2[~G?CglCI2c^vqI-]}X=z>CNyzztlp7_\>9$3ж6sMX6i?YWT~ѸgiU6{=6e^6dƶV]xLdY+uS@6|ыms(Ifl mjTmҠ{:cYZ @K6|QdZ;$3ж5})8v, ioEG6dƶ{:4e__Ϊb:;gg>͠$m/38l^Y<=;Bc[1лm3(Ifl m~sh6oU6;v޶ܝ=p]Ym_ 6dƶ{2=;3cJoL>o^*g*͠$m/ծMٖ{>Ix%s˽Y<6nȀm3(Ifl y9ms'}> _v"M$m/UW8I~sOU<ֶ6kEI2c^l4Sc[~P۫y/7%؏ kJn~q _\~|,m%Ɍm{ ]Kk[DF{.g>[ytJc~z'JB`[aCU_VmG']ya>׶߻yo/e'JB`m=RFyvFJQض\wTy]G]oڲeLjȧKٝ'lGI2c^vs{rJ?fU#Zo[enT>u_6dƶ0{CgP۴| l;lks$m/ 5vm{T=Je?Pضmkܑln* :c`[Iۆ/lAI2c^xĶ: ;6FI2c^m3̈㳶mCmӝ ll˱$39ۊ9N7Pmi\Pm(Ifl 3C5'@ݼmmv+$3:slw֛9\gGmlPض?mO/vj?{Uܑm9Nv%Ɍm{aڶ,6H:l%Ɍm{aҶQn7PGl%Ɍm{aҶQF&ǶQض&m+ؖl+\oc(Ifl VN4cl%Ɍm{aʶsm~fmc$3)۴mw4mk$3Yۚl;'Mmc$3)m؅ec(Ifl Sս&Vl;csPضBۜ }{rl,qmv&}Fmޣums(Ifl Msog;tmGo:G]QjۦQضF^^ac}ҏoOuCN|v%Ɍm{aͼwC׶ՁmD0g[n79գ- (Ifl mm7ڶ*fʜmfq_`4JȶGۖq'ތym(Ifl #2cBiߧaNDVDl;ij{GQU5{^Cw|/m%Ɍm{!"ٱK_N}ԡmh9gF4#ϩ߻zױ (w{I]ѹ lFI2c^l]{M;S`;R\cal~fPضm_$5a) /j<΀ms(Ifl m3 B~7_u,~޸[ߑ/20c JB`ٙ[mVAy7 fPضۺ,p}׻QOBJ?ha?>on̆7Ź fPضB {CCa= m\U";ZSstО}⯰m%Ɍm{!*~, Zq]+$u';V:۪*/}]}Z˝Cm+؍mŶ*ݧ|/;l[l+G+?ln>?E:+ն%%B~Rb*^g(Ѷ}$絭#Smxmm]R=R^;2ضVi۾IVYecʯi[ym#nlkjY'gm׶}gJEc`3X۪m׻mbmZmböm(>=46RakR1ÅlxM6 ³ v3w՜ͭQӭgm! ьm_Q0;σ,lS1ˮ-ٱvQ67@f <~׃dNjpYZפ0u<12)ҏo[|PLG|m>A1WTܭ kWt\ʦ6aȕC|á |{4i(}2q*kɺm5K[k퉯 l Omu'lkad[ mD € U8{m%Cq.yVzm))m:%A((ၛr˂vt^5qdKX7'n6ʅkl[9jͯm~cTʶ6~T}ؐ+_Sݼ7IIUu7KeCu=KuxJ27 j]'0j62=l ضfv J66LSI}ؐꛍBwfak{k>yVzm5vۯoc?ұ3-J;oS+tYJa_ mN~XF ^ө>lHe\c/"?J@O<]p6#`[n;s:ֺ[ ]Va_ mN~XJDl+|+8R M8a[=f-3PNҗ4np2-a `-8mc+\”9=6tz]NO`=۶rރ[Zg9ell cr(poǃwbtvAypoDz}7FT ~ssZ\qv\6#fﰔяrTK_  B-Bnk;g8-?T Vгm+OV ML_gq˭>rtݟ _f_^vPޡ{?d sO\,yT~ĶCpc[nz70i)s\˞lm.`A2\/&l;٦VB +ٶ9Euc][}G_.3'MC-1mOoB(u}LWZse}J>[Gk$x=3;ϱmonTʶ`-lR̙6s̙Y5,;kr֤ۆ]{.Pw=C6ȑmNwtaV9WrO:yWr=v7kagFP|~sUjl+nAZ VГmk_XLvt2;ؖMtXw4o?oYQS-3f٬vG7[z褝tm$wlџq՜ͯ6{%yTJTͱ۝`w AйkGц9NۥAZ VГmlS$Qu;*=Z8mƶ=зQeփg6u&kxmGwX[}l{ԍuf8jg[C՟sLm.`A財9YO;#ck\VO VГmtXbʶ/r憿ܹ^Qqmnwֺ+6 m|iMЯx6/ض}mZm;9a)1?|肽#]9ضSg?^_Ƕv챴#l^7msmv6N[}Um 4G6p):.M6Z_SW˼Am0fw/ͩA_J ,Ce|xL%Ƕvk`[mζ)6;7qo ~ƶ$Kql4Vmy[O.xS IQ5ol =EGØ6] ض< ]f| mȶ麇* zm~vlV9ʶd!vk[7mI37ܦڡ>gm-棶n Rm;ua2v[t l+f.ݕsc'MmYlEN# nܝض 9kb=aHmmZl+;y1e3 Xa@7jsf79%RU$z¶& no|ƶ#̗yw6a&loئ}6Wۂm݉#ۆ\й!C{HIVSm nl O2-Ťmvʶ;5[6#}Oj6ж-APJbێ_¶/WoY/Col?m@lemnPK}H_ xl{mk Ќ*Fk]V-Ӷd[l;DVdw ďn+1rB%Ri}[ߖsmcmmh[9VV7צW޷mhsXՕmv{2G4ۮkyMgWlkmmKV50ᬧ/Q=ҷuێ͜zq&z9omͤnl;EW7ӏv|4lhۛ_mW= 7k!`A\5 DjMrlmo-yBrMm[vp۷Mf2[гضM_Ҷ!`ABlp6f*zo_~%ok+ۂcs6mW @_oKcY [+m;-vm[`Slp!v7jeUg=unWo6\+96md!7-mdڇ.6BN Hlb78nNf+nd[Oaq8e[l{]҃[ dYy+ێ6܋k/ٖ9>R,HmvX' ?\"Ir6=?}\'LD~vslhV8T1OTr]w2nӶ ՜=aۊsS RC۩ ť Wf̶ׂͮ?͜mC76^;l DJ|/ɝa\ OӶUٌmUms`۶Z.U6ͭMm6lHn)*z}c۪ۂ]V[ȸ$V~O06m|+k!޶HIVѓ c o܌mu&'oi/V>Ŷg`O6lHn%RUƶOvfm{›Or{̙jk kaSfo =m!޶H_COv0;2gmͼm]pk#^ӷQ$d3SMlkbܳAm6v7k㒌mkmئ -k)of橶UOmn\ Ӷ Djz[!pm:}6y ͽmۆm5i0v/kI^vӶmH5\m–Տ'9i0v/j ?Vl$#l3BӶE};N>6-X [Xɓs)mVO^Aݳm[~ضvҶf6}~,u_sfqu-m+ymC\Jb۝KVٶyĶƜ˻᱾vAβ͆>?Է-X [V66n̤bmקKV鶕'٦H *Iڝ64ih[nJvʶѶkA4=Eۆ5m_>b[ue[mm+)&Y+-\MFJE 6vr~v;X F}굶ө1\m6+{mۮl{wζ}2m[Ulvm_ [ЬmXm'o=#aM?45]ۚvc︑G]mfۚOn[;2ͮWl6nmxm]$-|Dmն 3#Hm80p=ݶ);lcy;xNW ڦ+Nv ۢmm+:sa._5gJp&viێSOfl+?m:C`[.`r6W[ۆmmCemڀSK}Ƕvl7B#ۚ<PWymmFkmkًUTgbۆ2oa_$ҮfɽK$ ێ!6#v 6ӄmE_Ҟ9KƜvÞFeƁ3:ĵma5z +k%н9J٘le̞mؖ啳N%Rىmcr7混hMROnTeRͯ;|&skf.mC5uoW6HNOWOl[0u[zkrcpdAb[a63؇JٚN/'}_=ٴmYo9mٱٱ ؃! 9Ubw!~ucfi7FIˮm*3 l̤mC<=s DJ {8mzj6+^9i`]O6?V 27ëQj:1Zm~-mmvC*B[|c[;m>)ɂ62 Wmrʶ*E#(w`dbNXe݊mۂZ]6TFk4Z0{lʽ}jʰ~pfEfqK$ afuI5e[Nl ۚaQyZa;T9-<6ڶ6h޶C B&n kjSOö5,_&?UًmͰ=f-㶁mͽ(=aƶl >lnGi_76j΍8ۂڶ!`C뷽ټmz&HId/f;L֦h[ۺa1vRv`[yeCt:\6Tئ\Ok!Զ R{᷽oQK0 DJ {ͬC쫵)ۆoEomĪJ9) l 77O_=-ݾ7jꪛTav}SsvvƷy-Gpd&)ɂŶ} Nmf擶|/][fmUlӨ~m6_Mf/7mA|*{;9p3HId7U;B cm~m'ԭmrmUm 67ORaI6_MsVoouk_@g_omS{387updAvc[6wU/h&mK;[Z;sʏ7|hV%8Ԥ_0yB2h-+._omS{;l{yֶ*szpdAvc9țKTBۚ,_v~i*etmyOh[7tvۆjݓn`[6:mS3cُ't/)ɂζkq5P $iTM:x%a3Y{U? yzWo \tp+=jpg~$5ޏmq ,){IGwp VЛSUNPq/OimBW|J'h|%6vY?JRvE"4F| WbP~ ic}om堄+I;m`=l[۞'3ֶ=)~:m!u@~;ͿBj  $ö`[g= $ۣmy7ܐ [m!3/JRbk}yV϶>mۃm!el{mBb]σoUy74m $ۣm-;N?QѶoF؎džȃn$K ,uli%)h϶eض&!JRЎm;ߺGlϮFm $ضbѦCIm/#gW6yVvl[/Ojw/] li%)hǶՃmޠOi?*& JRЎmOk[mӫX`d[*g& JRЎm+j[m峫|6 yVvnUV`ۮɃc o13V\^m} *ŕm~RhV-H^Gz8wo_t͔tv_mm]fv-Hs𶙚conzf*rv߆AZI zmy;iei6{C[@s6>s7|[sWm=lb9j>e~ֶK+i]Wfz47w;mo;o[_Bym/zv߇NWTyV^cSɯm;n6s[sfLWmD|:r{mv͟Dl{0wae0ץZMgjmԽoqkۛbۡ:?7yS̶ fS5CWٖ5͵6wmUhbZ4M|7[J{sqvu6Utv+P^znl+>+|*{KnmygNjXOg,VسD'ӟ5Y!JRKl;۫k:nl״Cxvוlm>VsUʘIץ׶eb+ɧ̒g]3,\>h"i6Sq>AZI zm斧[{~ol3E5S];uUs'QrWœ1]S߈5oy~%+z/OG>6lI>al!JRKlk%ӵmigl3ѵxΆ[|v?pkE/:wUt~omEk%.n=;{twm[a]>c_ntauEhi7:ֶ} $6_wйVqk[]e=.4&(lo0-*weSwZQo~`Lܵ+6,]7Ӎ9?aߺLIl{sѳ>AZI zmZms 4Ɩmmp#;2wۦu$s{Zgѝr-m{3Krp'k̶] ]g}ȃۺMgcPٝl[߷#̗y'͜ºzԶZd.ѳ>AZI zm)S2ol;B+c/ֳ}g-wd]Lږdn]m31(%g|mwɃmMvɄmW;Z'ww cXF>4m[$V9ޣm'g 6ؤo|o2߷m~vߖܜ3wm>e6 a+u;>~īc۽}[ $mm[`11eiFm4dlGuaӶl>l?I뤏o!ۖfmɃlC:ҶZv8 Ƕ%LrCyT&o[ݷ%JR+m+1qm/Ƕ K=n 㶑m!t1cxvߖAZI zmmmΏv~_Og~6d3-i%)Ue[ۖ=d[msIټжpUwDi%)E&mY=Yl3=|+mgm[!JRЋl+\ͷޡmK]e[1kovAZI z{7-[Qۺo[5vm[!JRK߻mCeڶAZI z囀l -{ض,̑akl 6yWyV^&Æ viim+eڦ|[7&lBgl?CWo6x׶l~oi>1tvv߇OFZGl[4郶unGJ0\öVW m{xl3#%5E',m%޳lm} $ȶ78gM嶵vtDi%)hgoZm/mlkO{ܶnvr} $68oaۺ8m+zжr>!JRl72y`a}&vږΎL>mїx׶j/;'r2ş¶Oo[n&lz>&l˿m?o[>rv_Ubtm[=e[d4ټm|8tm[?/Fٖɲmm3@S@8u+-'$ݵ,M2/FWVV,zT^txۆnsƶjʶmviۇ.ƶ i%)hgYls߭mai[퉆;w1uwrflWmAD}ԶbAZI zmٔmGMw6WmfmwڝJ݌7ې3oma>8l{7c=d۪Nm;n}'6om{Эئ/m{w_i);jlKޕ\Of߻Ϯ XLEp@:֧*̦`6]Dj?twڏi%)E#efÍm{mm?j9nOčlӻԶ=i[_&s;ֶ~m"_=tn`~k[sm[m.M&#JRЫl;\MަqǶፚ7܋;}K;tE/ݘ`d{c̴m=־#i|lmuCi%)Uc%morWF-s^s^'g zol~fPpʶ߂t5Xcy.Mϣl3E5ói%)oK|^V] ^Me}ug&#o7uæ Vdnζ~omIbAZI zᛀ?@ڶj'g(P?g0|k[uʫ]Ҿ}n83n8y|"mgȃ·l3X3 <#ٹ 7}+S1|k[g Wv`^<9uUmdhm[?/Fئ[eVi`[}m[wuc(h/9ņ "Vm>ئnn,N~4־bmڨo6/Ff0;e[;6 ͑mx;xi4 S޶j$om^{6;+S>,X`}Ch}ζvY o[Xh[xmyV^c[gZ~ʶ7 Ugsz@og&rN{c+_x [`m֗5ّɋl oY#v];fbAZI zm\9oo"jK/ycř1MjQ3XFɭ;Bm3Ovg4To՛7Fï멶$|Sͅkm8;s39viQLa4yV^dȃml$JR}?*l#JR}?-bIENDB`Dd.   C Ansmo subsumption extended features subsumed results posb!zb|ޫ& nzb|ޫ&PNG  IHDR` s.sRGBgAMA a cHRMz&u0`:pQ<PLTEٟ pHYs\F\FCAtIME #ɐ!tEXtSoftwareTIFF Image Printer 7.0f;ZIDATxJcL&[By-X@#[GhNBG)X#4#c WbHӇ4F"7}u*}vw:IѸ23/[{8 l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l VEzp/?lջ fl b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l Ƈ q6lU9|:#2'[en=~k޺f^.]mNٞl-MGJ(}|6*[y--}W9=ʦi)[v M]ٮZo%58eyQ_8;PϿ_Կ[] 8f{·WtB?UQ0_W89mʖXG%[t[[6*퐍v.RllNDIdhȶ'1P2]XdhlT 퐏"F;`s (|x+[P}@&F{`e7O6F{`{λ2l+C22+rق+? Vbŗ4Ne[|Ej )NeSsɼ8M˶-ѡR~~+Cmȶ-uȶ>-S!۹&⫠g°c=5 o"#:G admu#f~.[%lGتʉ9?.gD7d[]vK fm}+[s6G٣l*1JU l`G>,n>lÿK.,“* W}tf'٣lsN~W-ɖ [vzx~݂rzS.~h}4gŞZ#MWd .Iɏ-c7٣lFl^*leG6sA(SBռEz)[QMY xݑdˮoimUQZQdR@ڸ[)jǐ-Bۓ-Ne\&[)v Eڻ]ʦgIہzIjb"[l'Fڻn֛@ [d{>gO v^ v!fyxA!fI >~l%46Ky;1Br>oC ftl%ܠ[6KTD 6K铻7m6>7d,d-Ud,As 탬l~-_]Y#m(#=Am:g VyT-UȶUӧ`q %ȶUt{vcq >FٶJjm "VW5nٶ]g'jf pJlԏ烔"V>g}v޸= 6@ԛmcLl[%(iOf v>mR^l[%۸ݽl[ݸݽ1m mǘl-B63d\" v,ȶQɡ5>d( WQclnjd'2QclŴ;}G6.޸}#B2#y_~#B}4 mdFj=l-ͮ$mxFP;/d&sև?}zQAF#x"vd gjV+·}Lp-} Eׯfy@ ۜ0[~lWod ^Rc4c'A ۜ0[8l} _^Rcn>lslfA6Y󖾮ͷ}D2f 0ۣ=;| efC6cYa6d3zfC6c )փFMAEu'k0#e3z_ i0| e{ eVL8&m^Y1݂囿-m͊dċodۼlJ' vķm^Y{./vȶyٖ,>?lodۼlKF6 uwoIg%*tamm-(_国^uLJ(Pm- &_ea1jU?J3dظlKlWf/ QfLd- &_d4e5~ĄVqٖٖ$553 \q<;ۖmI5Q[ׁW!m\kWFx5< ٞl[Ean 5[ެ揞Ӹx-m\Ea˖ N5=;lmQ5ă2[xJdm^5_Lk ٞl[Ea 5d+ojl[Ea$Gn "#[im-|"_~>5K-oca˺mȶq2 _UߢS3Elcl[CW^m]=m>i'9_?&ޗ?;|MߝȖn\ŗj{{i{!mq V?޽?&[8nzTl)Fm˶0|1:5SyD/ȶmt_??{5.òٶ-˜*?o}rx{~kDm۲-+L_?z?m_V,lۖm2/WʹźAM Uݑ˦6-%*&ؗ??6.9[mad*J?˿?XZ5Osl9[mad|_OƲT mZaWl?dK-0+&[6UQzdf!ۦe[f{Erԯ~52-E-˶8|?x7:-lmqOO:"SCв3i"mYaW$G班 ´l~ٶ,0+UZC@6S6,0+U}iU} )"'班z6d3eòY^?*uVff{AO?j/dbt{?.ۉnve:W[s}dX#]U]&pLkӳ!E1+]yʯo9?[,g2f7|_xsY33)œە&'ϭ9uQ֝9Q^]lb/H[ASG]/9ە&|_Bx?_ʏK֌s̸ќfQ: _\sߢs/_T<j?]3RޞOeodۮlva6QU˦+mW60|?YLJW"lە.~Ivȶ]lmY| fe 9W@ffsX֐mنٜ竬l6<_n6+m䓾mنٜ'G} fe 9W@ffsm7lەf *vȶUD竬ٶ*| Ves0䓾m9p9N@`^*!VeW@fEq+M*vȶQ٬K'nUmYN*N>iCȶQ٬K'nmٗNUmٗN'm7lEqrԷmT6a6*vȶMٜ竬l۔Ibim9 9WYmS6'a6*vȶMٜO>nٶ)0odۦlNlUm9 9W@M&̆l [0|odۤllnUֻlvN>i)dۤlj&G7lٔ0.Um9W?1>lAg]0&-d#%G1^~)ǔP=~/Tgy6 :WO6*wd/q/[٦22_W?ɖxyx %v);Ekٮ%kG/v ~Әl[~\<ǒMWt[2u]깙N~^6_%⟲EYaRa+#me.yb)~Ǥ-[y}]2v|xnMQ)[;M7lOMd1k! J'Sgײ-R%F[qKy &2Vї/_-KlU~SgZYxs2ɦ'bS~ANݰվN[e{IU칑l֦9dSZKUaa㝶N%a6WFJǒ~N+n:pr'[ak|N0|{=HxTLM?LnY6kփ[eWG^N0|{IƟ]OdT<LXt2*H]O>k&*Ue{M~r>V%?-?蠐-e;eYlAak~v0|Q8lz[8t2१ak~v0|s:j~g}BNǒ-g%RnE$PUvڲT~MSzI锸JUG RҢf|}n^fs.)?CgMNe`E©k)58(l5Xƪg{MUy95}w}ղǒ-O֔:_v.&鯙llcv0|?sэ|6lCm-H*Z-ۋlU@6 Ehl/ 9WYo,dlU{tY z#=uavȶ\VE.tVV_1B9IisȶXbjUkfeSlU#s$G} buy*z([|Kʷ{pm7lKe.oWGޛ7@gqT'*vȶTNdEe a-J{*l|ٺ5-x-UNUml [ٲţ#-7σ%Iʷ͐HVlc+hM6mesqn̅#bčgUFkmZ$9nfWL>_g/'{ֳ"mZ$_e d#[# @dժ>&.ֲ-~v6CnaQ6ꥥlNml+/ղ݆ˌlcR[\| ʪ]ֲ߱]6+uZKxiV6*v6Gb&e;7+ge.×e F>t@ec- \J->7.Umlɉ ZgQPN/-|od![e{IV=[K+#]ڎ6-Z6KmlɊm֐\ުl\$G} M5l}}N[CKmEʷ͒jаlի:RۖEz f.C6ݰl5F([Rۖ ٺl\+z[ށ ٚKղesm7le%a-[oƪ.[g"ƥԶesm7lsd Ơao٦)̨_J[6O@6seu8sV%KUͥArԷ͐pPu}-ԊBۖAʷv,mlҙ4u)9nݤamjZ6Qvf.[dְ 6b`Pesm7lsdjҰ 6b𬓱Aʷ6i>l͐wol@N.u _ed3-L 6XŶEGA}ds_|tbq '} l'X:7 9nf76X=^طl*vf,06MLh߲| پ~z1x`xv6}fm7lƲ%F%X5O02NlپeW@}n԰75[6|od3-.K':gPlL>negp- m8c5}fm7lƲCƲ5l*#پnT6`mj e>_[e̊GMa߲![mW h nsl*f.[^1;Qن2VS<ۙ'S*6W6uc@j*D6|OlƲXM\f> mlc>Tͅl*\-;m}|#s&FĽIf-ߏle"g&f-ߏl/mj3/'f|;CSWތфV֋M6cDUWތQ2vlMv 6.lUۑ-u.6ĶM^n8͘~ft׉d&c,cQ}w}>*֓*,h)m7{Y֏Ql3_M6}EfJ'M/<1}>OU7ͱlSSu3VSŎf26,<̲4Rle3ΎSnzElͳc<ׇɋ nhu6_AU=㺑ͱl&ƛٲݖa䓍wCxy^L?&ޭj#\F3~3Vfa6먬mr}X?f7wdX6W;cefiM 7vr0Bscf\-le3 ӷ3Vfa6e1͌/̓~̌K2F W+cef^w1{ǚ0~8Z%̸E6:mfa6˖olW_xG&˘|fw|Na:`cp֓lOVélS53Va6'IK>ŗZ![#cefNX~礕 Nb Wdk`>0w39}DۑvGM٢3R6z|.+˲Q3^/:ne3x2d;~d|nQQx5%cy{c[aZ Kgi#[ "ؖ}M>Xk>U?s[Zٌe371%eۜZ۲_o͐l}F6>l)9m-~lv*eS2Ғ:Ήգ1mr,>U~]dK룎CfM;mw3-L>&FGd_ͥlN[ж.9Z;ӧzA֛A#l& Xl}v\kDz֓MebeXjU3ꟈlΛJeF Ulkڜ8}Dz'\6oXݍ˭1JA8ò5.RDF??  uٚegd3ͤ*|f(E9ZfkZ+>dk| mf%lތ0}هUWyl=֔ѱ@6S ΍qQxul۲1S]Zn)[ٜ66~f[a5d:py{PF`6Eg٬ >led3͠*|ěf. N@Y[rz&'Zބl.ey_g5֫!䫞;fd7&l&sT6]?%Uhi #|sGed_iNle?x)J X}ꧤ#dO,AUeKfލZ}NV;ozrcvϧWd3UQi^lZ6;GweƟ/sv?w@PSi3e6#[t>ϥZjfPZ.s`[as;?OF[Lߟ˳@ɶ <FWdlS![g-E5Uؖ}Ӝ?@Rl-d3 1$j pYvZa3/{z)Dd>>n+?sjoeSL6BaЛWk]a3?v2 @#;~6_m&÷hds(ۂcb_a-?Le3 _p >7_Nc]a1oN%ixs ڬ>,G#rܷ,ymStg/L <8:m5Fهٜɶ Wf]aX'#[j*tUnRNnlaOFԝl bieGYr[ e޳Ax(e'}OFԝl3-eymهErԷdd3E n^ӷdds'ۂ½=w벏ld[Z"m|of(t9[f]c\ɶ$Z@6벏*PB%:"|oȶl Be˓'#3lƚc|}G63٦ u6[[ٖ竖\fxʌ#[fe.Ե4 [ٖl?\f ?gqʷuds$["%mofmVcqr͑l.Xmwd,ԵZjeS`ds$PLA)&uP8_5u &A63&wL(`j*ϖ#ͺCyd[}هuȶ5b {sn@6Y}5 Ub1]WAqΎ@&e:(A,ԍeYn`IX!kd`Z6%o M,˵ ,lL,mɪp*[Yo⠲ >9[*ԕ:m9#"SI>Dbdkc>l7Im#@pP RwG?ͅl O/B uc>"dkc\&uѼfc&u` ŏeo3mj^]|o ͶyAeZ}![/1]5F*հK:m˄l1]eTqozٟlP+$҆lLTK'}M}깎E6WPZ6u̢5yæ2wi۱lON_ӲM/<@6}.4ҲN.e Twx5KleX(_U4ll_~eQZJfVz[VXؾ$_U6llN.e#5 RO ƫ_1m5&75eVWe ~Kٶꐗ`_cdUI/Ӷ_"J˖]C|UT -dKv+[Pв}/&u!tٰkӶ_ټk%d/󶲉t]}Ll؞oriۣlKAl  (W-pTE=~^`)|LAdaKvv+[&TMdCm|֐r6kRVpe[!c* ۰H6C;VH Xx<.试tiۣlIC6) VX}>PӶKT7)%[Uad+ 1]eѡ"p6d 1]ei@=Olٗ}GQeS7W@z*'-0meF'# N.eD):NeVEY+)ӷglkt]6= FWEsiۥl^;0m*|>ohu*_{3u>+,e[%cPamU^1l+bwhsPcԾшsS=lW7tӟ8eƱ*3ӶGb.[x!X\Qxv [+_Yt([gmS)zT& ~s죹36uŽtd#7MlO[Xg uby ;(hha;V2PdJ M,:lYE{Mjy .:mMOx2뿙6vRbeVktv)% e=[v粙NXLkh?L,tfӶG'ZGQ{!XXL7:7m5FixGo .:mz5E7OjՁlr1]_k}rwo:m{-e<]xK͋'!O@\npnӶGT~9׮nŐ9-§e'o~嗺l> o\t(CՆz/:MG@`Ql={Y9asiۣlzd*R|cOf]6eq!ۂEm+]1"5<\Qx`j_[аRH Y2}~U¬M[ڰwڐHUYl} } ٺ&W#Mh/ Yf}mPWr-n$ z6NuMrָQ'5FMfiC.#IzE^>3.ݮӆl]FΉꏁGySciC.#DrǠ}dӰvڐpꏝIh>v m ٺlCGG6X]:m6g$UͫB `iC9$@%}G}j%] dL~?^mݮ3w!ۜg%ck޳}nwV6ddL7zOȚu}![B][y5zZ'K٬:maLձ"<έȚuGdXuڐÈl Ym. ?nyKkiCÅ1['f]ZfRNuM2Ǿ6~g=LWc![B]ɘY.![a$cM3>-a4úMӆl uEg5Tٖ}fiCòj;l>tՒvNu>%1>l>{g=6d0|JD/۲]"gNu<%EWOKb'-![D:ЕnrlrwڐÈl~*EYe̍l;mfGE!qZHŝ6d3#B;ر Kk|d=u:ۇ#ْ#flQxؒn܄>T}f|L}l lm ue‹kVrNM(\єB.XiC6@h'G(` / W4?(hY &^hgvaxY mlEDzN-/ O;WQuڐͰlEi@etuڐ`U vrBNXiC6@vӆlmψ|Qx;XQ.9f茬m(E6dk3tFVesuڐPU Ezw'6Rӆlmd[%mFE6d3ݡUL6>e6d3ݡ6ӭ},!m_8RӆlϭmgWrƂN>g;jb7/[)Huڐ`* l^ӆl-d[)оpnG6dk1TR}JGӆl-d[)۾pn#]iCCdr1sO *])۾pncI ZlOkc7.RfwڐPJ v[fݜiCC@h˶6dk1$J1N&~3eU-Z bd+:mbd7Sᵾ [Bx5ٚ }6n$(glE}6ddk1 j V[贙G lVڔl" d[-k+Sht-4nbm6TcTtdcl1ݖl>Ҽfj m˦:m![L|CGu+ɀl%zmCw FmPwn[-}= 5mn;9J"o׺E&It۲-(Vٚ 0@hg3l̗mB+9Tu<75(]1 Qֲ]LfqGmŘn+5ٲNWfٚ˶fLTGiCFd?1In! ٚ>R`l#_f{2}2>lMe[3oC>mٲm'G74&볙~f;kt۹meg3.hC6ݑo#~5}B6gV, OЍɦݬ?ٜmE[m}'ӆ˶nL2'AdLm4v+UbZdk/ۚEF|[Sˤwϼ f>u=}h35پ~5-hCU&m}J]Ӳdk/ۚEwVG}Zdk_nLܛiBz +2ټ٦R{-s3Wl6lf/ >1ceL"^mƲIlS-4>6ߍl[meq\Tqls.D[=ESD/m"3el٦h\9}m<قq@VlS-4^/x*~/鄕韻l-z +D@I"[$lS-4%elȖLdh6S{Vg*kFw"5I<9`s+?B _1nD6%[lSfc#?ŲNҶt[[mrl>E3߇ly:-6`sS˔ v]VA0^_NUl 67e:t5.d+FW/Ae1,YʚFlB]Q٦ln$GVhfh*~7fܢl 67_Qޫ/]e͐Ȗ@n^]MULcދlLZclS1ݖ_> fXe^dSe.k67T_,X*~'So)[PWbZ_Sa,]e͐vɏmm2k2Ui*kr|n'0k71d:knQ_c2|/1=~zNl1V&ޟ|Cgu*k!e6(tLLUWec}{[l=UrSNtۙۯ޼p-cFRͪ|՜⬕_X]6 6@h53Y+,}mX*ݛ VVs^l#GMؠl&y)^-KXDNjCRO2m6+TO"c2|7?_*WZ1VlO6n+9:/ֶⲏTNdi\;-[\L2R}Wu u&[i]ʚ1Õqmvu-({ԏQk}̖-ymκQq#9KG=ⲏT>dk-*kӏo:ۓ$L4Jq&;Em`2|eݐץe c]Cqz uOLbDӬؼh$ٲ8>-[Pצe_r)ߞl&1fiV3w^]1\قK#%e=~l^]l=UrtY)px }]ꑤ%{ž~iٮX*|lRYP^]1\ Zk+&lN6(4/m0[2ɓ3PEաS+&Qh⌛W ߅l za.[]e A03;׺vjt8òG%[W/蓭Lf8qD#?#6]~u`e.dl7M*(ۊHq&{2p*~75Ol'-[UaqeGKSV5F;Z(,*WL6B+9jLZ}W w"2%2m#KV4e3:C1 ߅l:UV'?r鞲=^zvDL6n3jZ ͫ>SMd+l f[#Eeuj_.Pe~ddS7c*#tճX6!1ݥyP*~ ~ AdKedke⍷j^`R;oT}ܔ:^)dˎQE e 76#dY֟߉lV~) TW~Vl+A06#de=Kc2|%zeK%FlsXލ7bϫ].bGl+A0O #MN3!|jT>dKaOlm͈6#MF{/~Cq3S|)l V m RٗO-ӟ߉lQ/XVhlUft]{eVmNd{Hei[ꏽ?a szFDxLės~l+A@hv Njzn7b{]D5n1GVmF%t^>([v*[vۜe[*|FLm` z*w)Iumkv}ʺoE7oIt]鲏d`*TeuTy[1*#-]mNLٵ: H P7IeKt4Pqި\0qN˲pSل ̓ e}=ЉR-YTVHG'ۜB5)[8O-UH~xJK0$tt(_ymNMضDGO*~GOYrn*91VokJd2}63ɖq~HHY1V&~bf1;هlzK˻$WՄYcˬW>+mMe!<}5e< Wp* 6`uS-5uF':M&9<e}B6y frt* {IC6V6}/`^Lyٞ ne.d{|`tgiN*rHF1f4uF=ڕỐśS5c$LU|'ltd{0vj%>:Ȗ-we[mB`4I;}ȦS8 67lƲNHG;#ٮo3M*|nL&6JEI;-mFna2CF1ݦ&cYR죝߉lGmnL%ۈhoPbjNe~d[ʶ5:hv\y7>lsHi,Y }+ÑM l#Yd@Jf*~/U3]0;짍lk,SѪ GmvLH)XHf z UϏ2;oYUNipв5.͗w2hw![}P'cT죵;/l A@hfN*eV*~']sem^:N\!;zlUϏdz8%LQ3 ْ+n9u v_ƨEjtՇ~J<&d>]v5aAy-<B5*ÑMl L^'Lf*~eW7q_#ZU bl#l2EI}|Pt[ U죱'-V#גmALٚ eIEI~K}lkU[6uJVbeT>dKnfZ-I 4ZT*ZYĹ[j5BO]D% `M&jNdd AXmeo}/LV}4*w#ۜ#[NHȶ(LkLU1j6$t&UXG=l;eٮݼnC܅l̷eKKȶ(7>2m&IT/J 4:M'Re]6h՗tmQL)Xoi:Y}S;:m[mQLi:M'Rez*~0&Xm ֗u2HVr5Fvekd{1dU*WQKD6 d _@h 't~lT< &ZɪT죖Gt5tѾdU*(۳ ljCt%>j][֓tmaLi^UbSˤT}pډ(2i@Ն ,N 4-LU5Fe*V8 *ɲ2֒myLo}VN/}1xd[*|yL"k&U`GEd[mȚU`GC)[Z:E3+2|ͥ:C'Qf"8YM-xd[K 1f,'XQT#ZU1f&ČM-S5m-,bh#0k>'2֒"ېU,WQmpQ`cI}pd[I6B#jH'RѲT<}HZ"`NJrj4 GeI dT?κ?1Sͮc!6 fJEkz*۸|TN꣼(T3ב,Е>`mYɪTV6wP@el]zf>͋]?UC^_n@hf˝J%I_.e]leK 4Um^*-P3o@t[yɪTtjwl A6"¼dU*[}I?2G'zlVؼdU*:֩ w*[q*'5Fe^6BU$&d>TËlf@hvdlaYod}sb *6n]<-_v#ۜҦlk A6E"`n*M]?o)X_6ۘnoYJe>6~![_C?7v3? 8roEa.mLg'Rٲl=j͍6CI^^n@˖{dU*:Ov1~[ïU  "  92y˶)*߭[Rdn-92$>…\VybӲI}>mE֖AIg[Iԗtǖ/ZپH s6-2],7ٮom7?b1ݺl-dҔ[fWf\|1Z&ަeZa{[7~[]0WUƣ"[Тe1zeq/4ձl-`ǮdS[ 8r~$#Y%Gn^dO6%ۂ!u7nj-*,Z66j99JQyِ͜VQU 's ]GMٜ הItMnjd돩-اд~JKhղɕ} o {^lƪeZf/]#6a\ cUul6-2Xr.e{x_ ۫H29jS&; uO!]\&ѭZ6ѩe/[OLmlG;< BD곉l~UwOgB֬EnlerԪeC6c~<f.[/-:,星GQfӲN-s8oFoW6TMtj}m- ~2x_WlfZ6ѩe![|UM_0hVᤞ)M&:>d.K׮r##k=3fQkԳN-‹͠fX_c/F~UH5YlSC@edSt+[݅ZfyY N6{l3G9b3vI Zjo$ݲݑm&_ZSJG}N6ѩen֛X_6 ]l;CfލQd3ƍleSBM.ΖMcZESfLveH5Y&ߍl2 A~/FM@jMc'] *inaՇT2;ׅbdsS'fGwln5FMFg\RM_h)=]FFWltrfs3Xh?-[>3ntAe-]U652ƍ^m,#IUx}27Fl.ǂealw~cxqTs}6\I@Vnjd>;[Gv5Cl>.f>Q :l9 M-#>yo3g ͘: }Cq;ٲn|qS10>#ۢ !jԽ; I B6ݢܨ 3cga6712/pQO6_0aoM!!?X$ɖ/f|qt9ELc%vbl}73OI߇3.bde;Mq1~۸l3 [戛 } [7a ev![>mOVr܈H\KuIrT=e$l!Wcق|)ͬ %&}ȦSd m] -͌c2+V6'\\>![:ft\YUɜD6:Ufv_&/WMyNulnrSDy6*|.j-d\ErT̑m\Y'eR{{lgs RB{lmfȶnJcFeqFV-/8܋}[6!wtŦٍlm.XNd[lw_|c\UJCvɶvJEc-]nڹ*ν2mznlPnڹ*νX>d\[OQv#E[#VZvw^ꚷnJ6-tٜ͈FvweӓUԍl[貹KBJ^]+|m\Erٌ?L>Ff* >v![{ٲkhVsU xev"yBo% J٬U`#flw)vcp!6ld:V[rPvmFmt\jv"۵>su۩C_kB2pXD̾D6>5})[~YUB-~ܬ7!Vc Zvw#YTo"Wp*I}CrBly'M^M(flwyu;i:K\M\!UٚK@ lUQ%EUVlNRSZ6_KdL͉lRe{-nV=;Yl}LrD:vK٣leW,x~Ύf6a*}cϲ}{~@&álUu 2-ɖ_@S9ogQ%.D1ڇl8[;}+SxҲ%eFyQ~,c?=3o lAv*LKRKl&lZv?~ޔ=NF1fT.PNdUl2ozgl뮴‰lW]݉leBVySK$DYaTFm&Wp!P.dk-}WEțW ?tD6!U)\$Gىl劉-oʲC=H9U5Fl}lU)\"TG٢.MY {"ͤ*|3EnbBֲU,kTkuJC b&m>HhVRNn#/Hp+|T l겹Ĩ([UŒ[s l(Ww5 QZk-۝4d'2VwɜmSMrTcuV=S=\ɶ.#a>d,#H -MJ֕-9G~/d[6!T#uS!3؉leDl7m+Wj ^+[ܕ' l+Wb ֦GٶRȢQײR\rfMζrU 'ĞQWb%ψnٖ A^Ir٦1tLbl}#ٶes"Lad lKm:_fj=b4꽹mk*}22eMOœM ٶGB6d#ONmUK l.WIlʔ}H.zߓ6V^B6{?=P\FW)ۂ-vٜ$6ej#]UyA-mO.m]67&~d{]=l4-19--dJM\䚐mk/2뫊綦_/ʋ47ÕBbYobEuٚQ#}P&/6lifH^d N.vX"SDmwLh'E d;ӱ\UZz)؉l#Ɔ l3Wub&Q<o[&6sU Ͼ3)R"ow1?ooe^Mxw](VliBc%3e䑽yl4mDمl3òUo4W~Om ?uӟZ$FsUH~Zw'% ΖO16R{}u+Qc"g AjJnݿ1ڃl"m]{_DdWѯNd p].Ȧgo^-VsUlݡD)~Q]'ְlUkH[$>v"F!U0E~2a_>d{K'ְlU0ۯZ}P#7[^um2_ [Xm˖k >'[&ڋer὿Ze8lWW ͭQ?_~W[OE-ۜ!Uj}OE}9m8WφlSP9W;4aùn>[᧬Y}?lKlsdny;gÅy8lK23 lVd/>ز-ɍ AH.7e+rQee[Pa$[@r*->`XVl_޿GxQ=YUNOo9WwRճYT^6ZefE>dù4-*u獜aYl 6R1w ?ef6!PצwيUYVct-}l.[1n4~,+8\˶hDQUCEsm>bDcDN]vmY%Ҵ\&#~,+8^6:j,[n񶝫RZe'w[eSز--H?YPe]4[ &v0)jز_Ť*}-^OKd>ز5׮TS 1_,Ϗe-UclGlْOF؛WяEeGg tjQǢClo7-:9DC`eQ!l5'QT~XD?M-s̖-榩ljؒl,~,*8f6; }Gknj,/k[4A[Z/`-~,1:h˖یe2KdyYےA[YW"Hճd@`i-)HڲWەލZD?vlfD'}8biYے(oމlO: K G*#s^M,((ew[j iE?}56|.Y?ي]eKOsnG*a6}K>bwkf6|.?x0̦Y2o2Qitd֣-!v5E?f}=i{oɇ5˒A}yB]{c%яeIu~}PL zZbA)75a!X͏e2{ӷ-W趙؝3! 'Ae GHg- 5Z?==l{F*[qؽOiQT!e뭕~A(яh1-?ES?eWH.;ǐnq*>ƇQwJbG4Ey_'7mk~}tC-:%5*{=lAߚl"'m&SˌܯOR߿wNWYӐ&V'{u{IYӐ"~_?'!CU{Fg01mGg9lmwǨF1 >F}=l_v !;#WqKoR?\K&2!elL^n΍fP1ݎRH.GT?]^2)Dphl8aek\=tS=lcMMf:6'bjDEni[ϝdxv;+RmSL@?R?lQ|OJ:(!Cƍmvhcrۑ4=(:>">mJF'꣬Q>&{#ڳWјg[% U#\;*ʲMtOcvj(̓q׶rjzڭpۇl_=vԭlIQf~z=,?l9*|ϣ :m+[ezRa ÷Ȗd{H7U|@sl rl{WyWʶKO_?E!sC0x;l9y!]ESӶBޅIz6x;l9lWһ*_(]+[O"[N&Cz"\N!(d7A$/HN6f0}l9lVS;=+;yIdˉ7IUr}ܽ?\ Sǭ/DtG *Q֙ihvdˉ;7M$=G3=1TQtT6oGMsQ|ƨ>:ζ?4t(b? 1aW%ގ"O2IOnԻ_()?1 S>9j_egZ]F=|ȦO2/J݁) AO-s x,-a?(ٺFCz Buo8zȤijy= +RdKu-iJ\sdkퟍ>+\GMU"[nfjr{ގ"ôg=8U5LFuݤmz+|l< ~0}ֹvnG/]žȗ>k_G;˦@U5fh_G;Ml jtgT܎^6upĽٷԺvF]|b6dkЖ ^Go;ll nIb`a:wt/Cĭt w:ڮJٲ#}~AUϐ.U]ʩeh~rymd g.4;ڜ֡,XxPnGw*R%۷"Vb˹*N5{Gm1 t[d`&~r-[ҕ~+BmhvK^aF?uy;cيK.)[~EUxP_~8+b?N#u(d[|U=l9\W:-pV%J<5}{#hU?yls\7r~oqPz|ٮlۯA?{#,cWme}-[~llO t)e-7=*4enlأ"[B6B:ZR-d'˖x_TX;_A9lkl׿_R֏etQ1* )Z~y:Wr8.eœнSi%)﷕0KѲ)p]vQel^^e*9e+u)Z6Da|7ۜl*Zd鹯OO~R.yf}bhY?KNjY9-[} .Z~-dI^#[eXo+Oeu4MuewedޅT/G.yvNq-QK^S-[vN˲v!]˻hZ]G=ʖtE>9k/_kD"[W-[q-nG([~lwK.L28ulW Ѣ|Jp/U75QZ!]luݥlQ5G&[^( (uqֲKْF,Jbboy]OpֲѼ~rM}O;-veӠvSl]ܵlݚ?lj ¥OS;zMh>Eql:ߎQ65nvTwqg+~Wj![-eSs}l]\l:CG9\l:{.pUV.`lu4>T16uqڲ|Ǔ-3CzpUϖ$QezpڲhxT Sރ>ޏ*[ p۲eз5J-娲5iTeSQSr-j/BeS!8gӨ?yۋ冐m36 ~azV/81a v+[r%M٢Pm,[1FU<-kZ *Q6}T &m R6S]Կta0ҲQ.~eBͰ[٪@G%3?lGltvDy#fe7[q[ogpyv)K&[7eeKmmv)o4Ǣ,x~vYal+GSC^8JYae*\]eZßߞFȶ{͏敺J0Bثllz(yl AX]vzdy7]ЏwnUeixi"RK^R9H* e{>%JQ6Z8ll+qDhV∲Ѳee[#F=JQ6Z8lV∲Ѳee[#F˶Gm%(-JQ6f\ d1 @6@ d1 @6〲l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l QdSz?mv,[x)ixq~+Auإlr[5o_!*W7%Mmv)dIgF9} WST=q.Guدl.zav Mme+[P[ɥ|ayP7e5hK<e.=g=0_ *VMPnUlgIl[`e\7- *Q6}Tȶ-v) U]lU~e>l`lOÐm([/[Ul+qȶ6-|-0Bإl.ln*d[][-1/ et{V!JQgFe #d[]eZBߞFȶ{͏敺lw~V!JU[Sw?al-,sݪ#d[=y齬 Gٮ[^˖eKn5o)[ֆ jPLK5d ,d[}_kiڐ[||C)?.;d겝kҲ>eK*e{PQd{ FSs7ƛ4?_ ƛ?_雾 1d~WSo/Div+jrl)Yc艮l1)m1?l)Yd6e#Z6Rl)Y-) [-Gos.MBRdGo#UbTvʧB6 -GoudWMs Y% [}g=tYs^ZLM ȶU]u߂KpIՒ|+!r~6B|7eYDm$28J褣acwٟdf?}jk~51o-~ߑ9f^:drG?=[ 5Gs[_>zl||(B%9 CVd%/;lo?@ |k{)ɨlr 2^6q>lՉξ%㡡Ey7ָ/oM|/i5/m\.Qlc]oPoM|4ߚi@gt5T=_6q>l5.\6?/nqNrt˷_>O<)Men3oy>J65uǾol?k|fl|kcǞwز=Z stIbkC^6q>lCWSV ev.zlgײ] ecv3M\ȹl%kvROQ6l ee{Ӳ K**?*P61Nv9lL'gu#rfgh)Hw#:xS{J޻5Z4IDtMތQmkC1˦I=0-4/6ljif$~eWHk4l*-9y6ljlɚM7FƲݲnlle!Ʋ'c3fCq,WQ!ƲNe{ff&zX60-:ɖ̒-4-L6}Q]m cԜMdVMmD6504"n6ljC3tۗm'>F1MOl56iY"ư,Mo-;qfϛMOM_>M-Zx|(/9Ʒ:·-~!*NOw?55Wښ~*Uͮj /Vn\ǖC|kˇ g/ޢ1sv&(_?Do-a_6q>ljQ`tQBre~Zo>M<_N(Pb>7PTS>4fz6drH#]YpRõr\]κUs1鵫|{kW-6vUjde3\:`d\qb Hl&K@.\R4]6hq[ښmofϭM~*krq\ mc[N$M.flwCyy3Y;ulw.e#eavr1P![x[[ Ż[W(Zdm-8̷L虝ïVGX)\u>- %ϭ&gQoM:YfĤBo>yO6 l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b l b qZُYls@6+ze1٬6GxY+fNdWtt F }=ldF }%6YvGAG9_! Ȇl3@6+3lC 6dlV  dYls@6+*9P%O/c^z>A [,ɜ{z!d?7-->mDwE:敏̳l_q_=SRJ?Pe Qf*@ ZPeeބM =J;V:'mZ6Tlu8@Ҥ!6,ʭ:˖?kݼM"[dk~CΪz+SߙN*z \l5-?@֖-ѩG%=+]Ly-U˦_ی"fEMBkC6/ݹo*d_*3pb$}irylq[P#[ d󊇪vʮ]6 ^lZ- Anue~fE%T:eoqLw <#( [)[S&rJ3em~.#8)(eS?mn)@;#[de TK jɖ5[*)@)wUVi}i]rW l}M%M3ފj{˦,Alѷ)I}JdklZ-t9 愺HԲ^?dl1M-O>UW*uosut >llE[Z!|![e;=RC [CpB6_8/D [C6oXX2MbId e{ɖz}d eKdWw!l.[<&[vfll5^Cl@1J'/y4N"[dk-.eCm6V-We}Dd@)#[PTHħZOOiaFn\V=Il~˩x^zٺ [%[RL'ׅ,K^LbɢxRy\_dlOټlo RRU15oAlУqTlB-cUjt%bFW1RtcclŘk9ˇ]Uglljn\Ȧ2E,-Q]Aʾ|ߤmdrVD0Tf{[}A![F/_Rz}Tgȧ_sZ^_hll@uU*ٲ[9e Tb4TM,sxo^^&l)RO{UȖOuD~?e-2 ۳Ի-jʦg+BzB%[vP"Bd3OޫIY--jҢgdc.V)&pޟru&pֲEZ;f ol@ -զO:Sk 4A Ue$b l b l b l b l b l b l b l b l b l b l b l b l b l⼞ls@6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1 @6@ d1<^[^?IENDB`x+DdA.  C rAZext pos 15 May 2010 logistic reg subsumptionb*I7GEVRj*+anb*I7GEVRPNG  IHDRg`k;sRGBgAMA a cHRMz&u0`:pQ<PLTEٟ pHYs\F\FCAtIME7*3tEXtSoftwareTIFF Image Printer 7.0f;Z)IDATxώȕ3wcr 8@7Z{|Gg@7jrޏaװd /%K5 5.mo$ 5zHux|s~W+0D3G:vPZ D`[}vbZt=[)RW\@ئBS+y8ES+Ya.6}Jm{W]\Q_a8*mZ<*RnʑvN6MS3Mm. 4ٛ2nwl;/s͙jn-6!n Cܩ^=O"ئ9Z'#M}o*ok%{8c`FaVئϐjWX+mu߸1}.km(4!ඞ,y6P?tڶCfL~1@[cluHRu?p[ljoʀ`U׷ɀ`U׷Mk - Y~}۲>?p[lj_p[lj~ۖ۬Zz{+fzѷ-Y^m'l[ljۈI.Y~}oY~}ۘKljo<Ʌ6֯o96֯om6֯o6֯o%'ۖFb۱0]4NlsA&νPzц"So,b$G`[-2KeG79 IR+yqy1 zelcmieWJ2s_y˱Y$&73ݶKuܔd[6׻d6o[鶝tHTja]\řn[u>dvhmnw=$`mh۔/F;lsAʰm˻ۊƶ#9 mivlRhԻ)o6dۖfml[nyqnIkmnw=m텑V=]mەmcڶM{.ٮ=-t۔Sv%rPz韙'0>TEm{W]iL6 po)[@d.*7[IvRcAfڶM{TԬW l*TOvֶmt= Pىض4mnDI%ٹO;;9lsAkq|_dCf۩kۥmw={-6F`۾C'Ljv3zlۨ?η ]&v2Lkc@I6Ľmvη ]u/}ԝ m3s~۬ZmaL0m7掲z6ămnw=sghy:?R5B'bg7W왩ڶ/d-sNmnw=sUͷΏ٦Dv˵ڶCMsgjq~e̶et:6ϻĹmzp,6Ώٖʢ6)XaǵUimw=s\rl[1tȞmڎ6MqoFY,c궙'-ˈmG޶skW6ĵmgeed `]ڶzޫk^w%MQJ{V6ĵm:go+YdlvQlsA\VcŒmƶZs+6ĵm! ,L϶*nl۫ یgmFO`)]z׶mJ%,Hvt=)]6ĹmlmSEZ14;6w~ pmESނR 2ҷ)mCMkۚ1J%,s2ݶm?8m2ҷ{)ElضIedf۾`wımf{JX}ۥmѨmn6mkK_S*aY?ե&ٶi`]ضeoەSmۚUJlsA\.Ƿq~AazfS65ئ&Tmm֝eFaoQ'V*{6mknT²h#`f$ݶ8JHeۺuV/.l׃Z݂R ҷMͳMc&Jb]ֶvpR o;aiҶQ*aa?٢Ҷ]/-66ımR*aY緉mCM[Fy>jmۜz]UT²<'Vh\zR۔JX ^ڶΒ{ض(l˻ l[h׃8tIer6RՀmv=Sεm-fs NmkOJ%, }UWQ JX6l{iS JXmw׃8mض4Vmʶ5WM]Y , }UWVKoP*aQM:qcDMfTѮmv=Sۚmɶ]>/JlsA\v-KFqmz[2J%,s:u`׃ΰmillՓ^5w-󋕃6rz5!Ie̶]96斳]ҶfplT¢<2dzoJe'I;wĥm^R 2f۶Yܵll`Bĥm ^R ܈Av'%le6ĥm 9;l),#FoV.V >k6P~6gĥ)˰malܮ}1>ض7Xٮqĥ틺ض$#F>mtFuO:B۶k>lsA:>z1J%,mmen&hYF/&ud%`gĩm{EM%1mo߫kY=GurlsAn百t; -c[]GlڥcLE9x2b]ж,y}}T²_:mB٦oA\۶9)C^Qˈmv=CۮU f~۳hJmE>^!6gġm%eԶmӵMY9 mkcR K2bImy׶Sc[Hvy|lжZUW˗R 2bY٦ڶ&kNdǷ]ۦζzpLT’vA_]mۥHE#ٱ `]ζI"ξ};l)(cF\;ǞmgsXOgmMlhqg[uiC0dԈγkin$ŮqgCF%2bmcwe'f,ۖ۬Zm3>((fm_'bۂ`UW~h+uzWF6/jqS*aAͪC[٫{Tq`UmS`Mi]uf[6Bwpzж)xY^޶K tGZ\aWbC.۬Z{m V]տ}h+۝R ˁmV}٦⑍m[nr<{h++BMN,؂mV}VzM߶Ƕ%6֋VuB{{mpnzsۮT‚`Umۙ쯶vnjJXljӶ{݄%sЦ:+mےR mVv}0ysf۱_bK]v۶JXlj¶N63L~8*5mmKmV:mЯB^]?6&Yljmv?]֯B^n>($fڃm*8`[-Sg࡭ n7}۬Z/o_M6m; <wJ%, Y^޶[mR2nmV|C[po{ö%6vPz$#Γ'|hn7}I۬Z/oUefnmS56V6J%, Yb{ַv9m﷜ "M߶T‚`Uk/ ٦k%Ov3ަm9Uljm =9g^6?zh+m߶T‚`Um{zml'ܦt6hܦ5l[ ljmUU=m7w'~Kۮ۔&Y ljmRUFVwwy߶C%be[9*oSQ*a1ͪugIJ9N'[ӍmoSJ%,Y^6VyT9mS7Y%ΦOm{nCvܦLTb`Umn]%vQT^<=;C'meoTR mV_mJ=9m?<7x- Y^6q |?Bp_,=7Mߦ?S*a1۪ͪ&;\6pN=Oew`[CeS.ߺORԿV|;6Gmߞx6^lSJ9d[o>ɓ9Wm[>][3ZAPXljͶCV3oԝ6=-E4nmQͪ7~߶GoBՓ9),O>M(fڛm޶vmGuF6l+]]sK1Ŷ"[Om;۬Z/n: ջ\8ŭm՟esA'#-Yfۮ}||Cp;t̽+I۬Z/^D~|S #ڶ}bc[p5=Q*a9ͪ*@_Co䤶=A^R KmV}R.N2\}f3:ns&DB`Uk?$w]۔iKUdc9+km3v}C"-T3@ñ+n/gkvɂl=T9u}RZm[QߤOrllmMop IkvMnۥw9IGl:FV/ml;nCHmAvfP]:6c]Ǫ`mz5KTBLt;m]2׶mHlmTBmߜ;S2l[`׃̵xβS*a)ۖwm;]m`d@?rmJX̋ԅ3-ANn_:I$*ten!a}-Af>uöeVucWvem[mKzzʶR K16uodvGl[b׃̴7wQSJ%,6Ujq%v=Lz۝-P*a)&ۦ{3admnDP*a)muᣩvƶEv=LsQ6l[ɶihږz  3mnwCƶwl[ mm&51ǶEv=\6ׯ!I=eK) MqEv=LzYo4`c( m;vmS zضg[opK4P*aeV7&,#m]2϶vMVJ%,+ۊ ۖ lnc۲L0m7Rۖ lnKKmm5"yKڶ/Mnbmm]2϶%$[( 7F"t=zhۖ lnBP;J%,čw9zкd)V_?mmKzyuklT2qW.Wه6sRQn5.ϨȮiۦׇZ=Q*an8޾e}o?mضȮg['0r[Bmƈ;CהmmzYu7kt W$pcm]/rYضöEv=,ۺl#y{F\n v[̮ۨe[wp;ʶ6R 7| vvc̶ۖ lnoF"8e7E~`a[mzYuo VSHJ%,C߈"ԃfKl[b׃̲wm7"9>ҵ8Ue<öw=<ۮ|ᪧlU&Y7u_ö@d ېdcoởo SS{M%C6$mS/R 7"տʶ3mې ^,g%ONh6dtoCf:r JXlw%G:?6dtSˍm3}pi?6dw!Ic[gDk\ja]2p:w!ڶ-gDٻ0OGl{%ۮ'.$iFgl a[sz9ӹLw!ڶ AqjzS65l nν>J5J%,6"o /_꼛hqm/d5 v4Z|`2h#a?~~Z]Ql[m6oۮz>݅$͖OJXm?c@ױt- Fm_=um;ai:uQ$oJXmu(mף9 -I?&lט$խ_Ƕ@HVZ ,Bψo뭞mvJ!IS*zJX$Fv]2pۧO=9S*aϓV@d\ I6Q*anrꌶl[m|Q*a ؖwm;aһD~8lᡐ$-M FVD~8 yRd% ùƗCWvnknmz~lkZm $mjy󃆑zV~ۮࣙl:?hO0o6D~8׋44JX`[eln$x խ_ƶ`H\;Im715VkmKz\CmoJXkʹ~ۮ!Is7J%,u9/v=p@`ƈn:5_l#3/R KpcMؙ:dmKz\\=-7KTo6D|8yכJX#1!ׇm*G¶%v=pk $YxQ*a n qsVbһD|8dkƈӶC:-%ڥpضĮN؃d}R*a F0}W߶mgFpƶw=p.mBmJX16n_mwm%-As~4l[#m:#p*em zmf Nll&aat.uOFuCn¶%v=p!Ic[A%1mo+ӎضN;0$iʶ1tήC?!yu՚^ضĮN;0$il;S*a FPw/C-m >ضN{[1B}fv4YƶYĶev=p!I3Fa%ltm}%v=pÐdk`vz]NضN;}y4ףzJX#*۞m;۶&mۖ bv!6J%,ef%L)mS`Dx8d}==`eӾg۹^%\Z=-AngŶE3BD:Eun,cDx8d}I`̈\շ`"M͒ -AngONve3⤂ GyR3ڐ$-A\Ť>n)cF\T;֏zضĮN$$ymߟVœG6i]OBu|C2bmcᴃ5c-Y~vó#ӑ ,Y~ g!ڶwJ%,Y~ ۚ!IcR mV_¶&u~4OoJXlj5OCƶ۬ZmӐ_`lsY~ ۚjlT{ͪK n? I։mٞR 6֯a۶~1OٞR 6֯`[3rvQ*=flk$[(lj 5C9n=flk[lj 5CJ%۬Zm<רm`UW Iև[|P*=f%lFB+H'uY~|,Fom6/`[3}kSmNJpY~ۚlkoJpY~dk܃mV_zp{,$YJ^1mV_0i,$F`U I֫f{lsY~hHjqljՃۇml;(ljd]*ᰥTsͪuՃ!ڶR 6Vn?]L`RmlsYN߶zp{<$Yۖo(Lljmi3llTsͪuՃۇ1WTsͪumd}ENJpY~۴;!x+Hv YN63mm6f4IQ9f:ymdKF`Um-BmJpYN޶zp"$Y'oHv YN63mlmckͪumdcam6f&X1mm6 6fBm-\mV n[$ʶ_9f:yhV-P*5f:u]Hjljmfp.$YJ(>5f:uNڶŤ mJp YN63mOJp YN6} v!6J%۬Znܶ Iֶ(ljmzp2$Y_\mVSMn[$Krɰ5f:u۴g!/ljmfp;lVpWls YNܶ2$$o;۬Z'nܶ I6?Ols YN6"66ls YN6=ml6mVMnۆ$R mVMnTn56;ͪumېdS*Tcͪu'6ն#_7f:m.6Ťjmwls YN6=mK%)Lljmzp>$Yv~Tcͪu궕SBumVӶMnۇ$k.mVӶMnۇ$[(ljmjp{BH.Pf8?umVӶM nOI^m#y-f:m۔hB59+۬Z'mܞF[ͪuҶㄣ-f:qBtR`UmSSB6`UmS !f-f:id$o۬Z'mܞlm6`UmSۖI5-;܂mVMUlm{T[$᭢y/'9 G nO I6glssֽ`W=c]2ѶbR5m"-:(zI$k."u*T߾t8jp{ZH.p~:?U#13ev{Et8zZHcWt.=^/{p6di!:y<|:?U3ݶS+ަ4wt8Ǐ!ɫm燾jVt7J/nLll;qZ3AhۦFۦ|94_L6 t8:M I6$y)m˻2l[f׃Lm;5$ؖmNn[Qwto9 SG nO I6%mLm˳pY&6 r8jp0$ʄ$Ujۮ40 9 SMIۜ26u3+OImw=Ȕ)&$C>e$o;E`GжmMG7y&`*vۜ2ٶ̛˭TxZz)s|lvpHvar8%lm{6LM-KѩrNڶۺN`׃Lm3e1.mnvڦz\] S'NI6.I"m_~)۔3ضp8l79$Vb[&v2LkcV\UlsA&N89$$om TWOGl[h׃L8zMI6)lOpmc S BW۶Ώ}L0m7RŘm]2p7AH)p6\*S?m_`l[j׃L8 $mpci{A[uض -A&CllM1mnr*tnz6 SlBMlsˍYl5Y^Qz wP M(<ֶ۳,ۦhmSbۂp.NlJ%`[noϮ.mV Bm'lsʭm|t6FzéĘ6]rc v[k۶϶= ?Mlۜrk;CWTa[]b8$$\ rc=~?XFp3ֶ+~{o۱W6˓bdmd{̤:ƈζ}f ے-撾wVע''~׃X%wֶ_1ݬkl;V`]b}8?eR GlsI߈mjږg϶l`}8L&j ٶ@߆`[($ږcKFlS,m6p7QH-mN?R۶mlsAm{νm+HvI߈"GLmM>ou69ֶ lp um;m)`S6ՖM66p? m y/{l;ap.WץD:mqo3Kk=s&FCnl2l[c`׃)ےZmw|="FCnl0hۮ¶MЕ6pU^mK1C#Lշmz>O`]bu8j7GmmsVw0UlsAGQ#F>mtFhI۶m :%E}HvmTrVT%U&`׃XrYu3cĈ]-}̮:@Zv:?:ddUDRE1m]bs8sK%\#Fc!6}g,kWm 9ˊmj1BŵNJVqضԮ9KfץJlsǨmU5`ۦk;-Al' g]=wm;նm]bs8Ǚ\mr0#:۞mE]0iöv=fR Ǎ_-#Fv7B8ڶKaRp*OmgzfسVm zÙUHєJȱcFNֽu9ضԮ8Y%1#rV߶,-A,'[, ycFBKYv2aRpiJ%`;ƌcAlYpmz9-v"yF d+_ lsA2c`CsaVF,mlj^n ͪm_E.p6͊z۟=l6-YfuMكR 餮6־l"W=N4v8l6\mVl_֙.~m ۚ۬Z/oVj7wsbza_k n6־lSZ]vgm75W ljͶmv-{wPO)@3ͪujT6ܜmVul+ɜd Y^޶kKfۿ:菰9f:m7Gm+D6+C+ͪublk\mVCvs3ĶWdGg=fmV?vY϶~|?Wd'@[,[/o^ՙKlۦE[\l Jll6v̓7zĭؖtǶP,I]mVzdo^Wf9 lHqY^޶vT 5^+ͪҶ]ЕDNYN{]nmo;9lFʍ#ͪtOvVuIG?l]6}z칭pb[[*mVͬ5yV;Ғ施mos1}P`Uk/3q|pU;-36bpS*mVئ+5<6ݷ"6y llmxIޖKo:TǞ܌Gckͪ򶽕mlWU7'f6'R 8۬Z/oۦ,[ۍYvዱ675fzy۶w(AM+Pmm#.ٵ0M?\6}jR7KE;L%ܜz7cmnIvN6DTBuָQzqeuUm:>ڷ99gW۶N6zVjmS/< J84U̖<3bߖ)>ew+NRv|%o^m?=-LBm-FE:A65W-2m`[Ce:S[.ߺQgm?O2@$m-V/>/agmUľ-m/Z Nֶ|gmY~ٺ͑$Zږod)m/we[m/(ZN|ųo٪im# E2lqko߳꿹g^ lkQ ,m^lkmvU7MI6Zٶ,+ۮ;G[`[C۲m+0ضڶLo٪ڶuE <2~ʶk-BXMf"= O&ؖ}dҮm9h6.o:tm™mͬ@,`BDVNSl Fkmmy? m dqӝ'4жm?;Z*Z㶛'gm&m)[7|3Og]Iy MaIۺ~^l;mlsm0 PJ%ij%;LOhܙlgAW%uF}5652f~E6U|oIpҌ*vld">ttVr՟+l뒔ٿ?qM i7ZF-x_7y%mͭ2`l2p[nm˝L/'ꆍmIijtMwn[+s۵T!(ئQ}jm ݪe_uڶqI i}^*vEk۬Vr;I޸$4!(tgz,@O\M+$iҎROLbzLJyhl;ckmVP!Ic[^$u^wav{Z*$ uuS֬pS{|8'߶up̮7ǶP;+KdO6;FKmE%ٻϵT&ds7.m;wmR `'s|wJv{qܼ&oư6Oβ&!%6O9)Ls"?&3b\K%l xr8#V^lp믶m xv8N+^ l xv8ۺIHyM]zq-m xv89fv"%6um$&kl{M@HxM0 &mng@(MlͶk;*& mܼ&m$?m}m `׍}r`o]K%`+mG^lobGm `'6l^l϶TIyM@~m/ g[[Rn^l/`r`6l{#6lk6lˎ[o;6m+6lJ7v KmB$ g5y/ؖ>& m$6l;+ g۵T¿>& mmmؖ Imm&6!lIHlĶ ےZ*aY6al &ٓ:& mlKlƶ)7m|v-mɃmض%&uMO:-uM@ HIlƶ  :|-@M`0}`[`]K%r:& m& d)7mv-mmF[` ]*>w 6^mJ(-qM@۪f;;`]K%$ d۞5'&qMW:HHlƶ# j[0 ImٶŶ6al˷$$ j[T% c[! qM߹R ؖ8& m$ c R}_K%r8& mglKl׶kRnmmFM`u `[`0U%&mM_:-mM@ {Rnt6~m&!&mM@ ێIIm)@M`@3IImBٶ!&mM|TkJ f)7im<]K%6& m'lKlʶwxS*ے퍔6m&6l^ IImBٖa[`϶uJ%r4& m{lKlٶNRn`6lۑr4&[S*ےmKM`Pؖ6&m$$ fےx[*6l6l;}r4&WTBN@`r2&m$I mF[`PU/5RS*6lvL@`NalKlζ=I)mv6mJ eM@0ێؖ4&mR $$ 6mٖoIIl۶nlKl̶bCM`߶uK%`[``Up$ 6m J &aM@0ؖ4&mR $ fے6m&)c+M@0ԞIIlݶnvƆ (& m)7mڶ+{W`u r.& mI m-mmV=.& mlKlݶn`6m&! ]M@8Nؖ2& m$ ߻tJ%`[`r0&@ o[T)7ɂmږa[`uK%r,& m{lKlжꖒd6m&d6!mۑr.&mR ؖ,& m[Rn<*&۽[d6!mTgc}l߶nRn5MRr*&mR $ 6m໅IHHlҶ7lKl6jHHlnlKl6oR[RmOM`-]M@ۺHIlҶ# nJ &UMl6uKIM`uK%`[`ND6lJD6!mS)7m D`J%r(& m$ 1s[*D6!mS/D@`e%)7mB-LšRmBڦvNM`onMurؖ&& m;R[*$46amb[`J%&& mlKl¶^Rn`6am &UM@z-MM@XIIl¶^lKlֶ7RnR`6A$ܤ kے[*46mc[`/HIlĶ^a&!LM@`v$ b[TBN@`mIIljwM}RIm}7_ٮ¶^Rn4bֳhwu1_myoIImƶ"r)/w_Luk+mImeG2&$D6Mcȥ0_V}L-UMؖڶ o^4mWO9 5am;c[`\wqem۾<]' b[0 IImvK3'ɻY ہm46Mk۶m;e"/1v@`O+O_ϕmF J6Mc[mon1ڦ &EMSyl:_6LK c[0 i\h6zWlۓr&ئ:+Z꿛v`ؖ"ئmSZ]mR ܤi۶mQilyJ lR4)+@M`ulS?:clM}AM`fĶYaR ؖ"ئMM z5]gֶ]exK:R|\m;n㰭_e:#fgl'$Զ]l;쪛vh f[T)7m-lDf^>alkyh[6ōmɁm-m+#W$G]"mWb[`FO<٦WMj1/@@r`0|l''F۶%%)Lde~lӃ$$)㵭_*ےzfڎ6:RuҶ~alKlS45m2ܺ#6u`$)L=]'w}eiìm'lKlg|Kұ#T)7Ɂm,mkEhےTU'iEZ걀Hl&!&9CIJVϭmq%)LmɁmM)*ԛmrSծ⨹uS6! Vv]\*K"'yS*ے4GS5^Qy;/E)7meGۡ).m)aՍR ؖئQVF&M}Rrئ9efMRPHGX0 PDR533``K%rfo 4/E nwlӣ9Im:gm%kV)՝ض-IIlC?^~wlJ(HH lmܤ f[TkJ o+%&5M@gJR{K%`[j`Hl# 5M@0I-5M@xe# 1M@mmvS*$6ئ0 )7mVb[`lۗ$ 7_*`ۮ$&9M@8-1M@,r& ۶%% g[T)7mbmSb[r`po~Rn` m6l)p;mbM_7RnTqH@ 6HI lmUTZ`p&!&-M@ c -M@8nJ%`[Z`(lۗ$ 6{mimN&! -M@JlK l=/@@Z`(l#K761froHI lж~lK lm$ hMalK lmzB2)7ImbmmvS*61vƶ6FMR`&61f^)7ImvS*ے`9Rna^8R0 )7ImdpS6!HI lm&5Ra' iMR`6!m)@@J`(l3ؖ& ۴g$ iMlK lmʑ& m7I-%M@+GMJ`8lӯ R|ߔJ &%M@6!m)@MJ`l& ORnT% Ķ6AǸnJ%`[B`l#&!M@$m՟ؖ& m7HIlm$t6Am)@MB`HlOlgd)7 mo \܊~6f.]6Am)PF) 6T6qVcmmv[1$M@$ȱ-M@PnK%IRb^Y`[2`Hl3󓋌$moK%-M@lC؃m"O>c[2`n7*vA6f'c[:`ݔJl=Db1y۠`l6A6am)P|T6f')7mvS*!mm"\<#&M@Tؖ& m7I)7m"ͼ߰-M@XnJ%2RnRDb9lKlm{GFM*` \o߰-M@,շ؂m/pξ~ŶD6ئfLU߶ z@` ƶJmG{Nocx; M@L_|ax;M@7w0ILlmjd}F$ M@4}D6mJ8f_~%L& uz[F@"`^aC0I`XlS`/M@4$m+PF!6ؖ&Il+`\0I `l a6ضmO$)m+`F.IRcۮ=$$ m$ ж0I`xl6_$ +PJ$ mmsY胂qM@7vTB a6v}X#L& F$ m[0I2  Ƕ#L?& { m[TBa6}mIW}Ic[HD N!ɑcW*0I`xlkgImTa6qF$zM@w]K%ImTB)Nc۵TBI$~M@hK%&lmR %a6l#L9& mCJ$уmԳ0I` J L=& n.a6W/0I` &)  ǶG$nM@pۆK%&vqjPDl5 L5 np0I6v6 B۶_MpJ L9+m+ȶ^0Iܬж۶!L3km;A0IܬѶ/9bm}ܶI*΄I"fmflJ( D*mÿNC&L1m2mJ%Ifm潚l~I"f|m{P$L5mdD_?$R6^M$%aYmTBI$fVjۼl;$V6s'mK%&VJFmaxYm[o燅IJ$>Ja0IĬ϶dT0IжvMdm$Bٷm{\*A5ږϽ6$Fso%TBI$^h[=&_&*6wH$VVi[1V2&J%&Uvy+޶'J$ѲJ!t1v@$Nirl{\$L-봭rn2*. aHYm16t DJm3Cn[&#I0I63n2&_a8YmY0II$VVj[>+(mCC$NVj[6k6IL  a8Ymm{V*$L)%Ee}0IҶzdʶ=+P&UV4e6$QJd}a0IѶs#4&:v Dm3m_^Ķ'L#kkcʶuJ% NTXm`dʶb0Iжc=k* O%L볭l۪;ʭl)L2xOE6ۦƶ{jL؄>& &t~ۣűV&9fQ=ԶWmu soZaas%G5T/{;xbM%;LbV| жfˎ¶TB5I )f6CDe'Ydà ڵձmjk2LYშl3pJ$jׯ1+Mń-m$&2x0ɔMm6pHĮ^/_b:dlɨlCmÃ/&)y|gVCVԶ囌¶uc  <ܹ۷g[^۶o2.MA <>ݹ?/ l;ֶm囌¶(JɃ#/&yZ>4K}Sl;my?<p`oӻHW,*m21x7GL*~0[rWmb}zni^~;R i[`p%$C#YU` ٶ+C#/&,y|ݝ/_Jp[mḀ],et5W;w/&(y\9[~ۦmp$w%B#Tm1%Զˍmŝm6mx64p|0dHehVg<6mm;x0Iw2hh:JYmR'֟m&)Zw7`S:5< _l3uQ a";lжwI;No~C @$ן!]h Z;X/U~g۹]?8ߠm#/&ɿdlvh[S#olPwm; ^ Lr!mNmWzҳiVu ]MCoضvVͰMڧmgތ߶s$7mvjV}|d|ٳ?KQv?~uh*%LmmudUۤlH?Lr֦]0m/AZ~ @$=:w#IYFn}֬mwɇI.c3q5[l3m/i#=naøN>LrSۦƶ! @a8]WO/fyѵ17øN=LrVyDѫy}n~~TٶR 5ELu*y\`֋n_m;v׊ݶ$'Ur'l\M:OۚZ6]s3zu &^<S,Z׶ۘmN++m PΫn9 N;LrT9EDIbnM_7g0AjnZ6X*=^otdYљ؇~-kz%9N1Y#IIkQl+uxkۡ !Kr;!N>II̶lڱmww& knh[;0,g%$l3;h><]&vB{9`^1wȵ $&)m"Vᵮm¬mc';p$i[D)7볭xY*mmVvcoj%b[w 0am$϶d= 'YwB_8TU ۺ# I/l(YT]ٹ8lSF,F;l}̋IFr>N7&'M/aWWŒ;vd$0#JYmzs;z~VmK%t s?&α\5VNV{ Xv3l8wVqD)7+ ]J _myrFm \ 7$ж2׺[V񫹒mtg[_vmK0I{m/h<ɐxVط/?+S?_j,=Kޮ)ڗh4mZk]tŝmֶ?$du 0I>K' `/jzw~Muwm vN4LrmZko$v轳ci @L˽`Y͙ƶIWJJ8ElMvN-LhhVh޶K.ێCu53@ v-b\QjmI)L$lhgaȶSv̚RUQMz ĊTR Mim!cۡczѼc'^>$ԼpWl^J)'E'됤-\w{Ϣm*Mimuձ5+ovM8e6Fe Iܤ*bIXmafd=(;dm+-wmSZ/o$m\R 8J%UĒ^f\xްS$O(L(TMcXm49otm;'&qJ% `yf]*΃TRMims(aN^؅>\XRnVkۜMc}uJv\a۔nm[)_mmLWz[ܬζSY]Y)֮fJċ*ܬжoֵڕ}f;^*TKZm;ֶf5Hq&fm-eSA+m$-my=m"y.TIjm;Ĭi3`F'qK%UDJێMow1nkaۄ֮m{{DJsxyUu;C$I8h &d@l |%9L%UDJMt^O+=cܺMhڶlնw}3owa*"$6c&;,ӣ?;rVmZcC9f|wa*"ڶ)jͶٶlv6ωȶɥse퀛TRE$)7m۹9h\*"IYmzޜtl$ogh\Wx{2Vhۛ伬ڶcd?~~F;8$K msˬTʈޣ2rqo^IiqܬѶ){/^6Y}~~+6Tq؆~MVBnN3 JFV=)Wm*"l{T^[mnSIq$ҶGeT[1I!Ilڤئ$w168hlJ(#:}]YloҶl{ItTREIRI_[9SI&ͺcJl;d~M3,p$ `}m/QU|gѮ 춺[MM/?mkjD8TGl;նm䛌 :[ٗi;KZRy7QQ\ǒ]}|1\׻5V$VxW_ ש(B+ӋO囌=*=6,,v(RnVh[˼dLJ%J7=ʠTRE)7+SInslӳ=ʠ"OQܬжr&KΉƽmzo1%ݧ*Я\mP"'&dJ&-05(RnVjy0/a.4m$ԶyJ(Md>m+Z䘰ͺ5=Dl9]`lܧ*HXm%lSa"6f&vi.&ۄ*_Tnz,J7C_-(حm )7&%l;g_d[lAERI1ܬѶٯ&7 ۲ӷ?lm9jJ.V(l"lB.EmbH6Q&M._jJ.7"m1rj J.RNCl;c!۷ÿVSrdf#- Jnfֱu2VgĶ| nrrm?I\BVg[ZsGWLu; ٶvnີNn[@kTB&ͬ϶ϗ-Z^չ-Jj^re} ^öf-mɌRngip+vim;ӹ-J f}9H~۲N{Jܰh&y)7볭dZY4ۢUR"HYmfzmg!mTRM藇ml;Fӹ-HO61Cdngrmx[3A1lh*"| 6}[}s[4T>d&@s[|M"xlsALw)≋/m>Dev`nEi'`l}nfEmTRljm_~n߾un˦*'`WL6;ƅܖM%U`]kl{xlCYS:eSI5M@Tml:7Eͪ5=~umv9kKmTRv71lmu6z6UܖN%U`Ukl{x9 p=,JrmMm)Lt*mbDi8Iow|*ًǗ4 }Ѻmn?s|*۬ZS᫻.V?3Nҏ zܖO%ՄN6q63Nr3s[>Լ(/{y ȶ.3$3{ܼ$'`Hm'/^;7O6fJIn|vn>RIM@͌d|*ͦ5Y1/Nr;Uc#Tlm$s;7)7& ZInOCu;Mlm$s/<&x ֶyq!|~<+*B`xm'UܼN66+NrԹyI%UN66'Nr?s4o+| ضYqssJ6f͜8K'T8 lۜ8@ދO*,Zc=3$:7OI& jfI";7O lh.I1%jfI_%OCM@2I$S`6z;{^^:wb< \N6ϩI& DqgFlJwrͭIyLo mұM'LWGyN%ջcڶo2!q|oJ?dzm7m87yN%U`h%l7m8HY;T4fL6AKVωmB6z fl2%qOSV-ۙrZl'yo,|;T~ȋZ636m$σQߩtgjmdRM<luxy[Snj۬`XZM<6}|Ujm$琖m$OElSm%}W_ܴslm$Om6^*"d H̶qs!RI^vZmmricPSTR}2!j665N4(yξv=+@*"d HͶqAɺon7gZUbxkl8ɳm_%{AH%"` HζqgAɦo+KH%U`Xkl8ɳdaoͺ JmҳmbIPةK2^2@*Zc\~<`o'4gWTR`m'yhMe \BASnM@M ٙnU:+D*9 `m'`G|WTREȔlm$C*ٽm|WTR vny֫IѶiq1M~/fV4f?9$m'a~϶K IMYmܗm$w.}|& JBNa$mSS?6/.VܪN JNmS+0)N JOkO-&pI%;XB1i6)N JH^9{E&pI%/mjkM=f='j۔8Ig9od3?>  >.ƶJr.lLې`WvliҶm%j۔8I#9g?W~/(T:\l;VF)q`dӷm<^2P*>Xl{ڥmEm^Tm'A^0L JwnyYmuf' F4n u$lۆzmr # ᷿ J}j+Zۚ8ێكG.Ci¥*%ζy l['yl[fѳ c=+\*۞vk۾T{7яl#I¥36vumum`TơNvWTR=oԶwɵVIm#G$&6fXmk݌`H ڰVSICۚ sҶIl#G2O J_Xڶϙ͐m:NbʋȑA2KL%GmO[cKtt50|y~/8$0~GI'i#豌%V_&ЎWkۊ˯`dhPrO J_&ЎWhFrۊ̌ #FOԞM J<&lyhP??6T=kMsr `0a??6T, `8>ݾmFX=T{<+hu; =m>ty˰͊?dy[Vh~r/a[۷=9#A^2p*>l{֚*@97mAg^2p*"XlZI64}$c'pN%U`nmXs6ό%Ļb2\l+ns J> \6x*9}ҶJy 0SmTRElYm#Yc`MTRE$sm| :m o%8lkMmV}hPr/yJc3{"MKf6u{%Q$ܬ϶&d'd omcco=|*"4e}q^¶z JǶНO%UJ6/a]6f֝O%UJXmMi| ۬ƃ=]'p]Ik%lFV7%ˢIKXm:mVmQGO^"|*>"l{کmۯa]69ͽy槜3B%жװm,(iwgXKMs=kԶbnH5l|z Jn-vVO:})@O+47$*Fbޤ`ۓNm ʶѠՙ4EH2X m9dz&Ƃɡ͜mRnhilv"$,f.D*F;]fg$P*m˪l JZSrmVehPV. rmV)`5c1~n8K17R+m4(i}6X&nK6_8QHehProo`ۓ!l+̼~Xm#AIKlHPҾyTwJ6̀m涧?.'F}]eն%W)P܂m[ިwHg9#E?WbSw6|s8$L mSfo>;M@xvŶr]m4Ͼ~ݥ rB!َm'e3|m7}VfxProcD naRnVh= v> ɨ~Z]|vF;Vr>.ƶKwԪU$hͶmBiL naRngiǶQVZ%޺5IM8gdǶzڈIvs6:/Kf%ano3?Ķg%NHL%$϶mag54݌άͶng۱mujA'V3K(eX&Ll;<FG8 )7+rk_ضrtuBi<`m;hK\ROv;g}%#eO5ڶ/ۧK~^4cAIۄxrB:;84p46]6=c}MX ܬжwɾmmUimS5m,(i@,IYmUwwwVq{BFnW( )7f#AZ66~X8x\mw/'rhmzh mUL.($Y:Oޱ"(9 B۶c߶*Z^oY>0^ am]l;T"V`%p '鳐Z: ږg*#FFG8l{ziNw̬rn56rhBi,H6C&vQT=Okm4(-l{zi*w:Ny轙hXPzXm-ڶ9N.!N^"ݭ ٦JٶO$Xc6L.4D {6u<;wm+>YٽQZmAnHYm9mo*m JP"f]|wm7Yj*((VixPrl{}6ymb%SK( rB VΔLmV6PHl6hP2[Wqu]Lq'CꟚ?_12TXP2DmުxfƃQ٦쪮mض|ǶlmxP2Rl{ڱmǶV%qml_JjsAkk6*̭+ofNZBi$vmv~4S63mѠdZsAmSgm՟ئ J&P `}mϰ͆.!Nvi[Q۶Lkri>|gm,`BC~% 1L*L f}jz?kjnul4+l3% f}63SOr׵M)Nf%AɴJܬ϶rK[+yȶꞒxPr' fʸN*NAm c)li%HYmǻjQ}ۊLcئJnBW}+ IQvҵmnm.IŸ[wBzґ}9]V[6کm yvn;L⯾b4(ԙBG쭎mΩl PmívԶgW]t8[J'ѶE=&`.@#[盵( PmívϩN vo_XP20$l3tlpP_mc)lI%bpmkn;u(EO*Va, TBl3ԝWǶK֬pS;]SzKe<(N`pm;h0#XmcaJ`6Lhl^FfrmVlX$R)7&`Ŷ%Vۉ)7&`Ŷ%SJ(xdXm7bƃD ^¶[6L*{jmx+MOllmsjw6h%RIm"TPd>TmfT߷Ik6*(ܥPmv6,-RI+aſg' RllܶۤS,PBv6)yJmkI)ٳcB Sni[mbOgqDŽJܬӶ \Sl-[SnVi fe䥓P=f6cߞ>!ڂmCۖa ??; %>5v¶9yz)$P;fmEN!RI+Mmb1tnaPV6)Ƿϧ7$NXmFrN2E8tJmK9uۖ}t-RI볭09UۦٞPmC]VHnflrt-RIm?$3w ۆZ;m)H&wl;̿\m{5WI(Ŷl;79dzrƞI(r6ygy/3R)7kmsoK(Z;69jy-+-ǶٌΓL(sl˰m6ۯ0H2 MLa]m>cmeB Sne[lm38eȝb2 VifpKH%ZĶ\_c3[h~{פPmڶǶY>FQN%s m{/mǷDzTJmS۶6||%RI+mm(,+I(Ŷ.mS/pXJd甛ٶ-mgB$zNYmol%{ 9PmڦXm͛JB.xuѷq܌N"AI)7m[bLhϕHBߔFLr66}$2wolcm6Ů}$P7 `mm6HB)ݷfddFOA" ~Vh99?Pmo 9i$MXmng<-loM]FB$UFͭ6cFB)ݷzklOJ#ol;bX^ vڙm'lsԭDJl+ S f[mNHBהv6'A ^Snf[s+sSH(Ŷnm;c|ln%Ps~[Il,Rlmضcc[ >SnVicBBϔUڦo%mS3f6TJ}ܬӶ fV >SniĶyXJ"S+LJ!gJm`\lnP3 `Glԭnamk6[)$LXma[i9u+Rlm޶m3-RI*A^B)ݶ^ fV )Vk[7mM$[vz ~oV0H(r^fm 'J5EԭJ=`lSNC SnM)n%P1`f nyL6ئBǔlm;ۢO(3h"& 'UEP1`.}B$lm['bMkl`^s<&`lXbO(Ŷ#vSO(6}_ ˩['b[5EԭJ%`l3XN݊=-N,bO(vm1[kmqb9u+RoI& GlFPm'SbO(rmrV kmqb=0lm5S"-DԭJ`l9JwBlm5S"O(rmˊ;s.&ѢN(rmᰱŨnyK61sb]1RlƶH;G ۩['b[5EHU M5ΣN(Ŷnkl:@mzV mmb=u4mC ^Muض }`[bƲjl3L% `[˩[ƶ,ι[ֶƶUv3L% `[˩[mysmm1Oz xI6~D9ι[ֶƶgK U ($^RnMF3L%`[*Ea/)7&L" `[ۢfDKa&X/XzG ۪[5qI|$`lc`2 ۚ5ӦnE&mϴ[1 ۚ5SzG&mb|dԭ27lkZc[Ԝ'eG >Sg=b `[fZa)7&nzgcdlkZc[6~?0lm7LUI|`lv (+a[fԭ8$Jclaԭ8+yH6vԩ[QI&fI6vU1yH6vԩ[QI<$`lar<7l[c[L10L! `-G#Lqö5Eԩ[1I<$`leԭ &c["gԭ$M2uVnmSL>I<`(`nE&Y> `EX0LklS ,mS+m5[IO6v[U$fZc[L_%M#-0 ? $*-n? {O݊/L&9N.mlmL_dlmnE6?) lٛ$ S O6lIpRbKqrmԭ$ |0y 2ئ[c[Hn&mmL6G2XYd~a$S\a}m^3 [%I,[Џ6dVdaѷYޗԮF=oqϨ" [9Ķ46 -mԭ$mo-oJ6|Ul@2u+)7\mil;fڶ"[dv }_>?mԭRf6K,im;g36 Ydv }]OmMf6(WlvthrkNLmۻ66'0h̰r,~mS"/66,ɳSnxnm*򸻝@mCHnE&@6݃y`Tr[SdmYRc{ݳ*`[qIf'з~POdl,VM26,~T7D`j7O6hV\f'зMM l!WdmYRǶӀm76lV\)nsC6mlSOMF*L26,~ѱv3[E67冾ʶK}5mU ̶6_2gY?y'=n. "AUܔ)]mM݊*L276<[QI0fe^ί63'mlA6u+sSnlS}[-s6lVTaa|=~bJ56{j:b>ݭ` I9|lkQzj[ nU80ɜOlk+0-z-`[Eq!0ɌӸKm-svHnœ6㭖z֛Om(90ɌE>lkiloqwMl{tVD)nqv*RYkCknS0Yl"m-m;nԭxRIcfs m-^aCjGm`8 ff5xMaIƶm3zTv0+ԭh$",Ka> V~Md}I-[ƋiMR^O<56+lQ~;IDd65;RUg'~Y~leFeH &KlabAS>^RcW)m%eFH$lMS}6"۪$b5YԷ6UG-[Sb),I(n6!ˌm^leݢ-&06ml9zA ݕ5_S+ٖSVGrc meG}6}TU7/VR Hi9m5qSjތ`ek+JY1cV$aAʍMq; Mq9gla[)n˱mlkNCv47zڤާ=fԭ8$eñP*VwuHR?5mP$)n0dlǶM=fԭH &ON?jl{"VL݊#L26vWm:9S(<5 zlD`cL݊#L26mlm3u~LM6Y71MmJZ '̙Edbʍ6'̙E%ُmcl{œ[QI&LɁ69SHqr3eD`L݊"L2-f6'X­"mml1n&n4`fM݊!ml c&۞0kVa)/rat nEQ0yJmlmϘw/AdmӖ 613kV $L;\lmϘ5u+0,n6YSgZ IƶM=#&~6`l{Ƽ[I&$L,mfԭR&f_NmnE&O6m"yS7{ۦQ6yS"اL6nϫPlۦmOwQ:m"mSI.ٓm6QO Z0=DžmmSIwmlm8%k'ǶM`VJ@JǶMrrpBI6 QL Z 9w&Fq1u+dmjMbVa&66xb0I6m5&q0u+d[p{Θ; qf/&yhlᶦ1Mq1u+`[(m6q\L &6k q\L X09AY`lAac6ܭۙ׃mq\L X0eq6qL &9döͻ66[gq~8|alԭp nvlm8$X$uhϳƶMYbVKշ 6 6 L &d߾ 1x 6 L v!3fm:fA(Lrξ =6f[RNU6pg3m:f[$mqMlm8`r (n6fM|#LȾ~Umclԭ@lm$ilm6&L}W&lp) 8fCnsۙm6L T ˗n3ǶMfVUw4^;]>yΎblI ulm7ojMfVy~6ӵEf`lԭGM۾Vs66Lj]s6` n5 z`&lp4uն-=+ngm6M23Cvt66+U`ۥ 6+'d_^ `Oy¶vf6lԭo>6f[#eE 6+M*O59lD`3֒clsۙWm6+\M:NcƶMYf 7m-vR'clԭ}Jr&۬p5u+smnq;Jm:f[sblD`uf3ۦmv8opSD`n|n+ 6y%6lԭ}yps2m"GSmlmv8o3nNƶMhV66y%6lԭ}Yk;6f[Aԃ66;.n&۪7c&,qd[&,q3u+pߦl ۦmv,DfK&,9:y ܷ,6f[T-alnos*9 `l-3(P!al̩['ݱY,Kgֹ66S*]ּ~s@:w&&0uЄG|o7}~pflM`ԭ\i՟&<#Lrz8mmne)7y3G@ݷ 7lm5ud_&<# ?U 5 `fM:Hz\j#>M{p669S.ys$amfչ669S&&uk8FRJ(F6q;lm>uV&Y}{̞alem߃I) ۤʱMMc[3(HPƞ?IܰMMcԭ1&)} `4&N*iۜt+^>alq̏Ó"iI:f iL8FRv,[n|ZOalƴ%ܞHJaJqdʱMMcԭi%qdʱMMcԭۼl42m7lmӘ4u}x$-l=alƔ:@Oc$4*ja iLMsmm*6S2Ǡ%L2N}p66[1[-WL}p66:@1c0yrq OݲKUK^#r O*c$Ri2Z{ClmSe#)05(( N:eSnp$cO66ncB\( `$ʱMMfd֤I̢hʩʱMMfd֤IHdzȔ66[o˶޴66|pǵZk![pZ;lmy>u8uaھalө[Sc$$sDlmy:ukj\ `B֛6 ɜSc$aQmm2Ϧn-8mAqȚ)n&ۦ<9F6 @WLyp66=Ų]aIe[{Lyp66S1yD/݃bl[I: `*Y5/z6u$1&;8N>Ffj_?xm.n,80IJ=&6Muggr:Mm:nb$4yZcl49ShVa_خ5L2-,Xiڶz;6GSd1mDW׹6Mc~О]m.n c$45&)۲lLjoctLݒHkd9厰MQvgR}M;R#)]Im=#6MmZx7K`H#)]̉#<l[꯭91O޵i63sHʶ~p6Mkۦh˛N.m%0n֑͋+lܰMS_sXlv6n Hbv9e[{Xչ6MWaۊmyݤ`-1&nMvKئ"l`#zS27fmc3Jm/,۵o׶lq鼱H4;fvCЖ2`;:90p|{^֛C?l3;8s3-M%I6m3_9%Cٌ d~󇮹mN8lJ@M&K[G].El3;ԃ=o@Uܤ`G3yQIdi]aفC~z?68mV29!+,3oĹe[b[YjUf^c^&db$09MY޽Y`> Fޱ Eݥ%/zLi|]MRm_mzlq-s#)-00n؛6uKvvaWFne[MĹM/8Ĝ0۫^׳mb> z5}M}vV^AD\mI9ZB?=뷱MͶzƜֶ+}baM?RiNBZc:Pٶi*] 6{uH9atgςئR&Ÿ,VM4FRΨlkjglBٶkj%<6nc$41xm&Ÿk+^"GlkPö]R!;QY]aYI֒hwA68eV{8+͊mo$&~G8FRE],6p,ٗ'c[Kc[oRwj7>HJqyӾZ*۝m;nZM]G,7[vVmlkyhqm.8|@hMm˧l{~O;olk]ܐ_ &Da7˶ lYm-yضîPŷZM%B4%::w{ȦIeknY0 CVS]dMOKp^}{m,E+^ڶfls%qڧٍo/ׅ;2Jvi "L 8^1,T:%6R"۟[ۨK]+y&,76Bu*mzS\ϣ5٤8^A==<lu+mp)s''9|5L2^(tzۓ7ldҺ@Yk[sj%;XiB"e[25v4Gtvil&3Vp\md< `y`EF& "8/I#H0mS=JrЧJ vϚm.mu%Fh`?J$vM~@ۄ, l$i m-gRnPBǫmE҉n=\{dZ2mv>6۷.YM>w_6!f"XWKXlksfFo3ar NsLL8hzlRvӧ,VUmض ,%,gVBaXV+MEHۧ=MX4?~6lk_Zl&b`S-a(&bĦZre[6#j Gllm1bS-aV`lj a&ۢd< `tV+MExlmQ2^-aV`l4lmq2he[6'calmq2V-!pD`[,:m"-NF$Km`[<:m"-R,Zlm4LI8-i[%i$gܶ }<-& }>`l: ۢcl:Lb6l` u.M%9{ϳV&6ؖYv|lmip̲%`[T;yہm- NYO `lKK튘. 1 alKc?bp6ؖ1 alKi `["i `["dטܰM%37lmipQȘܰM@Dp>?p6ؖ_~GD)n&Ҡ `[mY4)&$u)Mc`lK=ޖţ DsIjbRm-<Ɉt6ؖ&⚅…m- vF9 D8չ۱m-M]yx |<6lKK[sB7lmp'x8&Rp|@7lm]mpMBofipsMBKhݰMڤn0m-η!alKsM%TsM%  d8n~l2 dȷC?  d(vC? m-u`lK0)&'elm/9n&^)~]bt%xp-Glmvbl{Q^$K`H"I&^sϋLZ6Bu;-2`+1P,2 `K1/2 `k1p\d ~7K6jܕ-~Ā &%Ā ףpv6RNO al{Esbm%薿mEi(WmܰM*ϟp64_p6s\` ` slFp69ʟK al{i_6߻^"{?c]m/I]a6^3mAv=8~@Lx[]mG\^$Czl{u￞A!-ȮWJrzl{uKd׃`۫sm6lm$zl{y zl{yoA9fK,Nmɱ-ȖX `sʖX `sƶ^K[Iz drzl{}[lA9[6FMlAm Am ԓAm P5p6߻e6ضnal[lAmmmmm`]m-lm-lm|zl['Jbl[nal[{׃`*p?Q`*p?u `*6߻ց&ց[&y ضO6ضO6ضu|$ u| umw=%MSMlAm%(mm%mm%6nmm%(mm%m_{wi\a&+{x /a踃^I+h2uބ:?/% _T^?6,t]Q]mڈ <쨭tԶת˩r$qn'/IZ(Im^+nwrO*o'[*BK;Tۍ۟m-l}[OpԶǰK?ar\JRj[mr+QnQ7n"OqDQےS㋴۾~$o+2 %-\ E?j g[ħ'k;ŷ$]6j[ru=$t`cj+2 %-^ep\%98xw 8,ݢ6j+)EmYf$P[qYEmV"9P[qY(ImVenQ-j3P`QVEnQg[f@m6ԣ<,ݢ6j+"K̀CmGQ[yY(ImVEnQS,P @K,ݢ6j+=QV%JRh[f@mO= BIj3-t @GQ[X(ImV%nQڞ?(j+ %̀J-j3QӧEm%Z`$P[XEmV[f@m%Z`$P[XEmVf D ,6j+K̀J-j3QӧEmEt HnQڞ=(j+f@mEt HGQ[/6j+[f@mEgO=ڊ4$P[/ݢ6j+={QV %̀4{P[S2^(ImVK̀DmOziBIj32^EmV LJRi-j32].ݢ6j+=wQVK̀ 5wP[SB]EmVK̀ EmϝzjBIj3B]EmV(j{ԣPsJRj-j3RQSEmP T3nQڞ:(j+̅f@mt TԩGQ[f.6j+̥[f@mN=J5s$P[-ݢ6j+=sQVy %̀ڊ5oP[nQkBIj3b[EmVfֽ XJRk-j3b[EmV,j{ԣ\nQk-j3rQEm嚵P \nQڞ8(j+׬f@m嚵t \ĩGQ[f-6j+׬[f@m墶'N= 6g$P[,ݢ6j+=oQV9 %̀ 6gP[ySY(ImV9K̀ FmϛzlBIj3YEmV댥[f@m%M=J6cP[y(P[f,6j+ٌ[f@m%t dGQ[f,ݢ6j+ٌ[f@m%M=J6c$P[K̀ڊFmϚz;P h[f@mEgM=ڊf_(ImV4-j3Q۳EmE/6j+}P[ѨYSJRͼt lGQ[ %̀f^EmV6j{ԣl慒f@me3/ݢ6j+[cKQͼP l[f@meXnQͼP p֥[f@m.ݢ6j+=gQV8-j3YnQڞ3(j+u$P[K̀ GmϙzκP p֥[f@mL=Jg\(ImV:-j3QSEm3.6j+qP[)SJRθt tGQ[ %̀Jg\EmV:j{ԣxf@mų-ݢ6j+ޡ6j+m$P[lK̀ڊw2uCmVh.ھ}ڐ-0_XU۔ tv[m~$߾@mH-j;;ѰvPᑬ&~ȏ>| ta?侶]?edj#Ԗ=ߏd叴QUݩ6o_6JR'mjׄ}RZEmZk\mn'u?W =j˞=}t{F>Hn)ejԆ$k$kߦNk;U 7R[.mI6ꚫW a?lkrzzm6o/٧}m_| kEmݱVUË kEmqn}mW`X1jK~S۵ZWfX+jKښMZWfX+jKHqշ֊ڒ˧;Biˢng̲`7=j{_K <<^g̲ڎ[j{ˁ_K Z(i1jKrM~-10ky,mIm1k;y,k7=Jg̲ڪ=%f3fYsmWj{Ӂ_ ܩ]hM{jH~h8ETS{Zd`BIY$wai#j;׬J~Á_ Ye1^ۯJ=;o9k=|q6|I0wYڮSsW*5IDATK,O[Y Tc8M_Y_WT~7NSXo=} vqP?8oiқczfk;}Cb}mioJYP&J^Sm~duw j{_O=Cb_Bm1k=鵝5&v&30˟^%%rqS>y _YSœuO.Gq=~g`?u%\t`cj{_可Ꮻ$=͹7OxB0+t_2/Bgt7SmN6|7~<xښ/ݔq܁xftW?acwO+Sų/[o̟ _|ז;jۇIw? PۿhkkvZ+k-'ˌQvf/ԶS5տwsb{;e`-,eLDږ:B}6uakPeXmOujRꯕ;T_v*k`>7 jo^!:w }meqchP?vge7A+j{ԙRY86@cgؚkKr3/c3)CmhRAgf{vGNm/:B>6=DဳÞmva[smG+ŋ3/fGKKmx*O9jo[j{ԙTCTEX[7em=(ںWWMy!-\lo-) ,/aݦ6-h6,v^4u慴Ds0 /Î,=Jk^4u慴Dؖ8:A\[X%a|{j id_|4vWP[lR[k|zj{ԙ8A$95- UOj{ԙ8V{^ jCm:B`mDvԵm#fVLy!MKBj/Y>oxsmnA'j i;Ԫm4k`2cͲm'یeM荟![vΫ6cR˦μ=\ EHmsFsmj{09w ƵBRۣȹvQkμT`)7,_"k`%#%w7eF_@go,3P 3~}2vslg+<(p/L -iqã67~2vsgm V#-|dl0^ң d _[Z0c k>(Ϋh53/yk/Q %֖?c7a-~t_ 5 z2|jnne⫱? P#/6?6~+^ۇqƴ%d H~T K-}mnG3w08tQ]mnKn7=ԝvSxRjs7iw@@3xț F7 tLg3\%xIo<^?m⾒}^j-Qܖ3/q[>}L r^S+QEuWa>]'^MJG{Yz<趼;D6T8i5Φj3-2r=YO6ԶŮ6'8ӯCئO`fl,3.]^ݿkۆY<6v6{pդWvdt_?W^uU~c?.PnKa6Eu3vס߲J >~G?ky@)cSN@'ݼ~0j݇דFccm+*ly_<njg*s|4ko 3Y݈&ӟ)40/﫸jO!R}aVV-m25Uw>;jV`P`NT㵹|sXaڎNw>5Bm+0Ik 1}:[/_[ykvHP,ԶNXm]cW?o V|R,ԶڎHmh Qð>׏6o~ڮ}jV`PHhj;kUW?~kk7 bm5V?dmԶ1҈@vuk7? ΃?\WjVV[oY#bڪ*kKǼT[FSmn]R#ןjS˭LjkKǼ}F_m>j[ǵ}NmsӃ6>i]y6񝍇ƺͿ7mSԶ6ښ6= 5A)am}m>j[am bms;s>jRbmnUSk;jHnޟ:vI: 3j[_jGk G욻V*};նm -V[mT[Cmm\x/UzK jkveP kjk~[[wk^jRGKmxP[ב_ӗ*==鰶*n}fԶ}mamj; nvUv7K'۠YmFkK&T[$mo݊ϯ_[ܱ{*Yv/ЩmԶU9Rb Զc]&֦?^V`fjmnۻ&Gm+0Vqm5xVqXm}yMz^V`jjmiW~Bmqk\|m 9~lԶkqmGb=c^޶4HQ k>K j۸ j އ?F:q]s%P k^s kնQq׭}ms_[jz5cؓϠ6}_?orp:n3|ckw=}fԶw} }m~@렶]xjܩ㠶?Km^bmp6&x;w]zP%U]΁?w8?_WYmn%XX^;WJ; jpjRms|3cmgjVஶeS܅A1V[yHߣq?nSLjOi]G|ӞЩmj[[mto~M1J~|ڱ.}{}m4 -VVQmv\ [{o;{ 5ÿV6 /R?VnKR{Cՠ:=}ymn_ M⯃ڎS ]6o~7,V6 e?|m_>[9b{q"@|Jk7nv3omj[Am/ ~1I1s-oDnf mӛsX%ܽx9ߞRփC6< vz n'^fv|Qu~:Sݷ pσcitMucá6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT P6@jT^Y[jT P6@jT Py]m2PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *PBm *\?qlIENDB`$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 575 55}#v #v7#v #v#v}:V l0"6,5 575 55}9a]pytb2$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]p$$If]!vh5 5555#v #v#v#v:V l0 6,5 5559a]pF@F G Normal$a$CJ_HaJmH sH tH Z@Z  Heading 1$$<@&a$5CJKHOJQJaJX@X Heading 2$$<@&a$56CJOJQJaJR@R Heading 3$$<@&a$CJOJQJaJDA@D Default Paragraph FontVi@V  Table Normal :V 44 la (k(No List XOX X@B Title$a$5CJ OJQJ\aJ4@4 X@2> @ Footnote TextCJaJ@&@A@ @Footnote ReferenceH*3%(,6ol 2IJMZ 1@$Q(+-'178;?D|H-MNNTWKZ~Z^afgQggjm njnnntttttttttuuu#u,u-u;uAuGuLuQuWu]ubuguhuvu|uuuuuuuuuuuuuuuuuuuvv{!Ƀ-xB+̉U q/},<=?EF̜֜ޜ $%3UW]chnoŝ˝̝%+067LPV\aghqu{ΞҞ؞ޞ #)+129;AGLRScdˤӤܤ#'-38>?TX^diopץ٥ߥ 7;AGLRSkou{¦Ʀ̦ҦԦڦۦ "#*,28=DERgik]^`BCS)Ojns}  '(pw~ ;BIOVWx SZagno*18>EFov}MT[ahiglvPW^dkl NU\bij| &-.bipv}~"(/0SZagno89q"I#dJKFQlgCD}\6Z$~N3&e7uDW=f#g @  f  *    <{)zG00000000000000000000000000000000000000000000000 0 0 0 00000 0 00 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 000 0 0000000000000000000 00 0 000 0 000 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000000 00 0 000 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000000000000000000000000000000 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000000000000 0 0 000000 0 0 0000000000000000000000000000000000000000000000000000000000000000000000000000000000000@000%@000@000@000@000@0@00086ol 2IJMZ 78;?D|H-MNNTWKZ~Z^fgQggjm njnnnttttttttuuu#u,u-u;uAuGuLuQuWu]ubuguhuvu|uuuuuuuuuuuuuuuuuuuvv!Ƀ-xB+̉U},<=?EF̜֜ $%3UW]chnoŝ˝̝%+067LPV\aghqu{ΞҞ؞ޞ #)+129;AGLRSˤܤ#'-38>?TX^diopץ٥ߥ 7;AGLRSkou{¦Ʀ̦ҦԦڦۦ "#*,28=DEgk]^`)Ons}  '(pw~ ;BIOVWx SZgno*1>EFovMT[ahiglvPW^dkl NU\bij| &-.bipv}~"(/0SZagno9"#dJKQ@0@0@0@0@000)0000@0@0Z00Z00Z0000 0 00 07Z00Z00Z00Z00Z0000LZ00Z00Z00j00Gj00Ej00EZ00j00Dj00Cj00C@0@0000000@00#00#00#00#00#00#0j0*00 0*j00 j00 j00 j00  j00  j00 j0*0j00 j00  j00  j00 j00j00j00j00j000;0 0J0j0 0 j00 0;0 j00j00j00j00j000;0 090 j00 j000;0 0;0 0;0 0;0 j00j00@0 020 j00 j0"0 @0  @0 Z0000!00!Z0000!00!00!00 Z000000Z00000000Z000l0m0@0Z000n00n00n00 0*0A00A00A00A00 0*0 0*j00 j0 0 j00 j00j00 j00 j00 j00 j00j00 j0<0 j0<0 j00 j00 j00 0J00.0%j0<0 j0<0 j00 j00 j00 j0,0 0J0j00 j00 j00 j0 0 j0 0 j0<0 j0,0 0 0*0 0*j00 j00 j00  0 0*090 j00 j00 j00 j0+0  j00 0 0%j0<0 0J00J00J00J00J00Q00Q00Q00Q00Q00Q00Q00.0%0;0 0;0 0;0 j0 0 j0%0  j00  090 0 0%0 0%0 0%j00 j010 j0"0 0 0*0 0*0 0*0 0*j0 0 j00 j00  0;0 0 0*0 0*0 0*j00 j070 j0&0 0 0*0 0*0 0*0 0*j00 j0=0 j0*0 0000%00$00Oj00 j0 0 j00 j00j00 j00 j00 j00 0J00E00J0j00 j0 0 j00) j0<0 j0<0 0J00J00J00J00J00Q00.0%0.0%0.0%j00 j00 j0&0$ j0,0 j00j00 0.0%0.0%j00 j00 j00. 000;0 020 020 j0 0 j0%0  j0D0 090 090 090 090 j00 j0+0  j0N0 0 0%j0,0 0 0'0 0'j00 j00 j000 0 0*0 0%0 0%0 0%j00 j010 j0X0  0 0*0Q00Q00Q00Q00Q00Q00.0%0 0*0Q00Q00Q00Q00Q00 0*0 0*0 0'0 0'j0 0 j00 j0:0 0;0 0 0*0 0'0 0'j00 j0=0 j0l0 0 0*Z000819,081081j00Z00Z00Z00Z00j00j00j00 j00 j00 j00 j00j00 j00 j00Yj00Y0L10L100|j0 0 0000000000j00 j00 j00 000000j00 j00 j00 00000000j0 0Yj0 0Y0V100n00yj0 0 j00Yj00Y0d10d100|j0 0 j00 j00 0j100f00sj00 j0 0 j0 0 00j00vj00 j00 j00 01b00pj00 j00 j00 01^00mj00 j00 j00 0 1Z 00jj00 j00 j00 01W00g000  000  000  01S00fj00 j00 j00 01O00cj00 j00 j00 0 1K0 1K00`j0d0  j0d0 j00 j00 j00 j00 j00j00 j00 j00 j00 j00j00j00j00 j0 0 j0 0 j00 j00 j00 j00 j00 j00 j00 j00 j00 j00 j0 0 j0 0 j00 j00 j00 j00 j00 j00 j00 j00 j00 j0 0 j00 j00 j0 0 j0 0 j0 0 j00 j00j00 j0+0 j0+0 j0+0 j00 j00 j00 j00 j00 j00 j00 j0 0 j0 0 j00 j00 j00 j0 0 j00 j00 j0'0 j0'0 j0'0 j00j00 j00 j0#0 j0#0 j0#0 j00 j00 j00 j0/0 j0/0 j0/0 j00 j00 j00 j030 j030 j030  @0  @00K0L1_Z00Z00Z00@0@0@0@0@ 0@ 0@ 0@0@0@0@0@0@ 0@ 0@ 0@0@0 @0000 00% 4TSvO%m3 y5Xu^,$''  5v6u?Qݪ :QcitܱAQlײ +5KXs{Vy%{Fb*yPsw[w Yu9;yD\Y '%'   'LE]F]$ G]H]>8I]$J]2K]d L] M]N]dO]P] Q]D R] S]D T]D U] V] W]| X]< Y] Z] []t \] ]] ^] _]t `] a]T b],Ic]Hd]5e]5f],5g]4h]|4i]<4j]3k]3l]3m]2n]Io]Ip]\Fq]Fr]Ks]Kt]Ju]LJv]?@ABCEDFGHIJK ++11177\\/^/^ _ _:z:znnaa .. rvv!%%x|U__+//""7EE  "!#%$&(')*+-,./0132465789:;=<>?@ABCEDFGHIJK =LL*urn:schemas-microsoft-com:office:smarttags PlaceType=JJ*urn:schemas-microsoft-com:office:smarttags PlaceName9BB*urn:schemas-microsoft-com:office:smarttagsState:HH*urn:schemas-microsoft-com:office:smarttagsStreet>FF*urn:schemas-microsoft-com:office:smarttags PostalCodeBEE*urn:schemas-microsoft-com:office:smarttagscountry-region;II*urn:schemas-microsoft-com:office:smarttagsaddress8GG*urn:schemas-microsoft-com:office:smarttagsCity9KK*urn:schemas-microsoft-com:office:smarttagsplace "LKJIHGFEGKBKGGKGKGKEKEKGKGGKGKLKJKGBKGBKGBGEKJLGKEKGBKGBKBGKKGBKGGKBKGBLKJGK0123:;DFGHQR[]abk2} V ] p w ~Za !"""#$$ $'$,$%%%%%%%%r&x&&&&&P)V)X)b)f)k)))))e*k*++++,,%,+,:,A,P,U,],c,s,x,|,,W.].x...................// ///0!000 2$2'2-2|44G5M5\6c6(707::,;4;OAOIO3W=WyWWXXXXKbPbcbhbtbzb hh3k9kmm n nnn n#nEnHnSnUn_nannnnnnnnnp pppppqqqqqqtttttttttttuuuuu$u&u'u+uxx%y(y*y-y/y2y4y8y:y>y*z0z8z:z!48<?sx)-]bhpÞ͞ dhcmݳU]`fnzQW$49)0%)18lw}gnDH}Z_ehz t|$)~NS38>AGN]d?F7;CN Wzgqw{ $f l r |    " * /       JP{)4<ERY| z #SX knok  1*255X8e8WWggRgTgpntnr9rttxx?FGOP'emnEkr`gOIpQzE5lgZCDH)STi}[\mikmv /Z6U~!N_23pDEWe7Z[MNtu(*Wz  (f #>?&g BCq @ c  ' f     ; < t    D E      ";<  >{)QH]`/0l-FGst()U&'L33333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333lmmnnnnt,u-uuvv̜E`h)On (p;So*FoMigPl N.b~0SoAQ{  JI'L%l8poj9 !lx8=po<+BtE^`OJQJo(hH^`OJQJ^Jo(hHopp^p`OJQJo(hH@ @ ^@ `OJQJo(hH^`OJQJ^Jo(hHo^`OJQJo(hH^`OJQJo(hH^`OJQJ^Jo(hHoPP^P`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohpp^p`OJQJo(hHh@ @ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohPP^P`OJQJo(hHhhh^h`OJQJo(hHh88^8`OJQJ^Jo(hHoh^`OJQJo(hHh  ^ `OJQJo(hHh  ^ `OJQJ^Jo(hHohxx^x`OJQJo(hHhHH^H`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHhhh^h`OJQJo(hHh88^8`OJQJ^Jo(hHoh^`OJQJo(hHh  ^ `OJQJo(hHh  ^ `OJQJ^Jo(hHohxx^x`OJQJo(hHhHH^H`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohpp^p`OJQJo(hHh@ @ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohPP^P`OJQJo(hHh88^8`OJQJo(hHh^`OJQJ^Jo(hHoh  ^ `OJQJo(hHh  ^ `OJQJo(hHhxx^x`OJQJ^Jo(hHohHH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHl=pol8JI'j9Bt                                                       G {T5(b28^H oX?TX^diopץ٥ߥ 7;AGLRSkou{¦Ʀ̦ҦԦڦۦ "#*,28=DEns}  '(pw~ ;BIOVWx SZagno*18>EFov}MT[ahikQglvPW^dkl NU\bij| &-.bipv}~"(/0SZagno89( EN.InstantFormatw<ENInstantFormat><Enabled>0</Enabled><ScanUnformatted>1</ScanUnformatted><ScanChanges>1</ScanChanges></ENInstantFormat>@qsxy 0@0z0@00 @00@00@00@00@08@UnknownGz Times New Roman5Symbol3& z Arial?5 z Courier New;& z Helvetica;Wingdings"1hF٪& [*s[*s!4d==2qHX ?X?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijkmnopqrstuvwxyz{|}~Root Entry F#Data G1Tablel\WordDocument.,SummaryInformation(DocumentSummaryInformation8CompObjq  FMicrosoft Office Word Document MSWordDocWord.Document.89q