ࡱ>  rFU@ ˂bjbj at^\4nnnh$8N(vvl$Rd5V("VV5vvJ###V vv#V##$ W|6+vv rengbT `0ɰbjjj+j+T\f#TTTT55dJ$j#.J Esperonto Services IST-2001-34373   Deliverable D21  State of the art on Semantic Web Languages  Ying Ding Sinuh Arroyo Institut fr Informatics University of Innsbruck {ying.ding, sinuhe.arroyo} @uibk.ac.at 22-01-2003 Executive Summary This deliverable delineates the state of the art of Semantic Web languages, which include SHOE, Ontobroker, XML(s), RDF(s), OIL, DAML+OIL, OWL and other related languages (DAML-S and Topic Maps) and technology. During the running of the project, this work package will be responsible for providing updated information about the evolution of these languages and technologies. Document Information IST Project NumberIST-2001-34373AcronymEsperonto ServicesFull titleApplication Service Provision of Semantic Annotation, Aggregation, Indexing and Routing of Textual, Multimedia, and Multilingual Web ContentProject URL HYPERLINK "http://www.esperonto.net" www.esperonto.net Document URLEU Project officerBrian Macklin  DeliverableNumber21NameState of the art in Semantic Web LanguagesTaskNumberNameWork packageNumber2Date of deliveryContractual30-10-2002Actual22-01-2003Code nameStatus draft ( final (NaturePrototype ( Report ( Specification ( Tool ( OtherDistribution TypePublic ( Restricted ( Consortium Authors (Partner)Ying Ding (IFI), Dieter Fensel (IFI), Sinuh Arroyo (IFI)Contact PersonYing DingEmailYing.ding@uibk.ac.at HYPERLINK mailto:dybkjr@pdc.dk Phone+43 512 507 6112Fax+43 512 507 9872Abstract (for dissemination)This deliverable describes the state of the art on Semantic Web Languages developed within the Esperonto project. KeywordsOWL, RDF, RDFS Semantic Languages, OIL, DAML+OILVersion log/DateChangeAuthor30-08-2002First VersionDieter Fensel20-10-02Second revisionYing Ding22-01-2003Third revisionSinuh Arroyo17-02-2003Internal Quality ProofSinuh Arroyo Project Information PartnerAcronymContactIntelligent Software Components S.A. (Coordinator)iSOCO Dr. V. Richard Benjamins c/ Francisca Delgado 11, 2nd floor 28108 Madrid (Alcobendas), Spain #e  HYPERLINK "mailto:richard@isoco.com" richard@isoco.com #t +34-91-334-97-97, #f +34-91-334-97-99Universidad Politcnica de MadridUPM Dr. Asuncin Gmez-Prez Campus de Montegancedo, sn Boadilla del Monte, 28660, Spain #e  HYPERLINK "mailto:asun@fi.upm.es" asun@fi.upm.es #t +34-91 336-7439, #f +34-91 352-4819Institut fr Informatik, Leopold-Franzens Universitt InnsbruckIFI Dr. Ying Ding Institute of computer science University of Innsbruck Technikerstr. 25 A-6020 Innsbruck, Austria #e  HYPERLINK "mailto:dieter.fensel@uibk.ac.at" dieter.fensel@uibk.ac.at #t +43 512 507 6486Universitt des SaarlandesUdS Dr. Hans Uszkoreit Universitaet des Saarlandes Computerlinguistik D-66041 Saarbruecken, Germany #e  HYPERLINK "mailto:uszkoreit@coli.uni-sb.de" uszkoreit@coli.uni-sb.de #t + 49 681 302-4115, #f: + 49 681 302-4700The University of LiverpoolUniLiv  EMBED PBrush Dr. Valentina A.M. Tamma Department of Computer Science, University of Liverpool Room 1.11, Chadwick Building Peach Street Liverpool L69 7ZF, UK #e  HYPERLINK "mailto:valli@csc.liv.ac.uk" valli@csc.liv.ac.uk #t +44 151 794 6797, #f +44 151 794 3715Fundacin Residencia de EstudiantesResidencia  Mr Carlos Wert Fundacin Residencia de Estudiantes Pinar, 23 28006 Madrid, Spain #e  HYPERLINK "mailto:cwert@fundacionginer.org" cwert@fundacionginer.org #t +34-91-446 01 97, #f +34-91-4468068Centr d'Innovaci i Desenvolupament EmpreserialCIDEM (Centr d'Innovaci i Desenvolupament Empreserial) Carlos Gmara Centr d'Innovaci i Desenvolupament Empreserial Provena, 339 08037 Barcelona, Spain #e  HYPERLINK "mailto:cgomara@cidem.gencat.es" cgomara@cidem.gencat.es #t +34-93-4767305, #f +34-93-4767303BiovistaBiovista Dr. Andreas Persidis 34 Rodopoleos Street Ellinikon Athens 16777, HELLAS #e  HYPERLINK "mailto:biovista@ath.forthnet.gr" biovista@ath.forthnet.gr #t +30.1.9629848, #f +30.1.9647606 Table of Contents  TOC \o "1-3" 1. INTRODUCTION  PAGEREF _Toc33275002 \h 7 2. SEMANTIC WEB LANGAUES  PAGEREF _Toc33275003 \h 9 2.1 Early age: SHOE and Ontobroker  PAGEREF _Toc33275004 \h 9 2.1.1 SHOE  PAGEREF _Toc33275005 \h 10 2.1.2 Ontobroker  PAGEREF _Toc33275006 \h 10 2.1.3 Summary  PAGEREF _Toc33275007 \h 11 2.2 XML and its family  PAGEREF _Toc33275008 \h 12 2.2.1.XML  PAGEREF _Toc33275009 \h 12 2.2.2 XML Schema (XMLs)  PAGEREF _Toc33275010 \h 12 2.2.3 XML Family  PAGEREF _Toc33275011 \h 13 2.3 RDF and its family  PAGEREF _Toc33275012 \h 14 2.3.1 RDF  PAGEREF _Toc33275013 \h 14 2.3.2 RDF Schema (RDFS)  PAGEREF _Toc33275014 \h 15 2.4 OIL, DAM+OIL and OWL  PAGEREF _Toc33275015 \h 17 2.4.1 OIL  PAGEREF _Toc33275016 \h 17 2.4.2 DAML+OIL  PAGEREF _Toc33275017 \h 20 2.4.3 OWL  PAGEREF _Toc33275018 \h 21 2.5 RuleML  PAGEREF _Toc33275019 \h 26 2.6 Topic Map and its family  PAGEREF _Toc33275020 \h 26 2.7 DAML-S  PAGEREF _Toc33275021 \h 27 2.8 Supporting tools  PAGEREF _Toc33275022 \h 28 3. COMPARISON  PAGEREF _Toc33275023 \h 29 3.1 General comparison of modeling primitives  PAGEREF _Toc33275024 \h 29 Factual knowledge: Data Models  PAGEREF _Toc33275025 \h 29 Terminological knowledge: ontologies  PAGEREF _Toc33275026 \h 29 Inference Knowledge  PAGEREF _Toc33275027 \h 30 3.2 Specific comparison between two languages  PAGEREF _Toc33275028 \h 31 XMLs vs. DTD  PAGEREF _Toc33275029 \h 31 RDF vs. XML  PAGEREF _Toc33275030 \h 31 OIL vs. XML  PAGEREF _Toc33275031 \h 32 OIL vs. XMLs  PAGEREF _Toc33275032 \h 32 RDFS vs. XMLs  PAGEREF _Toc33275033 \h 33 OIL vs. RDF(s)  PAGEREF _Toc33275034 \h 33 DAML+OIL vs. RDFS  PAGEREF _Toc33275035 \h 34 DAML+OIL vs. OIL  PAGEREF _Toc33275036 \h 34 DAML+OIL vs. OWL  PAGEREF _Toc33275037 \h 35 4. ANALYSIS  PAGEREF _Toc33275038 \h 36 5. CONCLUSION AND FUTURE PLAN  PAGEREF _Toc33275039 \h 44 REFERENCES  PAGEREF _Toc33275040 \h 47 Appendix 1  PAGEREF _Toc33275041 \h 49  INTRODUCTION The current World Wide Web (WWW) is, by its function, the syntactic web where structure of the content has been presented while the content itself is difficult to access to computers. Although the WWW has resulted in a revolution in information exchange among computer applications, it still cannot fulfil the interoperation among various applications without some pre-existing, human-created agreements somewhere in-house or outside of the web. The next generation of the Web aims to alleviate such problem. The Web resources will be much easier and more readily accessible by both human and computers with the added semantic information in a machine-understandable and machine-processible fashion (Berners-Lee, 1999). "The Semantic Web is an extension of the current web in which information is given well-defined meaning, better enabling computers and people to work in cooperation," wrote by Tim Berners-Lee, James Hendler and Ora Lassila in their Scientific American article "The Semantic Web". How to make the Semantic Web possible in a way that computer could understand the semantic meaning of the information presented on the Web. Ontology here is the magic-maker. It plays a pivotal role by providing a source of shared and precisely defined terms that can be understood and processed by machines. A typical ontology consists of a hierarchical description of important concepts and their relations in a domain, task or service. The degree of formality employed in capturing these descriptions can be quite variable, ranging from natural language to logical formalisms, but increased formality and regularity clearly facilitates machine understanding. Therefore a decent ontology language which can help in the formality on the web is the most wanted thing in the Semantic Web. Refer to D1.1 for further information. Various wish lists for the requirement of such Web ontology languages have been flied around the Web. To name but not limited, such as: it should be well designed for the intuition of human users without loosing the adequate expressive power; it should be well defined with clear specified syntax and formal semantics; it should be compatible with existing web standards, etc. The concrete requirement report on Web ontology language design is well-documented by Heflin, Volz and Dale (2002). In this survey, we intent to have the broader coverage for various existing web ontology languages, starting from SHOE and Ontobroker, and following the historical line, XML(s), RDF(s), OIL, DAML+OIL, and OWL. We also bring some attentions for DAML-S and TopicMaps. The structure of this survey is planned as follows: In Section 2 Semantic Web Languages, we give the introduction for each language family. In Section 3, the comparison has been conduct among and between these languages. Section 4 gives the summarized comparison result and analysis. Section 5 contains the final summary and future plan. SEMANTIC WEB LANGAUES As Tim Bernes-Lee described, the Semantic Web Languages are like the language cake (see Figure 1). The layered tower is the dreamed vehicle to bring the Semantic Web to its full potential. The recognition of the importance of ontologies for the Semantic Web has led to the revolution and extension of the current web markup languages, e.g., XML Schema, RDF (Resource Description Framework), and RDF Schema, furthermore, OIL, DAML+OIL, and OWL.  EMBED Word.Picture.8  Figure 1: Semantic Web Language cake (layered tower) The structure of this Section is planned as following the history of the web language development. First, we introduce the origin of the first-try (SHOE and Ontobroker), then main stream of the language development: XML(s), RDF(s), OIL, DAML+OIL, and OWL. At the end, We will bring your attentions with DAML-S, TopicMap and other related initiative world-wide. 2.1 Early age: SHOE and Ontobroker In the early 90s, when the current Web just popped out and changed the whole world overnight like a tornado. When people are still indulged by the magic and realize that they need to spend much more time and effort to get used to the changes to their normal daily life brought by the current Web, some prescient researchers already foresee the limits of the current Web. Therefore the early initiatives starts almost parallelly together with the development of the current web. These initiatives are SHOE and Ontobroker. 2.1.1 SHOE SHOE has been developed in the University of Maryland (USA) in early 90s. It creates an extension of HTML by adding tags that are necessary to embed semantic data into web pages. This extension contains two categories: tags for constructing ontologies and tags for annotating web documents (Heflin and Hendler, 2001 and 2000; Heflin, Hendler and Luke, 1999; Luke, Spector, Rager and Hendler, 1997). HTML -tags is the first attempt at representing semantics inside Web-documents. Their intended use is limited to stating global properties that apply to the entire document. The anchors in META-tags are not standardized, which can be exploited in software if one wishes to, but cannot be interpreted by standard Web-browsers and search-engines. SHOE proposes an extension of the HTML -tag concept which can occur both in and of a document. These SHOE expressions are separate from the contents of a document, and can be applied to the entire document, whereas HTML -tags are limited to attribute-value pairs. SHOE expressions also include binary relations between instances. SHOE allows representing concepts, their taxonomies, n-ary relations, instances and deduction rules, which are used by its inference engine to obtain new knowledge. Is important to notice that this language is no longer being maintained. 2.1.2 Ontobroker Ontobroker has been developed in the University of Karlsruhe (Germany) during the mid 90s. It applies Artificial Intelligence techniques to improve access to heterogeneous, scattered and semi-structured information sources. It relies on the use of ontologies to annotate web pages, formulate queries, and derive answers. Ontobroker provides a broker architecture with three core elements: a query interface for formulating queries, an inference engine used to derive answers, and a webcrawler used to collect the required knowledge from the Web. The gist of Ontobroker is to create a methodology to define an ontology and use it to annotate/structure/wrap the web documents, and furthermore to make use of its advanced query and inference services (Fensel et al., 1999). The Languages Ontobroker defines three interleaved languages: an annotation language to enrich web documents with ontological information, a representation language to formulate ontologies, a query language (which is the subset of the representation language) to formulate queries. The annotation language provided by Ontobroker is called HTMLA, which enables the annotation of HTML documents with machine-processable semantics. It extends the anchor tag with additional attribute, called the onto attribute, to annotate the web document based on three primitives object, value, relationship. The same piece of data that is rendered by a browser is now having a semantic meaning defined by HTMLA. A representation language is used to formulate an ontology, which defines the terminology (i.e., signature) and rules (i.e., axioms) that allow the derivation of additional facts. This language is based on Frame logic (Kifer et al., 1995) and introduces the terminology that is used by the annotation language to define the factual knowledge provided the Web, such as, class definition, attribute definition, is-a relationship, is-element-of relationship and rules. The query language is defined as a subset of the representation language. The elementary expression is written in Frame logic: x[attribute -> v] : c. Complex expressions can be built by combing these elementary expressions with the usual logical connectives. The Tools Ontobroker is implemented into two tools: a webcrawler and an inference engine. The webcrawler collects web pages from the Web, extracts their annotations, and parses them into the internal format of Ontobroker. The inference engine takes these facts together with the terminology and axioms of the ontology, and derives the answers to user queries. A hyperbolic presentation of the ontology and a tabular interface improve the accessibility of Ontobroker. Ontobroker was presented as a means to improve access to information provided in intranets and in the Internet (Fensel et al., 1999). 2.1.3 Summary The ontology language provided by SHOE and Ontobroker are based on the current HTML with proper extension. Therefore the normal web browser can still interpret the document by understanding additional embedded semantics. Especially the Ontobroker can be treated as the earliest Semantic Web prototype. There are two main differences between SHOE and Ontobroker. First, the annotation language is not used to annotate existing information on the Web, but to add additional information and annotate them. That is, in SHOE, information must be repeated and this redundancy may cause significant maintenance problems. Ontobroker uses the annotations to directly add semantics to textual information that is also rendered by a browser. A second difference is the use of inference techniques and axioms to infer additional knowledge. SHOE can rely on frame-based systems in order to deal with inheritance. Ontobroker uses an inference engine to answer queries. Therefore, it can make use of rules that provide additional information. 2.2 XML and its family With the exponentially increased web information, HTML is considered as too simple to present the document. The limited tags failed to provide some essential information on the Web, which limits HTML to only represent the layout of the information. Therefore the new language has to be designed to cater the new requirements emerging from web applications. XML is a tag-based language for describing tree structures with a linear syntax. It offers the facilities for users to define their own tags, which are needed for describing the structure of the documents. In this part, we discuss XML and the important members in its family. 2.2.1.XML XML provides seven different means for presenting information: document, element, attribute, text, namespace, processing instruction, and comment. A DTD consists of three elements: Element declaration that define composed tags and value ranges for elementary tags, attribute declaration that define attributes of tags, and finally entity declaration. For more details, please visit http://www.w3.org/XML. 2.2.2 XML Schema (XMLs) XML schemas are means for defining constraints on valid XML documents. They have the same purpose as DTDs but provide several significant improvements: definitions are itself XML documents. The clear advantage is that all tools developed for XML (e.g., validation or rendering tools) can be immediately applied to XML schema definitions. a rich set of datatypes that can be used to define the values of elementary tags. much richer means for defining nested tags (i.e., tags with sub-tags). the namespace mechanism to combine XML documents with heterogeneous vocabulary. 2.2.3 XML Family As XML has been quickly adopted by academy and industry due to the easy use and implementation, various XML-like languages are designed for specific purposes. Here we list some important XML family members. XSL In XML, users can define their own tags therefore additional style sheet information is required for a browser to render such XML documents. XSL is mainly designed for this purpose to express the format information for XML documents. Furthermore, XSL allows defining views that can manipulate structure and elements of a document before they are rendered. XSL even enables the translation of one XML document into another one by using a different DTD. This is important in cases, where different user may wish to have different views on the information captured in a XML document. XML is a standard language for defining tagged languages. However, XML does not provide standard DTDs, i.e., each user can/may/must define his own DTD. For exchanging data between different users relying on different DTDs one has to map different DTDs onto each other. XSL can be used to translate XML documents using DTD1 in an XML document using DTD2. Query Languages for XML The need for a query language for XML becomes obvious when comparing the WWW to a database. Query languages for XML provide query answering service which also applicable to semi-structured data. Currently, there exist still a number of proposals (XQL, XQuery, XPath). Most of them can be found in (http://www.w3.org/XML/Query). Resources  HYPERLINK http://www.w3.org/XML/ http://www.w3.org/XML/  HYPERLINK http://www.w3.org/XML/Core/ http://www.w3.org/XML/Core/  HYPERLINK http://www.w3.org/XML/Linking http://www.w3.org/XML/Linking  HYPERLINK http://www.w3.org/XML/Query http://www.w3.org/XML/Query  HYPERLINK http://www.w3.org/XML/Schema http://www.w3.org/XML/Schema  HYPERLINK http://www.w3.org/Style/XSL/ http://www.w3.org/Style/XSL/ Summary The first push towards more semantic structure on the Web has been the development of the XML. It allows the Web-page creators to define their own tags. In essence, XML allows structuring Web-pages as labeled trees, where the labels can be chosen to reflect as much of the document semantics as required. In general, XML is used for two purposes: for the markup of individual pieces of data, and as serialization syntax for other languages (eg SMIL, or RDF). In the first case, XML itself is used as a language to model meta-data. In the second use, XML is used as a language definition vehicle to define another language, which is in turn used to model meta-data. XML provides semantic information as a by-product of defining the structure of the document. XML prescribes a tree structure for documents and the different leaves of the tree have a well-defined tag. The structure and semantics of document are interwoven. However, important aspects are lacking of rules and constraints (i.e., class definitions). They are often a significant part of the knowledge provided by an ontology. 2.3 RDF and its family Although XML provides much more space for users to define their own tags, it fails to define the semantics in the machine understandable and processable way. RDF comes to fill up the hole. In this section, we will discuss RDF and its important family members. 2.3.1 RDF The advantage that Semantic Web brings is that the computer can understand and process the semantics of the information on the current Web. The Resource Description Framework (RDF) is an important step toward that direction. It provides means for adding semantics to a document without making any assumption about the structure of the document. RDF is an infrastructure that enables the encoding, exchange and reuse of structured metadata. Search engines, intelligent agents, information broker, browsers and human user can make use of semantic information. RDF is an XML application (i.e., its syntax is defined in XML) customized for adding meta information to Web documents. Basically, RDF defines a data model for describing machine processable semantics of data, which consists of three object types: Resources: A resource may be an entire Web page; a part of a Web page; a whole collection of pages; or an object that is not directly accessible via the Web; e.g. a printed book. Resources are always named by URIs. Properties: A property is a specific aspect, characteristic, attribute, or relation used to describe a resource. Statements: A specific resource together with a named property plus the value of that property for that resource is a RDF statement. These three individual parts of a statement are called, respectively, the subject, the predicate, and the object. In a nutshell, RDF defines object-property-value-triples as basic modeling primitives and introduces a standard syntax for them. As RDF statements are also resources, statements can be recursively applied to statements allowing their nesting. RDF emphasizes the facilities to enable automated processing of Web resources. The broad goal of RDF is to define a mechanism for describing resources that makes no assumptions about a particular application domain, nor defines (a priori) the semantics of any application domain. The definition of the mechanism should be domain neutral, and the mechanism should be suitable for describing information about any domain. The foundation of RDF is a model for representing named properties and property values. The RDF data model is a syntax-neutral way of representing RDF expressions. RDF properties may be thought of as attributes of resources and in this sense correspond to traditional attribute-value pairs. RDF properties also represent relationships between resources. As such, the RDF data model can therefore resemble an entity-relationship diagram. However, it doesnt provide mechanisms for declaring these properties, and the relationships between these properties and other resources. That is the role of RDF Schema. RDF has been and will be applied in various areas. It provides better capabilities for search engine in resource discovery. It describes the content and content relationships in formalized way to provide computer-understandable catalogues. It facilitates the knowledge sharing and exchange among various intelligent software agents. 2.3.2 RDF Schema (RDFS) In the Semantic Web, people are capable to describe the attributes of certain kinds of resources, for instance, to describe the "author", "title", and "subject" for certain bibliographic resources. The declaration of these properties (attributes) and their corresponding semantics are defined in the context of RDF as a RDF schema. A schema defines not only the properties of the resource (e.g., title, author, subject, size, color, etc.) but also the kinds of resources being described (books, Web pages, people, companies, etc.). The RDFS specifies the mechanisms needed to define the classes of resources, to restrict possible combinations of classes and relationships, and to detect violations of those restrictions. RDFS is based on non-standard model theory which different layers of modeling primitives are glued together at one layer (Nejdl et al., 2000), which makes RDFS hard to understand. RDFS in particular can be recognized as an ontology language: it talks about classes and properties, range and domain constraints, and subclass and subproperty relations. However, it is a very limited language and more expressive power is clearly demanded to describe data in sufficient detail. Furthermore, such descriptions should be able to support the automated reasoning. Resources  HYPERLINK http://www.w3.org/RDF/ http://www.w3.org/RDF/  HYPERLINK http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/ http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/  HYPERLINK http://www.w3.org/TR/rdf-schema/ http://www.w3.org/TR/rdf-schema/  HYPERLINK http://www.w3.org/TR/rdf-mt/ http://www.w3.org/TR/rdf-mt/  HYPERLINK http://www.w3.org/TR/rdf-syntax-grammar/ http://www.w3.org/TR/rdf-syntax-grammar/  HYPERLINK http://www.w3.org/TR/rdf-testcases/ http://www.w3.org/TR/rdf-testcases/  HYPERLINK http://www.w3.org/TR/rdf-primer/ http://www.w3.org/TR/rdf-primer/  HYPERLINK http://www.w3.org/TR/rdf-concepts/ http://www.w3.org/TR/rdf-concepts/ 2.3.3 Summary RDF and RDFS provide the way to represent the semantics of information on the Web. They allow the representation of concepts, taxonomies of concepts and binary relations. It is also intended to provide mechanisms to explicitly represent services, processes and business models, allowing non-explicit information to be recognized. They are the first web language capable to define the ontology in a limited way due to the lack of the necessary rich concept forming operators. However, richer languages can be built on top of them (such as OIL, DAML+OIL, and OWL). Indeed this layered approach does provide benefits. Applications that are aware or understand RDF and RDFS can still process OIL, DAML+OIL or OWL ontologies. Although the application may not be able to extract and use all the information within the ontology, basic information about class hierarchies can be still available to such an application. RDF and RDFS only provided limited reasoning mechanism, which is mainly for constraint checking. While much powerful reasoning service is clearly demanded by the Semantic Web. However, many design decisions in RDF are rather particular. Properties are defined globally, violating all modeling experiences from object-oriented and frame-based approaches that define attributes locally for classes. RDF provides powerful reification, i.e. expressions can become terms in meta-level expressions. This makes reasoning service for RDF rather tricky. Most description do not provide such a service. Finally, RDF also does not provide a means for defining constraints and rules, which allow us to specify information intentionally. 2.4 OIL, DAML+OIL and OWL Due to the pros and cons of RDF and RDFS, efforts have been allocated for developing a more expressive ontology language for the Semantic Web. OIL is one of the trials and has successfully generated the critical mass. In this section, we discuss OIL and its important family member, especially OWL. 2.4.1 OIL OIL has been developed by the European IST project - Ontoknowledge. As shown in figure 2, OIL is built on top of RDF and RDFS, using as much as possible their constructs in order to maintain backward compatibility. OIL provides modeling primitives used in frame-based and Description Logic oriented ontologies, coming along with a simple and clean semantics. It has a syntax definition using web standards such as RDF(s) and XML(s).  Figure 2. Three pillars of OIL The three pillars of OIL OIL unifies three important aspects provided by different communities (see Figure 2): (1) formal semantics and efficient reasoning support as provided by Description Logic, (2) epistemologically rich modeling primitives as provided by the Frame-based community, and (3) a standard proposal for syntactical exchange notations as provided by the Web community. The details are: Description Logics (DL). OIL inherits from Description Logic its formal semantics and the efficient reasoning support (FaCT reasoner). OIL is capable to describe knowledge in terms of concepts and role restrictions that are used to automatically derive classification taxonomies. The reasoning support contains class and property subsumption, instance classification, query subsumption and query answering over classes and instances, and navigation through ontologies, etc. The FaCT reasoner is used to perform automatic classifications of concepts and constraint checking in taxonomies of concepts. Frame-based systems. OIL incorporates the essential-modeling primitives of frame-based systems on the notion of a concept, its superclasses and slots. OIL also treats slots as first class objects that can have their own properties (e.g., domain and range) and can be arranged in a hierarchy. With frame-based modeling we make the implicit assumption that only those attributes can be applied to a class that are defined for this class. Web standards: XML and RDF. First, OIL has a well-defined syntax in XML and XML schema definition. Second, OIL is defined as an extension of the RDF and RDF Schema. OIL extends this approach to a full-blown modeling language. When describing ontologies in OIL we have to distinguish three different layers: The object level, where concrete instances of an ontology are described (Klein et al., 2000). The first meta level (called ontology definition in OIL) where the actual ontological definitions are provided. Here we define the terminology that may be instantiated at the object level. OIL is a means for describing structured vocabulary with well-defined semantics. The main contribution of OIL is in regard to this level. The second meta level (i.e., the meta-meta level, called ontology container in OIL) is concerned with describing features of such an ontology, like author, name, subject, etc. For representing metadata of ontologies OIL makes use of the Dublin Core Metadata standard. Reasoning support is another feature of OIL. Reasoning can be used as the design support tool for large ontologies with multiple authors and as the back support for integrating and sharing ontologies. It can provide automatic consistency checking when you edit or generate ontologies especially with many authors synchronously. It can cross-link the inter-ontology relationships and check for the implied relationships. Layered cake of OIL It is unlikely that a single ontology language can fulfill all the needs of the large range of users and applications of the Semantic Web. Therefore good ontology language should be layered to provide different levels of services. The core language should only contain consensus primitives. Additional expressive power can be easily added to different layers by language extensions. After two-year development and refinement of OIL, OIL presents a layered approach to a standard ontology language. Each additional layer adds functionality and complexity to the previous layer. This is done in such a way that agents (humans or machines) who can only process a lower layer can still partially understand ontologies that are expressed in any of the higher layers. Figure 3 sketches the relation between the OIL dialects and RDFS:  Figure 3. The layered structure of OIL Core OIL coincides largely with RDF Schema (with the exception of the reification features of RDF Schema). This means that even simple RDF Schema agents are able to process the OIL ontologies, and pick up as much of their meaning as possible with their limited capabilities. Standard OIL is a language intended to capture the necessary main stream modeling primitives that both provide adequate expressive power and are well understood by allowing the semantics to be precisely specified and complete inference. Instance OIL includes a thorough individual integration. While the previous layer - Standard OIL - included modeling constructs that allow individual fillers to be specified in term definitions, Instance OIL includes a full-fledged database capability. Heavy OIL may include additional representational (and reasoning) capabilities. The layered architecture of OIL has three main advantages: First, an application is not forced to work with a language that offers significant more expressiveness and complexity than it actually needs. Second, applications that can only process a lower level of complexity are still able to catch some of the aspects of an ontology. Third, an application that is aware of a higher level of complexity can still also understand ontologies expressed in a simpler ontology language. 2.4.2 DAML+OIL DAML+OIL is a semantic markup language for Web resources that has been created as a joint effort of the American and European ontology communities for the Semantic Web. The Joint EU/US ad hoc Agent Markup Language Committee was installed to further develop DAML+OIL. DAML+OIL is the result of merging DAML-ONT (an early result of the DARPA Agent Markup Language (DAML) programme) and OIL (the Ontology Inference Layer). This language has a clean and well-defined semantics, and many efforts are being put to provide reasoning mechanisms for DAML+OIL. DAML+OIL is an ontology language specifically designed for the Semantic Web. It exploits existing Web standards (XML and RDF), adding the ontological primitives of object oriented and frame-based systems, and the formal rigor of expressive description logic. As an ontology language, DAML+OIL is designed to describe the structure of a domain. DAML+OIL takes an object-oriented approach, with the structure of the domain being described in terms of classes and properties, and the set of axioms that assert characteristics of these classes and properties. Datatypes DAML+OIL supports the full range of datatypes in XML Schema: the so called primitive datatypes such as string, decimal or float, as well as more complex derived datatypes such as integer sub-ranges. This is facilitated by maintaining a clean separation between instances of object classes (defined using the ontology language) and instances of datatypes (defined using the XML Schema type system). From a theoretical point of view, this design means that the ontology language can specify constraints on data values, but as data values can never be instances of object classes. This allows the type system to be extended without having any impact on the ontology language, and vice versa. Similarly, the formal properties of hybrid reasoners are determined by those of the two components; in particular, the combined reasoner will be sound and complete if both components are sound and complete. Axiom The axioms supported by DAML+OIL includes: subClassOf, sameClassAs, subPropertyOf, samePropertyAs, disjointWith, sameIndividualAs, differentIndividualFrom, inverseOf, transitiveProperty, uniqueProperty, unambiguousProperty. A crucial feature of DAML+OIL is that subClassOf and sameClassAs axioms can be applied to arbitrary class expressions. This provides greatly increased expressive power with respect to standard frame-based languages. A consequence of the expressive power of DAML+OIL is that all of the class and individual axioms, as well as the uniqueProperty and unambiguousProperty axioms, can be reduced to subClassOf and sameClassAs axioms. DAML+OIL also allows properties of properties to be asserted. It is possible to assert that a property is unique (i.e., functional) and unambiguous (i.e., its inverse is functional). It is also possible to use inverse properties and to assert that a property is transitive. The meaning of DAML+OIL is defined by a standard model-theoretic semantics. The semantics is based on interpretations, where an interpretation consists of a domain of discourse and an interpretation function. Inference in DAML+OIL DAML+OIL is equivalent to a very expressive Description Logic, i.e. SHOIQ, with the addition of existentially defined classes (i.e., the one of constructor) and datatypes. This equivalence allows DAML+OIL to exploit the considerable existing body of description logic research to define the semantics of the language and to understand its formal properties. Reasoning can be useful at many stages during the design, maintenance and deployment of ontologies. Reasoning can be used to support ontology design and to improve the quality of the resulting ontology. For example, class consistency and subsumption reasoning can be used to check for logically inconsistent classes and (possibly unexpected) implicit subsumption relationships. Ontology integration can also be supported by reasoning (Calvanese et al. 1998). 2.4.3 OWL OWL is the web ontology language currently under the development of W3C Web Ontology (WebOnt) Working Group. OWL can help power automated tools for the next generation web, offering advanced services such as more accurate Web search, intelligent software agents and knowledge management. OWL is mainly based on OIL and DAML+OIL and therefore the main features of OWL are very similar to those of OIL. Heflin, Volz and Dale (2002) proposed the wish lists for OWL based on various case studies. OWL includes three sub languages called OWL-Lite, which roughly consists of RDFS plus equality and 0/1-cardinality, OWL DL that contains the whole OWL vocabulary, and OWL Full, composed by the complete vocabulary interpreted more broadly than in OWL DL. OWL-Lite suits particularly well to express the light weight ontologies. But medical science and technical engineering communities among others are complaining about the limited expressive power of OWL-Lite. What they want is much full-fledged semantic web modeling language to fulfill the requirement so as to define much complex and heavy weight ontologies they have in their own fields, the language with at least the expressive power of full OIL/OWL. Therefore the layered design algorithm is very essential for OWL as well. Currently OWL-Lite is functioned as sub language of full OWL. OWL Abstract Syntax Besides the DAML+OIL style RDF syntax, the OWL specification also includes an abstract syntax, which provides a higher level and less cumbersome way of writing ontologies. It also has the advantage of allowing a more succinct statement of the semantics. The abstract syntax is defined using an extended BNF notation. A translation from this syntax to the RDF syntax is available. The OWL abstract syntax is very much like OIL in that it provides for compound axioms resembling frames. It is interesting to observe that the OWL abstract syntax has reverted to grouping axioms into frame structures. The basic idea of having a semantic web language to represent ontology is aiming to allowing computer programs to inter-operate each other without pre-existing, outside-of-the-web agreements. If this language also has an effective reasoning mechanism, then computer programs can manipulate this interoperability information themselves. The OWL language can be used to allow the explicit representation of term vocabularies and the relationships between entities. This language goes beyond XML, RDF and RDF-S in allowing greater machine readable content on the web. The OWL language is a revision of the DAML+OIL web ontology language-incorporating lesson learned from the design and application use of DAML+OIL. Main features of OWL language include: Ontologies: An OWL ontology is a sequence of axioms and facts, plus inclusion references to other ontologies, which are considered to be included in the ontology. OWL ontologies are web documents, and can be referenced by means of a URI. Ontologies also have a non-logical component (not yet specified) that can be used to record authorship, and other non-logical information to be associated with a ontology. Axioms: Axioms are used to associate class and property IDs with either partial or complete specifications of their characteristics, and to give other logical information about classes and properties. It contains Class Axioms, Property axioms, Descriptions and Restrictions. Facts: Facts state information about particular individuals in the form of a class that the individual belongs to plus properties and values. Individuals can either be given an individualID or be anonymous (blank nodes in RDF terms). The syntax here is set up to mirror the normal RDF/XML syntax. The OWL datatypes is on hold pending the outcome of RDFS. It is assumed that the semantics of OWL will be similar to those of DAML+OIL. The standard semantics of OWL are based on an open world assumption (OWA). This means that we cannot assume that all information is known about all the individuals in the domain. Thus, simply being unable to prove that an individual a is an instance of X does not justify our concluding that a is not an instance of X. This is in contrast to languages or systems using negation as failure or a closed world assumption (CWA). The OWA facilitates reasoning about intentional definitions of classeswe do not need to know all the information about the world in order to be able to make deductions about the relationships between classes. OWL Lite The objective of OWL Lite is to provide an easy-going language for tool builders. Tools play very important role for securing the widespread adoption of OWL. But this has to be balanced by without loosing the necessary expressive power of DAML+OIL or OIL. In order to provide a semantic web language that fits to the various needs of a wider audience, the language itself must have the layered structure. Therefore OWL Lite has been developed to capture many of the commonly used features of OWL and DAML+OIL. It attempts to provide more functionality than RDFS, which is important in order to support web applications. OWL Lite is a subset of the full OWL language constructors and has a few limitations therefore. Unlike the full OWL language (and DAML+OIL), classes can only be defined in terms of named superclasses. Equivalence for classes, and subclass between classes are only allowed on named classes. Property restrictions in OWL-Lite must use named classes as well. The only cardinalities allowed to be explicitly stated in OWL Lite are 0 or 1. See Appendix 1 for more details. OWL DL OWL DL includes the complete OWL vocabulary, interpreted under a number of simple constraints. Primary among these is type separation. Class identifiers cannot simultaneously be properties or individuals. Similarly, properties cannot be individuals. OWL DL is so named due to its correspondence with  HYPERLINK "http://www.w3.org/TR/2002/WD-owl-guide-20021104/" \l "DescriptionLogics#DescriptionLogics" description logics. OWL Full OWL Full includes the complete OWL vocabulary, interpreted more broadly than in OWL DL, with the freedom provided by RDF. In OWL Full a class can be treated simultaneously as a collection of individuals (the class extension) and as an individual in its own right (the class intension). Another significant difference from OWL DL is that a DatatypeProperty can be marked as an InverseFunctionalProperty. These are differences that will be of interest to the advanced user. This document does not describe the use of these features. OWL DL and OWL-Lite OWL extends the constructions of OWL Lite with the following: oneOf (enumerated classes): Classes can be described by enumeration of the individuals that make up the class, e.g., the class of daysOfTheWeek. hasValue (property values): A property can be required to have a certain individual as a value (also sometimes referred to as property fillers), e.g., the class of austriaCitizens can be characterized as those people that have Austria as a value of their nationality. (Where Austria itself is an instance of the class of all nationalities). disjointWith: The full OWL language allows the statement that classes are disjoint, for example stating that man and woman are disjoint classes. sameClassAs and rdfs:subClassOf: those are added for class expressions. unionOf, complementOf, and intersectionOf (Boolean combinations): OWL allows arbitrary Boolean combinations of classes. minCardinality, maxCardinality, cardinality (full cardinality): While in OWL Lite, cardinalities are restricted to either 1 or 0. Full OWL allows cardinality statements for arbitrary non-negative integers. complex classes : OWL Lite restricts the syntax to single class name. Full OWL extends this to allowing arbitrarily complex class descriptions, consisting of enumerated classes, property restrictions, and Boolean combinations of these. OWL also includes a special "bottom" class with the name Nothing that is the empty class. DAML+OIL vs. OWL (OWL-Lite) Although OWL and its family are still under development, many new features or changes might happen pretty soon (we will keep on documenting the changes and status of the development for various version of Del 22 during the whole project period). At the moment, OWL differs from DAML+OIL in the following respects: The namespace has been changed to  HYPERLINK http://www.w3.org/2002/07/owl http://www.w3.org/2002/07/owl. Several bugs and omissions in RDF and RDF Schema have now been fixed by the RDF Core Working Group. This allows OWL to use official RDF syntax (in contrast to DAML+OIL). Examples include: RDF now supports cyclical class and property inclusions (using rdfs:subClassOf and rdfs:subPropertyOf), so the relevant RDF properties can now be used in OWL. RDF now supports multiple rdfs:domain and rdfs:range properties, with both being treated as equivalent to the intersection of the individual constraints. RDF now includes a collection parse type (rdf:parseType=Collection) which replaces the old rdf:parseType=daml:collection. RDF now has its own formal semantics (a model theory). But the relationship between this and the OWL semantics is not yet clear. The RDF core working group is examining the issue of datatypes and may produce a recommendation. The issue of OWL datatypes is on hold pending the outcome of their work. Some DAML+OIL properties and classes have been renamed. Table 1 shows the relevant DAML+OIL names and their corresponding OWL names: Table 1. The renamed class and property between DAML+OIL and OWL DAML+OILOWLdaml:hasClassowl:someValuesFromdaml:toClassowl:allValuesFromdaml:UnambiguousPropertyowl:InverseFunctionalPropertydaml:UniquePropertyowl:FunctionalProperty A new type of property, owl:SymmetricProperty, has been added, which can directly state that properties are symmetric. Qualified cardinality restrictions are not supported in OWL (. This results in the removal of the DAML+OIL properties daml:cardinalityQ, daml:hasClassQ, daml:maxCardinalityQ and daml:minCardinalityQ. This would seem to represent a real reduction in the expressive power of OWL with respect to DAML+OIL. Resources  HYPERLINK http://www.ontoknowledge.org/oil/ http://www.ontoknowledge.org/oil/  HYPERLINK http://www.daml.org/2001/03/daml+oil-index.html http://www.daml.org/2001/03/daml+oil-index.html  HYPERLINK http://www.w3.org/TR/daml+oil-reference http://www.w3.org/TR/daml+oil-reference  HYPERLINK http://www.w3.org/2001/sw/WebOnt/ http://www.w3.org/2001/sw/WebOnt/  HYPERLINK http://www.w3.org/TR/2002/WD-owl-features-20020729/ http://www.w3.org/TR/2002/WD-owl-features-20020729/  HYPERLINK http://www.w3.org/TR/2002/WD-owl-guide-20021104/ http://www.w3.org/TR/2002/WD-owl-guide-20021104/  HYPERLINK http://www.w3.org/TR/2002/WD-owl-ref-20021112/ http://www.w3.org/TR/2002/WD-owl-ref-20021112/  HYPERLINK http://www.w3.org/TR/2002/WD-owl-semantics-20021108/ http://www.w3.org/TR/2002/WD-owl-semantics-20021108/  HYPERLINK http://www.w3.org/TR/2002/WD-owl-test-20021024/ http://www.w3.org/TR/2002/WD-owl-test-20021024/  HYPERLINK http://www.mindswap.org/2002/owl.html http://www.mindswap.org/2002/owl.html  HYPERLINK http://www.daml.org/language/ http://www.daml.org/language/ Summary Since OIL has been developed in Europe by the end of 2000, due to the existing bugs and strange modeling theory which RDF and RDFS are based on. The evolution has been spread out the whole Web area. The joint collaboration between EU and USA has come out with DAML+OIL, which has been significantly used throughout various areas for adding semantics to the knowledge to assist computer to understand and process information more intelligently. Still the current semantic web ontology language are far from what the real life required and the Web is such a heterogeneous environment, the language for such environment must be layered. OWL is the layered language for this environment and currently is still under-going. In a nutshell, the current OWL-Lite roughly consists of RDFS + equality + 0/1-cardinality. 2.5 RuleML Rules for the Web have become to establish their importance since inference rules are identified as a Design Issue of the Semantic Web. Rules are also continued to play an important role in Intelligent Agents and AI shells for knowledge-based systems, which need a Web interchange format. The Rule Markup Initiative has taken initial steps towards defining a shared Rule Markup Language - RuleML, permitting both forward (bottom-up) and backward (top-down) rules in XML for deduction, rewriting, and further inferential-transformational tasks. 2.6 Topic Map and its family Topic Maps provide a formalization of the notion of a back-of-book index. They are been standardized in ISO (ISO/IEC 13250). Topic maps provide a mechanism for describing knowledge structures and associating them with resources. Topic Maps contains three notions: Topic: A topic represents any thing whatsoever a person, entity, concept regardless of whether it exists or has any specific characteristics. It allows the specification of classes of topics thus topics represent both classes and instances. Association: Topic associations represent the relationships between topics. Occurrence: An occurrence of a topic is an information resource that is deemed to be relevant to the topic in some way. Topic maps are strongly connected to notions of indexing in Library and Information Science. They might prove to be a useful mechanism as indexing structures within the Semantic Web. However, the lack of concept forming operators or constructs with detailed semantics suggests that they may not provide an appropriate formalism for the representation of rich ontologies. TopicMaps.Org is an independent consortium for developing Topic Maps. Their work includes the development of version 1.0 of an XML grammar for interchanging Web-based Topic Maps, called XML Topic Maps - XTM Version 1.0. The design idea of Topic Maps are mainly originated from Library and Information Science, where the indexing the structure of knowledge has been existed and applied for many years. In order to index the huge web document, some formalized language is needed. But Topic Maps failed to express much rich ontologies due to the limited functionality for class and property it could provide. 2.7 DAML-S DAML-S is developed by the DAML (DARPA Agent Markup Language) program. It is a DAML+OIL based Web service ontology, which supplies Web service providers with a core set of markup language constructs for describing the properties and capabilities of their Web services in unambiguous, computer-interpretable form. It enables users and software agents to discover, select, invoke, compose and monitor Web resources based on the well-defined ontology. DAML-S has the following features: Automatic Web service discovery: Automatic location of services, which provide a particular or specific service. Automatic Web service invocation: The possibility of automatic execution of an identified Web service by a computer program or an agent. Automatic Web service composition and interoperation: The selection, composition and interoperation of different services automatically integrated to perform a task Automatic Web service monitoring: Trace and monitoring of the services during their execution. DAML-S is an attempt to bring the benefits of the emergent ontology-based information standards of the Semantic Web to web services. That is to bring the notion of formal, explicit, semantics to the description and integration of web services. To that end, DAML-S employs DAML+OIL, description logic-based, web-enabled, and ontological mark-up language to define an upper ontology for web services. DAML-S has a three-level model that corresponds to answering three groups of key questions of web services: What does the service require? This information lets an agent who is seeking for services understand what a service is going to need to do what it says it can do. This information is to be found in the ServiceProfile. How does it work? The information tells an agent how it is going to do what is says it can do. This information is to be found in the ServiceModel How is it used? The information informs an agent of how it should be addressed and communicated with. This information is to be found in the ServiceGrounding. Because DAML-S itself is not a Semantic Web language, rather an application of the Semantic Web language, DAML+OIL or OWL, to define the ontology for the web services, it will not be detailed discussed in this deliverable. If it is needed during the late stage of the project, detailed discussion or comparison of DAML-S will be provided in the forthcoming version of the deliverable. 2.8 Supporting tools There are some tools available to facilitate the wider-use of the above semantic web languages. Deliverable 13 in Ontoweb gives the detailed roadmap of various existing supporting tools, please visit http://www.ontoweb.org/deliverable.htm. COMPARISON Artificial Intelligence has a strong tradition in developing methods, tools and languages for structuring knowledge and information. While meta-data annotation of Web resources is the example for applying AI techniques on a large and successful scale. In essence, the machine-processable representations of the semantics of these resources are the backbones. 3.1 General comparison of modeling primitives The modeling primitives here we would like to stress are factual knowledge data models, terminological knowledge ontologies, and inference knowledge. Factual knowledge: Data Models The data-models underlying the semantic web languages vary: Ontobroker's nonstandard onto-attribute is based on F-logic and therefore has the Flogics rich data model, consisting of classes, attributes with domain and range definitions, is-a hierarchies with set inclusion of subclasses and multiple attribute inheritance. XML takes labelled trees as its basic data-model, therefore XML and XML Schema actually define the labelled tree data structure data syntax. There are several ways to represent the same RDF data model in XML. RDF 's data model is a syntax-neutral way of representing RDF expressions. It is based on binary relations, enhanced with a reification mechanism to enable relations between relations, and statements about the statements. This data model consists of three object types (resources, properties and statements). RDFS uses this basic data model to build a basic object-oriented type schema on top of RDF. SHOEs data-model is similar to that of RDFS. It does not include the reification mechanism of RDF. SHOE allows the specification of classes with attributes, with multiple inheritance of attributes between classes. OIL, DAML+OILs data model is based on description logic and Frame-based logic. Therefore, they have the rich class, property, and axiom to model the world. OWL is OIL and DAML+OIL based. Provides an abstract syntax as a higher level to write ontologies. Onotologies built as sequences of axioms, facts and reference to other ontologies. Terminological knowledge: ontologies Ontologies are shared specification of a conceptualization and the corresponding vocabulary used to describe a domain. Ontologies can define the rich semantics of much complex objects and therefore well-suited for describing heterogeneous, distributed and semi-structured information sources such as things on the Web: In SHOE, ontologies can be defined by information-providers themselves inside their own HTML pages. Such an ontology contains a class-lattice and possible relations between instances of these classes. SHOE allows for local definitions of ontologies (or local extensions of central ontologies). Ontobroker ontologies are similar in nature (a class-hierarchy, attributes with domain and range definitions, and multiple attribute inheritance). Ontobroker relies on a single centrally defined ontology. XML-schema provides the mechanisms for constraining document structure and content; to enable inheritance for element, attribute, and datatype definitions. It also allows creation of user-defined datatypes. RDFS can be used directly to describe an ontology with its Objects, Classes, and Properties. On the one hand, RDFS provides rather limited expressive power. A serious weakness of RDF is that it lacks a standard for describing logical axioms. No intentional definitions or complex relationships via axioms can be defined. On the other hand, RDFS provides a rather strong reification mechanism. RDFS can be used to describe what is called an ontology in SHOE and Ontobroker. OIL, DAML+OIL and OWL are the languages mainly aiming for defining ontologies. They provide much richer constructors for forming complex class expressions and the axioms can be easily defined by these languages. Another important feature of these language is that they are layered language with different expression power on different layer to fulfil different needs to define simple or complex ontologies. Inference Knowledge Here we analyze to which the extent the inferential knowledge can be expressed in the various semantic web languages. We take the class subsumption as a simple example: SHOE can describe such inferential knowledge in pure Horn rules inside local Web pages. Ontobroker states this inferential knowledge centrally. It allows a larger fragment of first-order logic to be used, which can be translated to stratified normal logic programs via the Lloyd-Topor transformations [Lloyd & Topor, 1984]. XSLT (XSL Transformation language) allows expressing transformations of XML structures, and these can be used to express certain inferential knowledge. RuleML permits both forward (bottom-up) and backward (top-down) rules in XML for deduction, rewriting, and further inferential-transformational tasks. RDF/RDFS subclass relation can be used to represent the class subsumption. OIL, DAML+OIL and OWL can define even much complex rules or axioms. Table 2 gives the summary of the comparison on modeling primitives, where the range from - to ++ indicates a range of expressive power in the corresponding category. Table 2: Summary of the modeling primitives factsterminologyinferenceSHOE+++++Ontobroker++++++XML++/--RDF(S)++/--OIL++++++DAML+OIL++++++OWL++++++ 3.2 Specific comparison between two languages Semantic web languages are complementary technological means that will enable ontological support in knowledge management and electronic commerce. XML provides a standard serial syntax for exchanging data. A DTD allows for the definition of structure and elementary tags of an XML document. XSLT allows translating between different XML documents. RDF provides a standard for describing machine-processable semantics of data. Finally, OIL, DAML+OIL and OWL provide much more support to define the complex or heavy weight ontology including the inference. XMLs vs. DTD Main improvements of XML schemas compared to DTDs are: 1) XML schemas definitions are themselves XML documents. 2) XML schemas provide a rich set of datatypes that can be used to define the values of elementary tags. 3) XML schemas provide much richer means for defining nested tags (i.e., tags with subtags). 4) XML schemas provide the namespace mechanism to combine XML documents with heterogeneous vocabulary. RDF vs. XML RDF is an application of XML to representing meta data. For example, the RDF statement can be represented in linear XML syntax. However, RDF provides a standard way on how to represent meta data in XML. Using directly plain XML for representing meta data would result in different syntax. RDFS provides a fixed set of modeling primitives for defining an ontology (classes, resources, properties, is-a and element-of relationship, etc.) and a standard way on how to encode them in XML. OIL vs. XML XML can be used as a serial syntax for OIL, which is very useful because it puts OIL in the mainstream of tools that are currently being developed for supporting XML based applications. Validation and rendering techniques developed for XML can directly be used for ontologies specified in OIL. Central for an ontology is the is-a relationship, and XML schemas incorporate the notion of inheritance. [Klein et al., 2000] discussed a more complex translation procedure that leads XML documents to capture the semantics of an ontology in OIL by using type refinement in XML schemas to model the subsumption between concepts in OIL. OIL vs. XMLs XML schemas and OIL have one main goal in common, which is to provide vocabulary and structure for exchanging information sources. XML schemas and OIL both have the XML syntax. The XML syntax of OIL is useful for supporting the exchange of ontologies specified in OIL (see OIL vs. XML section). XML schemas have rich datatypes while OIL not. XML schemas provide a much richer set of basic datatypes, for example, string, boolean, float, decimal, timeInstant, binary, etc. OIL does not provide these built-in datatypes because reasoning with concrete domains quickly becomes undecidable or at least inefficient. XML schemas do not worry about this aspect because all inheritance needs to be defined explicitly. XML schemas do not support intentional definition of types while OIL does. OIL provides an explicit language for the intentional definition of types that is completely lacking in XML schemas. XML provides structures: elements, while OIL not. In XML schemas main modeling primitives are elements. Elements may be simple, composed or mixed, and elements can have attributes. While the modeling primitives of OIL are concepts and slots. Concepts can be roughly identical with elements and child elements are roughly equivalent to slots defined for a concept. However, slots defined globally have no equivalent in XML schemas. The reduction step has to be taken when transforming OIL specifications into XML schema definitions. XML schema provides structures: grammar, while OIL not. OIL does not provide any grammar for composing concepts from slots, i.e., an instance of a concept is a set of slot values. XML schemas allow defining stronger requirements via a grammar: sequence and choice of attributes applied to an instance can be defined. XML schema provides structures: type-derivation, while OIL not. XML schemas incorporate the notion of type-derivation. However, this can only partially be compared with what is provided with inheritance in OIL. All inheritance has to be modeled explicitly in XML schemas. While, inheritance can be derived from the definitions of the concepts in OIL. XML schema and OIL provide namespaces. XML schemas and OIL both provide the same (limited) means for composing specifications. OIL relates to XML schemas like the Extended Entity Relationship model relates to the relational model. On the one hand, OIL provides much richer modeling primitives. It distinguishes classes and slots and class (or slot) definitions can be used to derive the hierarchy (and their corresponding inheritance). On the other hand, XML schemas provide richer modeling primitives concerning the variety of built-in datatypes and the grammar for structuring the content of elements. Models in OIL can be viewed as a high level description that become further refined when aiming for a document structure model. RDFS vs. XMLs RDF Schema enriches the limited modeling primitives defined in RDF by adding, for example, class, subclass relationship, domain and range restrictions for property, and subproperty. With these extensions, RDF Schema comes closer to the existing ontology languages, while XML Schema not. XML Schema and DTDs prescribe the order and combination of tags in an XML document, in another world, the syntax or structure of information. In contrast, RDF Schema only provides information about the interpretation of the statements given in a RDF data model, but it does not constrain the syntactical presentation of a RDF description. OIL vs. RDF(s) One of the syntax of OIL is based on RDFS. To ensure maximal compatibility with existing RDF/RDFS-applications and vocabularies, the integration of OIL with the resources defined in RDF-schema is very essential: The abstract OIL class OntologyExpression is a subclass of rdfs:Resource and the abstract OIL class OntologyConstraint is a subclass of rdfs:ConstraintResource. OIL-slots are realized as instances of rdf:Property or of subproperties of the original rdf:Property. The subslot relationship can be expressed by the rdfs:subPropertyOf relationship. rdf:Property is enriched by OIL with a number of properties that specify inverse and transitive roles and cardinality constraints, which are not possible in RDF/RDFS. OIL uses the existing primitives of RDFS as much as possible to retain an unambiguous mapping between the original OIL specification and its RDFS serialization. Therefore, the RDFS contained in the definition of domain ontology in OIL can be easily understand or interpreted by any non-OIL-aware RDFS applications, while OIL-aware applications can be benefited by the added features of formal semantics and reasoning support. In a nutshell, any valid OIL document is also a valid RDF(S) document when all the elements from the OIL-namespace are ignored. In order to allow users to choose the expressive power appropriate to their application, and to allow for future extensions, a layered family of OIL languages has been described. The sub-language OIL Core has been defined to be exactly the part of OIL that coincides with RDF(S). DAML+OIL vs. RDFS DAML+OIL is tightly integrated with RDFS by using RDFS to express the syntax of DAML+OIL. Therefore, the existing RDFS infrastructure can be easily reused and ontologies defined by DAML+OIL can be partially or fully compatible with those defined by RDFS. On the other hand, DAML+OIL also inheritates the strange modeling idea of RDFS, such as restrictions with multiple properties and classes. DAML+OILs dependence on RDFS also leads to the consequence of the decidability of the language. Decidability is lost when cardinality constraints can be applied to properties that are transitive, or that have transitive sub-properties. So decidability in DAML+OIL depends on an informal prohibition of cardinality constraints on non-simple properties. DAML+OIL vs. OIL From the point of view of language constructs, the differences between OIL and DAML+OIL are rather trivial. There usually exists a one to one mapping of constructors between these two languages. In the cases where the constructors are not completely equivalent, simple translations are possible. OIL also uses RDFS for its serialisation (although it also provides a separate XML-based syntax). OILs RDFS based syntax might cause it to face the same problems as those for DAML+OIL. However, OIL processors would reject ontologies containing unexpected use of the meta-properties. Therefore, not any ontology conforming to the RDFS meta-description will be a valid OIL ontology. OIL and DAML+OIL take rather different positions with regard to the layering of languages on the Semantic Web. The tight integration of DAML+OIL with RDFS leads to the loose of the frame structure of OIL syntax. A DAML+OIL ontology is more DL-like in that it consists largely of a relatively unstructured collection of subsumptions and equality axioms. This makes difficult to use DAML+OIL with frame-based tools such as Protg (Grosso et al. 1999) or DL based ones as OilEd (Bechhofer et al. 2001). The treatment of individuals in OIL is also very different from that in DAML+OIL. DAML+OIL treats individuals occurring in the ontology (in oneOf constructs or hasValue restrictions) as true individuals (i.e., interpreted as single elements in the domain of discourse) and not as primitive concepts as in OIL. This weak treatment of the oneOf construct is a well known technique for avoiding the reasoning problems that arise with existentially defined classes, and is also used, e.g., in the CLASSIC knowledge representation system (Borgida & Patel-Schneider 1994). Moreover, DAML+OIL makes no unique name assumption: it is possible to explicitly assert that two individuals are the same or different, or to leave their relationship unspecified. DAML+OIL relies wholly on RDF for assertions involving the type (class) of an individual or a relationship between a pair of objects. DAML+OIL vs. OWL The difference between DAML+OIL and OWL is rather trivial as well. But OWL abstract syntax has reverted to grouping axioms into frame like structure, which makes it easy to use the frame-based tools such as Protg or DL based ones OilEd. In this sense, OWL is closer to OIL due to the frame-based feature while DAML+OIL is more DL-like. Because OWL is still a very young semantic web ontology language (less than half-year old), more development will be continued in this area. Significant works are going on in the Web Ontology Working Group of W3C. We will keep the sharp eye on its evolution and report the changes in our later deliverables. ANALYSIS In this section, we analyze the existing semantic web ontology languages in more details based on the latest requirement raised in Web Ontology Working Group of W3C and some features we presume are important to report here. Wish lists for Semantic Web Ontology Languages Heflin, Volz and Dale (2002) proposed the wish lists for OWL based on various case studies, which can be summarized into eight design goals of a Web Ontology language: The first is that it should provide for ontology sharing, including the ability of one public ontology to extend another public ontology by, perhaps, specializing some of its terms or properties. The second design goal is to support the change or drift over time of ontologies and their constitutive parts (ontology versioning). As knowledge domains evolve, ontologies must be able to evolve in order to continue to formally represent their domain. Third, it must support interoperability, by offering some way to map similarities between disparate ontologies. The fourth design goal is to provide the possibility to detect the inconsistencies among different ontologies or data sources (ontology reasoning support). Fifth, it should strike a balance between expressiveness and the scalability. K The sixth is that it should be easy to use in syntax and semantics. Seventh, it should be compatible with other Web and industry standards. In particular, this includes XML and related standards (such as XML Schema and RDF) (Compatibility with other standards). Eighth, it should support the development of multilingual ontologies, and potentially provide different views of ontologies that are appropriate for different cultures (Internationalization) Table 3 summarizes the features provided by these semantic web ontology languages according to these 8 wishlists. The details are: Ontology Sharing: It is mainly maintained via the reasoning support. XSLT could define the translation rules among different ontologies defined in XML. It is not surprise to see that OIL, DAML+OIL and OWL could partially support this function due to the powerful reasoning service they equipped. Actually this ontology sharing is not clearly stated in the wishlist. Is this the partial ontology mapping or two ontologies have the simple parent-child relationship therefore it will be easily extended or shrink by just broadening or narrowing some terms or relations. Ontology versioning: It is very bad to see that almost none of the existing semantic web languages support such functions, except the early development of SHOE. Clearly the versioning is very much wanted for the next generation of Web. Therefore the later-on developed semantic web language, should be capable to bring such function in. Efforts are already started and going-on (Klein et al., 2002), and as a result of these efforts OWL now contains some versioning primitives. Interoperability: this overlaps the Ontology sharing, or Ontology sharing is the subset of Interoperability. Same to see here, it is the role of the reasoning. Partial mapping can be supported by defining the mapping rules between two ontologies. Reasoning: Reasoning is the main concern of the design of the semantic web language. Therefore RuleML is split out to become the independent initiative. But the balance between the power of reasoning and the complexity of the computing should be significantly considered here. Due to the special feature of the web, simple reasoning function is desirable for the reality. Expressive & Scalability: This is the common problem similar as the above. The more expressiveness it has, the less scalability it could have. How to balance in between also depends on the different user applications. Therefore the layered structure of the language is very important. Easy to use: The language must be simple and easy access if it is to have a realistic chance of widespread acceptance. Tools need to be provided to assist the ease of use of the language. The clear and formal documentation for the use of the language is important, which can contribute in a major fashion to the reusability and ease of understanding of code. Fail to do this means the dead-end of the language. Compatibility: The reuse of the knowledge is stressed almost everywhere, especially the language tower visioned by Tim Bernees-Lee, which is established on such compatibility among different layer of the languages. It is nice to see that most of the existing semantic web languages have the compatibility with the existing related languages. Table 3. Comparison of ontology languages based on the W3C wishlists LanguageSHOEOntobrokerXML(s)RDF(s)OILDAML+OILOWL sharing No No XSLT can define different views NoReasoning can be used as the back support for integrating and sharing ontologies. Reasoning is useful for ontology sharing. Reasoning is useful for ontology sharing. versioningProviding ontology versioning function Maintaining each version of ontology as a separate web page and each instance must state the version to which it adheres. No No No No No Yes interoperability No Horn-Logic can define the mapping rules XSLT can define translation rules between two ontologies Partially: mapping rules can be defined Partially: mapping rules can be defined  Partially: mapping rules can be defined Partially: mapping rules can be defined reasoning Reasoning supports are provided to handle revision problems Inference engine is used to derive answersRuleML permits both forward (bottom-up) and backward (top-down) rules in XML for deduction, rewriting, and further inferential-transformational tasks.  Provided limited reasoning mechanism mainly for constraint checking. Automatic consistency checking Cross-linking the inter-ontology relationships and check for the implied relationships. Supporting design, maintenance and deployment of ontologies, especially the reasoning support from Description Logics  Based on an open world assumption (OWA). Equipped the same reasoning power from DAML+OIL expressive & scalabilityLimited expressive powerMuch richer expressivenessExpressive power in syntaxLimited expressive power in semantics. Lack of the necessary rich concept forming operatorsMuch rich expressive power in semantics. Layered feature for scalabilityMuch rich expressive power in semantics. Much rich expressive power in semantics. Layered feature for scalability Easy to use Yes Yes YesNot easy due to the strange modeling theory Yes Yes Yes compatibility It is a superset of HTML, which adds the tags to semantic data. There is also an XML-based version of SHOE. Annotation language is an extension of HTML. Representation language is based on Frame logic Query language is based on Frame logic Compatible with HTML with the added value that the users can define their own tags. Syntax is defined in XML Designed on Description logic, Frame logics, and web standards (RDF(s) and XML(s)). Core OIL coincides with RDF Schema (except the reification features of RDF Schema).The merge of DAML-ONT and OIL Based on existing Web standards (XML and RDF), Adding object and frame based systems, and description logic. Supporting the full range of XML Schema datatypes Based on OIL and DAML+OIL OWL-Lite roughly consists of RDFS plus equality and 0/1-cardinality. DAML+OIL style RDF syntaxinternationalizationSupport different natural languagesSupport different natural languagesSupport different natural languagesSupport different natural languagesSupport different natural languagesSupport different natural languagesSupport different natural languages Table 4. Detailed language feature comparison LanguageSyntaxSemanticsLayered structureRedundancyCompatibilitySpecial featuresTool support SHOE HTMLExtended HTML tags for defining ontology NoYes. Mixing ontology and data. HTML Versioning support,  SHOE tools OntobrokerNot available due to commeralizationExtended HTML tags for defining ontology.Not available due to commeralizationNo. Ontology and data are separatedHTML Description Logic Frame-based logic Reasoning supportOntobroker tools: webcrawler and inference engine XML(s) XMLThe structure and semantics of document are interwoven. No No. Ontology and data are separated XMLXML schemas provide a rich set of datatypesVarious XML-related tools: editor, parser and so on. RDF (s) XML (Schema) RDF (Schema)  RDFS define the semantics No No. Ontology and data are separated XML XML schema Too strong reification  Various RDF-related tools: editor, parsers. OIL XML (Schema) RDF (Schema) OIL  Rich expressiveness for defining ontologiesYes OIL-Core OIL-Lite OIL-Standard OIL- instance OIL-heavy No. Ontology and data are separatedXML (Schema) RDF (Schema) Description Logic Frame-based logic Description Logic and Frame-based LogicVarious OIL related tools: editors, parsers. DAML+OILXML (Schema) RDF (Schema) DAML+OILRich expressiveness for defining ontologiesNoNo. Ontology and data are separatedXML (Schema) RDF (Schema) OIL DL based.Various DAML+OIL related tools: editors, parsers. OWLXML (Schema) RDF (Schema) Rich expressiveness for defining ontologiesYes. OWL-Lite OWL-DL OWL-FullNo. Ontology and data are separatedXML (Schema) RDF (Schema) OIL DAML+OILSimilar to DAML+OIL but with Frame-based Validator Table 5. Strength and weakness of various languages LanguageStrengthWeaknessSHOESupporting ontology versioning . Redundancy is generatedOntobrokerReal semantic web prototype with annotation, query and inference languages support. Due to commercialization, many info is not availableXML(s)Rich datatype XSLT defines the rules for transferring different ontologiesLabel tree Define the syntax RDF (s)Define semantics Strange modeling theory: cyclical definition is not allowed (improved by latest development) multiple domain and range are not allowed OILLayered languages Rich inference support (description logic) Frame-based XML(s) and RDF(s) compatibleInheritance some of the strange features of RDFSDAML+OILDL-like XML(s) and RDF(s) compatibleNot frame-based  OWLLayered languages Rich inference support (description logic) Frame-based XML(s) and RDF(s) compatibleUnder development Strength & Weakness In this part, we try to put all these languages together to compare the strength or weakness of each based on some criteria, which are foreseen to be important for our Esperonto project. Table 4 summarizes the comparison as the followings: Syntax: All the latest developed languages are XML or RDF syntax based. This significantly enhances the compatibility among the languages. The existing tools for support XML or RDF can be easily reused for OIL, DAML+OIL and OWL. The layered language tower can be made possible due to the compatible syntax. Semantics: Adding semantics to the existing information is the main goal for the semantic web languages. But how to add and represent clear, explicit, machine-understandable semantics is not a trivial task. It is a clear go-direction for the design of languages, from SHOE to OWL. Different semantics needs different expressive power, therefore the layered structure is very essential as well. Layered structure: There will not be one single language, which fulfill all the needs of various web users. In a layered design, a simple core can accommodate simple taxonomies and relationships, while additional layers of expressivity, functionality, and complexity can be added for groups requiring more expressive power. Also the scalability and maintenance burden can be distributed to the different layers of the language, therefore can be alleviated accordingly. The languages with such clear layered structures are OIL and OWL. Compatibility: Please see the above section. Special features: SHOE has very basic versioning support, while Ontobroker provides efficient broker functions. XMLs defines very rich datatypes. RDFS has very strange modeling theory which brings the undeciability and too strong reification. OIL is frame-based, while DAML+OIL is not. OWL is much closer to OIL due to the frame-based feature. Tools Support: Software tools need to be developed for ontology language to really take off. Such tools should include the assist for user to easily create, use and maintain the ontologies. These tools should also provide user-friendly import and export facilities to link with other standard languages. Almost every language has various tool supports, ranging from editors, parsers up to reasoners. The summary of the strength and weakness of the languages are shown in Table 5. Other Analytical Dimensions Web ontology languages have a number of dimensions along which we can evaluate and measure their appropriateness. Here in this deliverable, the comparison or evaluation is situated in the macro-level of the language structure. The much down-to-earth comparison can be carried out to the detailed functions of the constructors of the languages, such as: negation; conjunction; disjunction; explicit universal or existential quantification/restriction; multi-valued slots; role properties such as transitivity, symmetry, reflexivity, functionality or the support of inverses; role hierarchies; number or cardinality restrictions on roles; both necessary and sufficient conditions for class membership; constraints. If the language supports constraints, how rich is the constraint language? Is the constraint language formally defined? rules; recursion; relations; multiple inheritance; axioms. Are facilities provided for the definition of axioms? If so, how expressive can those axioms be? template/default values; method slots (calculated values); Data types Instances. Does the language provide support for the encoding of instances as well as classes? Many difficult problems remain to be tackled trust, time, and identity, to name but a few. However, an ontology language that is powerful enough to capture most desired information, while still being simple enough to allow agents to perform inferences based on the facts at their disposal at any particular moment, will represent a major challenge to the Semantic Web development. CONCLUSION AND FUTURE PLAN The Web ontology language must be able to describe and organize knowledge in the Web in the machine understandable way. It is a very complicated task, which can be simplified as defining categories by creating concepts, or frames, or classes, rules and so on. However, taken into the extreme, organizing knowledge requires the facilities of the logical formalism, which can deal with temporal, spatial, epistemic, and inferential aspects of knowledge. Implementations of the Web ontology language must provide these inference services, making them much more than just simple data storage and retrieval systems. The Web ontology language must have the clear model theory to secure the precise determination, while well model-theoretic semantics is really nothing more than a generalization of data models, like the relational data model or semi-structured data, that can deal with identifiable objects and uncertain and vague information. Tools are needed for the massive adoption of the languages themselves, therefore the easy-of-user for the programmers to implement all the essential features of the web ontology languages into the supporting tools are very important as well. After the detailed the comparison and analysis within and between different web ontology languages in previous sections. Here we will give the summary and mention our future plan. Ciminos dream for web ontology language Cimino (1998) presents a collection of desiderata for Controlled Medical Vocabularies in the 21st Century. Some of them are quite interesting to state here for the design of Web ontology language: Content. Formal methodologies are required to support the addition of content. Concept Orientation. Terms in the ontology should correspond to at least one meaning (nonvagueness) and no more than one meaning (nonambiguity). In addition, meanings should correspond to no more than one term. Concept Permanence. Ontologies may evolve, but once a concept has been created, its meaning should be inviolate. This is not to say that concepts cannot be refined or redescribed. But unnecessary altering the means of concepts should be prevented. This eases the method for ontology management and versioning Non-semantic Concept Identifier. Unique identifiers should be provided for concepts. Poly-hierarchy. Multiple inheritance in subsumption hierarchies should be supported. Formal Definitions. This is very important for making machine understand the semantics of the data therefore reasoning or inference can be carried out by agents. Multiple Granularities. Vocabularies should be capable of supporting multiple levels of granularities. The layered structure for web ontology language is rather essential than normal because of the heterogeneous usage of the Web. Multiple Consistent Views. In tandem with a requirement for multiple granularities is the requirement for multiple views. Evolve Gracefully. Detailed descriptions of change are required in order to support the maintaining of the ontologies. Formal methodology on ontology versioning should be provided within the web ontology language itself. Recognize Redundancy. The redundancy will significantly reduce the power of reasoning and increase the complexity of maintenment. Proper redundancy is also necessary for provide difference views. The mechanism to recognize and balance the redundancy should be considered for designing the web ontology language. Many of the above desiderata can be seen to resonate with the requirements put forward by the WebOnt committee. Early approaches for the semantic web are based on rule-based approaches. Although there is a clear consensus that the current Web Ontology Language needs an extension in this direction. The Rule Markup Language Initiative may help to make a step into this direction. Giving a real semantics to the semantic web language tower as sketched by Tim Berners-Lee in Figure 1 requires much more work. Currently many layering ideas oriented to syntactical and semantic extensions compete which each other. Working out the proper relationship will be much more challenging than just developing one layer for it. The easy information access based on the success of the web has made it increasingly difficult to find, present, and maintain the information required by a wide variety of users. In response to this problem, many new research initiatives and commercial enterprises have been set up to enrich available information with machine understandable semantics. This semantic web will provide intelligent access to heterogeneous, distributed information, enabling software products to mediate between user needs and the information sources available. Web Services tackle with an orthogonal limitation of the current web. It is mainly a collection of information but does not yet provide support in processing this information, i.e., in using the computer as a computational device. Recent efforts around UDDI, WSDL, and SOAP try to lift the web to a new level of service. Software programs can be accessed and executed via the web. However, all these service descriptions are based on semi-formal natural language descriptions. Therefore, the human programmer need be kept in the loop and scalability as well as economy of web services are limited. Bringing them to their full potential requires their combination with semantic web technology. It will provide mechanization in service identification, configuration, comparison, and combination. Semantic Web enabled Web Services have the potential to change our life in a much higher degree as the current web already did (Figure 2). We are currently in the process of extending UPML [Fensel et al, 2003] to a full-fledged Web Service Modeling Language (WSML). In the continuing version of the deliverables in the Workpackage, we will keep the sharp eye on the newly development and evolving in the semantic web ontology language area and document and discuss the pros and cons to provide the significant support for the Esperonto tool development. REFERENCES Bechhofer, S.; Goble, C.; and Horrocks, I. (2001). DAML+OIL is not enough. In Proc. of the First Semantic Web Working Symposium (SWWS01), 151159. CEUR Electronic Workshop Proceedings, http://ceur-ws.org/. Berners-Lee, T. (1999). Weaving the Web. San Francisco: Harper. Borgida, A., and Patel-Schneider, P. F. (1994). A semantics and complete algorithm for subsumption in the CLASSIC description logic. J. of Artificial Intelligence Research 1:277308. Calvanese, D.; De Giacomo, G.; Lenzerini, M.; Nardi, D.; and Rosati, R. (1998). Information integration: Conceptual modeling and reasoning support. In Proc. of CoopIS98, 280291. Cimino, J. J. (1998). Desiderata for Controlled Medical Vocabularies in the Twenty-First Century. Methods Inform Med 1998; 37: 394-403. Gmez-Prez, A. and Corcho, O. (2002). Ontology Languages for the Semantic Web. IEEE Intelligent Systems, January/February Fensel, D.; Angele, J.; Decker, S.; Erdmann, M.; Schnurr, H.-P; Staab, S.; Studer, R. and Witt; A. (1999): On2broker: Semantic-Based Access to Information Sources at the WWW. In Proceedings of the World Conference on the WWW and Internet (WebNet 99), Honolulu, Hawaii, USA, October 25-30, 1999. Fensel, D.; van Harmelen, F.; Horrocks, I.; McGuinness, D. L. and Patel-Schneider, P. F. (2001). OIL: An ontology infrastructure for the semantic web. IEEE Intelligent Systems 16(2): 3845. Fensel, D.; Motta, E.; van Harmelen, F.; Benjamins, V. R.; Crubezy, M.; Decker, S.; Gaspari, M.; Groenboom, R.; Grosso, W.; Musen, M.; Plaza, E.; Schreiber, G.; Studer, R. and Wielinga, B (2003): The Unified Problem-solving Method Development Language UPML, Knowledge and Information Systems (KAIS): An International Journal, 5(1), 2003. Grosso, W. E.; Eriksson, H.; Fergerson, R. W.; Gennari, J. H.; Tu, S. W.; and Musen, M. A. (1999). Knowledge modelling at the millenium (the design and evolution of protege-2000). In Proc. Of Knowledge acqusition workshop (KAW-99). Heflin, J. and Hendler, J. (2000). Searching the Web with SHOE. In AAAI-2000 Workshop on AI for Web Search. Heflin, J. and Hendler, J. (2001). A Portrait of the Semantic Web in Action. IEEE Intelligent Systems, 16(2). Heflin, J. and Hendler, J. and Luke, S. (1999). SHOE: A Knowledge Representation Language for Internet Applications. Technical Report CS-TR-4078 (UMIACS TR-99-71). Heflin, J.; Volz, R. and Dale, J. (2002). Requirements for a Web Ontology Language. W3C Working Draft 08 July 2002. Horrocks, I. 1998. Using an expressive description logic: FaCT or fiction? In Proc. of KR-98, 636647. Kifer, M.; Lausen, G. and Wu, J.(1995). Logical Foundations of Object-Oriented and Frame-Based Languages, Journal of the ACM, 42. M. Klein, D. Fensel, F. van Harmelen, and I. Horrocks: The relation between ontologies and schema-languages: Translating OIL-specifications in XML-Schema. In Proceedings of the Workshop on Applications of Ontologies and Problem-solving Methods, 14th European Conference on Artificial Intelligence ECAI'00, Berlin, Germany August 20-25, 2000. Klein, M.; Kiryakov, A.; Ognyanov, D and Fensel, D. (2002). Ontology Versioning and Change Detection on the Web. In 13th International Conference on Knowledge Engineering and Knowledge Management (EKAW02), Sigenza, Spain, October 1-4, 2002. Lloyd, J. W. and Topor, R. W (1984). Making Prolog more Expressive, Journal of Logic Programming, 3:225-240. Luke, S.; Spector, L.; Rager, D. and Hendler, J. (1997). Ontology-based Web Agents. In Proceedings of First International Conference on Autonomous Agents 1997. Nejdl, W.; Wolpers, M. and Capella, C. (2000). The RDF Schema Revisited. In Ebert J. et al. (eds.), Modelle und Modellierungssprachen in Informatik und Wirtschaftsinformatik, Modellierung 2000, St. Goar, April 5-7, 2000. QL (1998). Proceedings of W3C Query Language Workshop (QL98) - The Query Languages Workshop, Boston, Massachussets, December 3-4, 1998. http://www.w3.org/TandS/QL/ QL98/. Appendix 1 Since the OWL specification is constantly evolving a link to the OWL Lite expanded summary is provided:  HYPERLINK "http://www.w3.org/TR/owl-features/" http://www.w3.org/TR/owl-features/  http://www.cs.umd.edu/projects/plus/SHOE/  http://www.aifb.uni-karlsruhe.de/www-broker  http://www.w3.org/XML  http://www.w3.org/RDF  http://www.w3.org/TR/REC-rdf-syntax; http://www.w3.org/TR/rdf-mt/  http://www.w3.org/TR/rdf-schema/  www.ontoknowledge.org/oil  http://www.cs.man.ac.uk/~horrocks/FaCT/  http://www.daml.org/language/  http://www.daml.org/committee/  http://www.daml.org  http://www.daml.org/2001/03/model-theoretic-semantics  OWL Web Ontology Language Abstract Syntax This document gives a high level description of the language features in an abstract syntactic form similar to that of OIL ( HYPERLINK http://www.w3.org/TR/2002/WD-owl-absyn-20020729/) http://www.w3.org/TR/2002/WD-owl-absyn-20020729/).  OWL Web Ontology Language Reference This document gives a detailed specification of the RDF syntax of the language, including usage examples ( HYPERLINK http://www.w3.org/TR/2002/WD-owl-ref-20020729/) http://www.w3.org/TR/2002/WD-owl-ref-20020729/).  http://www.dfki.uni-kl.de/ruleml  http://www.y12.doe.gov/sgml/sc34/document/0129.pdf  www.topicsmap.org  http://www.doctypes.org/xtm/home.html  http://www.daml.org/services/  http://www.w3.org/Style/XSL  http://www.dfki.uni-kl.de/ruleml/ IST Project IST-2001-34373 Esperonto Esperonto Services Deliverable -  PAGE 43 - OIL Description Logic: Formal Semantics & Reasoning support Web languages: XML- and RDF-based syntax Frame-based systems: Epistemological Modeling Primitives '(*,BCDEKwxy  ƿypgpgcccZZhI6chmH sH h hI6cmHnHuhKmHnHu hI6cCJhghhmH sH  hhhhI6cB*CJmHnHphu*jhI6cB*CJ(UhmHnHphu h*hI6ch*hI6cCJ$OJQJ#jhI6cB*UmHnHphujhI6cUmHnHujhI6cUhI6c#%&')*-.:;<@ABDEwyz{$ 1 d^a$$a$TSgd*$a$|Ɂ'ʂ        7d1$<< 7d<<1$R$a$-gdSgd*  }   ? A M N S U i k l n     ÿ鱬 jphI6c *hI6c jhI6c jhI6ch hi4h@ hI6c0JjhI6cUjhI6cU hI6c5hI6chI6chmH sH hWshmH sH @ | Ykd@$$If4\3}"9!OJ 04 af4 P$$Ifa$Sgd* | } y P$  $Ifa$ P$$Ifa$okd$$If4039!O04 af4 y P$  $Ifa$ P$$Ifa$okd $$If4039!O04 af4 y P$  $Ifa$ P$$Ifa$okdF $$If4039!O04 af4    @ P$$Ifa$okd $$If4039!O04 af4@ A F M N S T TKKKKK P$$Ifa$kd $$Ifr3 9!OS70U!4 aT U b i k l m TKKKKK P$$Ifa$kd $$Ifr3 9!OS70U!4 am n  TKKKKK P$$Ifa$kdl $$Ifr3 9!OS70U!4 a TKKK P$$Ifa$kdI$$Ifr3 b9!O 0U!4 a  zqq P$$Ifa$kd&$$IfF39!Oj 0U!    4 a   ; P$$Ifa$qkd$$If039!O0U!4 a ! - . 9 : ; < N W ] l r /  %&'/0?@IJKUVdekrst~߼ h"hI6c h"h"hI6c5B*phh"h!jhhI6cUjhI6cUhqRhi4 hI6c5 *hI6c hI6c6 jphAhI6c jhA@; < N P$$Ifa$qkdr$$If039!O0U!4 a P$$Ifa$qkd $$If039!O0U!4 a ulllllll P$$Ifa$kd$$If4JF39!`O*0U!    4 af4/) P$$Ifa$kd'$$If4J֞3j ]9! O720U!4 af4 P$$Ifa$qkd?$$If039!O0U!4 a~~P$If P$$Ifa$qkd$$If039!O0U!4 a &zqkkP$If P$$Ifa$kdu$$IfF369!O  0U!    4 a&'0@JzqkkP$If P$$Ifa$kd$$IfF369!O  0U!    4 aJKVeszqkkP$If P$$Ifa$kd$$IfF369!O  0U!    4 astzqkkP$If P$$Ifa$kd^$$IfF369!O  0U!    4 aCELpq3UXY|}~)lԼܧԜԄԜyodhbhI6cmH sH j]hI6cCJUhhI6cmHsHh@hI6c0JmH sH jp\hI6cUh@hI6cmH sH hhI6cmHsHj/hI6cCJUhhI6c0JmH sH jU.hI6cUjhI6cUhhI6cmH sH hI6cH*jUhI6cCJUhI6c hI6c5"zxvxmggP$If P$$Ifa$Rkd$$IfF369!O  0U!    4 a)Lmvpppggggg P$$Ifa$P$If P$$Ifa$kd$$IfF@ d9!\ $  0    4 a 3UvppvvvvvP$If P$$Ifa$kd /$$IfF@ d9!\ $  0    4 a  )ARlvppggggggg P$$Ifa$P$If P$$Ifa$kd/]$$IfF@ d9!\ $  0    4 a lopVYZƼƱtlƦaj hI6cUjhI6cUj|^A hI6cUVmH sH hShI6cmH sH hhI6c0JmHsHjmhI6cUhhI6cmHsHhhI6cmHsHjhI6cCJUhI6chbhI6cmH sH h@hI6c0JmHsHjkhI6cUjhI6cUh@hI6cmHsH&!6Vvppgggggg P$$Ifa$P$If P$$Ifa$kdR$$IfF@ d9!\ $  0    4 a !AYvvppggggggg P$$Ifa$P$If P$$Ifa$kdT$$IfF@ d9!\ $  0    4 a +6:InxvmggP$If P$$Ifa$kdj $$IfF@ d9!\ $  0    4 a P$$Ifa$ 6789::;lmnEWXϷ׬ϑyϑoddh@hI6cmHsHjkhI6cCJUh@hI6c0JmH sH jihI6cUh@hI6cmH sH jQhI6cCJU hI6cCJ hbhI6cmH sH hhI6c0JmH sH j#PhI6cUjhI6cUhhI6cmH sH hI6cCJj*hI6cCJUj hI6cCJUhI6c#3:mo}Dvpppgvgggg P$$Ifa$P$If P$$Ifa$kd Q$$IfF@ d9!\ $  0    4 a DENWYnvppgggggg P$$Ifa$P$If P$$Ifa$kdj$$IfF@ d9!\ $  0    4 a '(679:FGHabcdefhi~¾q`qqqOq jzh`k~UmHnHu jyh`k~UmHnHujh^UmHnHuh^mHnHu-h^h^5;CJaJmHnHsH tH uh=h^OJQJmHnHu hI6c:jhI6c:UhI6c h5hI6chbhI6cmH sH h@hI6cmHsHh@hI6c0JmHsHjhI6cUjwhI6cU'f <j4}vvpjjjpjjj0 h / h . h Rkdx$$IfF@ d9!\ $  0    4 a    678:;<IJK誜ueTue jv{h`k~UmHnHuh=h^5mHnHtH u*h^h^6CJaJmHnHsH tH u jzh`k~UmHnHuh=h^5mHnHu*h^h^:CJaJmHnHsH tH u jzh`k~UmHnHu-h^h^5;CJaJmHnHsH tH ujh^UmHnHuh^mHnHu KdefhijךxhW jb}h`k~UmHnHuh=h^5mHnHtH u j|h`k~UmHnHu!h=h^5hmHnHtH u*h^h^:CJaJmHnHsH tH u jl|h`k~UmHnHu*h^h^6CJaJmHnHsH tH ujh^UmHnHu j{h`k~UmHnHuh^mHnHu"./0234KLefgijktuvyiX jNh`k~UmHnHuh=h^5mHnHtH u j~h`k~UmHnHu!h=h^5hmHnHtH u*h^h^:CJaJmHnHsH tH u jX~h`k~UmHnHu*h^h^6CJaJmHnHsH tH u j}h`k~UmHnHuh^mHnHujh^UmHnHu"4k0_Q Q,XEv. h 0 h / h *+,./0>?@YZ[]^_hij´|k j:h`k~UmHnHu jh`k~UmHnHu*h^h^6CJaJmHnHsH tH u jDh`k~UmHnHuh=h^5mHnHu*h^h^:CJaJmHnHsH tH u jh`k~UmHnHuh^mHnHujh^UmHnHu%12KLMOPQST^_`yرؠ؏h-h^h^5;CJaJmHnHsH tH uh=h^OJQJmHnHu j&h`k~UmHnHu jh`k~UmHnHu j0h`k~UmHnHu*h^h^:CJaJmHnHsH tH uh^mHnHujh^UmHnHu jh`k~UmHnHu#yz{}~   012Kؠ|kU|*h^h^6CJaJmHnHsH tH u jh`k~UmHnHuh=h^5mHnHu*h^h^:CJaJmHnHsH tH u jh`k~UmHnHuh=h^hmHnHtH u-h^h^5;CJaJmHnHsH tH uh^mHnHujh^UmHnHu jh`k~UmHnHuKLMOPQdef   ´أ“؂l[´ jh`k~UmHnHu*h^h^:CJaJmHnHsH tH u jh`k~UmHnHuh=h^hmHnHtH u jh`k~UmHnHuh=h^5mHnHu*h^h^6CJaJmHnHsH tH uh^mHnHujh^UmHnHu jh`k~UmHnHu# &'(*+,789RSTVWXdefp_!h=h^5hmHnHtH u joh`k~UmHnHu jh`k~UmHnHuh=h^5mHnHtH u jyh`k~UmHnHuh=h^5mHnHu*h^h^6CJaJmHnHsH tH ujh^UmHnHu jh`k~UmHnHuh^mHnHu%$%&?@ACDEUVWpqrtuvx´أ”؃´rbh=h^OJQJmHnHu j[h`k~UmHnHu jh`k~UmHnHuh=h^5hmHnHu jeh`k~UmHnHuh=h^5mHnHu*h^h^6CJaJmHnHsH tH uh^mHnHujh^UmHnHu jh`k~UmHnHu!xy      0 1 2 4 5 6 ϰϟώ}i'h^5;CJaJmHnHsH tH u jGh`k~UmHnHu j̋h`k~UmHnHu jQh`k~UmHnHu j֊h`k~UmHnHujh^UmHnHuh^mHnHuh=h^OJQJmHnHu-h^h^5;CJaJmHnHsH tH u' 6 8 F G H I ""2$3$m'n'[)\)+++++++++$a$  7d1$ & Ff. h . h 6 7 8 F I t!z!"2$3$E'l'n''++++>-?---------U/V/y/{/s1޼ɳɋrlfa hI6ch hI6cCJ hI6cCJhI6c5OJQJhnH tH jŒhI6c5U$juoA hI6cCJUVmHnHujhI6c5UhI6cCJhnH tH hI6chnH tH  hI6chh mH sH hI6c5hltmH sH h"mH sH hI6cmH sH hI6cCJ$OJQJhI6cjhI6c:CJU"++------T/U/V/y/z/{/11111$3%344666 !1$  7d1$$a$$a$s1t1111111$3%3536334444444B5j5555556666G7779:9:D:::::;;";P;R;;;뭢hI6cB*ph hI6cCJ hI6c6hI6c5hnH tH hI6c6hnH tH hI6chnH tH hI6c5nH tH h hI6c5h hI6chhI6cB*hnH phtH  hI6c5hI6chmH sH hI6c hI6chjhI6c0J!Uh06666699999:::;;<<>>??????AAA  7d1$$a$;;L<<<<<<==>>? ? ??????????????s@@AAAAA+C,C:EEFFFFGٵ}u}o} hI6cCJhmH sH hI6cmH sH hI6chmH sH "hI6cB*OJQJhnH phtH hwhI6c5nH tH h5nH tH hI6cB*hnH phtH hI6c5hnH tH hI6c6hnH tH hI6chnH tH  hI6cCJhI6chI6cB*phhI6c56B*ph+AAAA+C,CFFFFFFHHHHH;J h88^8$a$  7d1$ yzzz|||}}}}}2~6~;~<~F~y}-/CE{†ΆފߊrhI6cOJQJhnH tH  hI6c5hI6c5OJQJhnH tH jhI6cUhnH tH h, hI6c5hnH tH hI6c6hnH tH hI6chmH sH hI6c6mH sH hI6c6hI6cmH sH hI6chhI6cOJQJhnH tH  hI6c5hhI6chnH tH hI6c,-./CDE†ފߊ & F@ h7d1$^$ & F? h^a$$a$$a$gd63 7d1$^  7d1$ߊЋыҋab\f܎ JTVڒLM“̖͖ΖϖTUWmoǼǼǼǼdzDž|shI6c5mH sH hI6c6mH sH hI6cmH sH hI6cCJhjhI6c0J!UhnH tH hBqhI6c5hnH tH hI6c6nH tH hI6c6hnH tH hI6chnH tH  hI6cCJjhI6c0J!6U hI6c6jhI6c0J!UhI6c hI6ch hI6c5*HIJTUVْؒڒUVWmnoט !1$ u1$  7d1$$a$טؘBlvϜ0A,%&#X\]کѾѭѭѭўѕ}l hI6c5CJOJQJhnH tH hI6cmH sH jhI6c0J!6UnH tH hI6c6nH tH h=[ hgLRhgLRh, hgLRhJh<0hmH sH hI6chmH sH hI6c5 hI6c6hI6chI6c5OJQJmH sH hI6c6hnH tH hI6chnH tH hOhnH tH *טؘ%&M#$($ & F' h^a$$ & F& h^a$$a$  7d1$ !1$ЪҪ16Ny()*2346cgwy)*+,-౩똊~m~m~m~dhijWhnH tH  jh-h-UhnH tH h-h-hnH tH h-h-6hnH tH h-hnH tH h, mH sH hI6cmH sH hI6c!jhI6c0J!6UhnH tH  hI6c5CJOJQJhnH tH hchnH tH hI6c6hnH tH hI6chnH tH hI6cCJhnH tH $()*456wxy,-./09:;NOP$a$gdijWgdijWgdijWgd-  7d1$$a$-/089:;NOPSVdefLop~s~aT?a(j&hI6cOJQJUhnH tH hI6cOJQJhnH tH "jhI6cOJQJUhnH tH hI6c6hnH tH hI6chI6cnH tH h?5CJhnH tH hI6c5CJhnH tH hI6chnH tH hI6cmH sH hijWhijWhnH tH hijWmH sH hijWhijWmH sH hijW6hnH tH hijWhijW6hnH tH hijWhnH tH hJhnH tH Pdef5dݷLM$ & FF h^a$  7d1$$ & F3 h88^8a$$a$zȼɼͼмѼ߼4?@DNO߽!0134*78E˳ߤ˳˳ߤ˳˳ߤߤוxg hI6c5CJOJQJhnH tH hI6cB*hnH phtH hI6c5B*hnH phtH hI6c5CJOJQJmH sH hI6cCJOJQJhnH tH hI6cCJOJQJhnH tH hXhnH tH hXhI6chnH tH hI6cmH sH hI6chnH tH "jhI6cOJQJUhnH tH  hI6c0J(z3`ab*37 $$Ifa$ $ !1$a$ !1$$ & FH h^a$^$^a$$ & FG h88^8a$ & FB h7d1$^ 78FYqkk$Ifkd$$IfTl  0 I     t0  4 lapTEFXYZfgxyz VX $Ʒƪ~x hI6c0JjhI6cUjhI6cU hI6c6hI6chI6cCJOJQJh hI6chhI6cOJQJhnH tH hI6cCJOJQJhnH tH hI6chnH tH hI6cmH sH hI6cCJhnH tH  hI6c5CJOJQJhnH tH hI6c5CJhnH tH -YZgy|vv$Ifkd$$IfTl  0 I   t 0  4 lap Tyz|vv$Ifkd$$IfTl  0 I   t 0  4 lap T|vv$Ifkdv$$IfTl  0 I   t 0  4 lap TWX|upb`O``$ & FB h^a$ & FI h^$a$ u1$kd1$$IfTl  0 I   t 0  4 lap TW ^F"#$/01gdEcgdEc  7d1$$%&UVWX IJK~,-.\]^_DEFGxj(hI6cUjhI6cUjhI6cUjuhI6cUj0hI6cUj3hI6cUjhI6cUhI6c hI6c0JjhI6cUjhI6cU8xyzvz!#$(/1RTWqsĻⱨ}wi`hI6chnH tH hI6cB*hnH phtH  hI6cCJjhEc0J!UhnH tH hEchnH tH hEcB*hnH phtH hEchEchEcCJ hEcCJh, h, hmH sH hI6chmH sH hI6cCJOJQJjjhI6cUhI6c hI6c0JjhI6cUj]hI6cU#1RSTqrs}v:;   7d1$$ & FJ h88^8a$$a$$a$gdEc9;{~ &'hFSZDH`cdel շ՝կ̔ՏՏՊ hI6c6 h}6h@hmH sH h}jhI6c0J!UmH sH hI6cmH sH hI6chmH sH  hI6cCJ hEcCJhI6chmH sH hI6chI6cB*hnH phtH hI6chnH tH jhI6c0J!UhnH tH 3wm $a$$ & Fn h88^8a$  7d1$ & Fm h88^8$ & Fk h88^8a$$ & Fj h88^8a$$ & Fi h88^8a$HJ~Tc/kw""? >0;<;BDKQ[]qr1ϸ쑈욁xhI6c5mH sH hI6c6hhhnH tH h<0hnH tH  hI6chhI6cB*hnH phtH hI6c6hnH tH hL+mH sH hI6chI6cCJhnH tH hI6cmH sH hI6c5hI6cCJ$OJQJ hI6cCJhI6chnH tH hI6c5hnH tH /HIJ  7d1$ & FfCJ">d1qr  7d1$ u1$$ & FV 88^8a$$a$$ & FUa$$ & FTa$1v9:#-$If$a$$a$$ & FY h88^8a$ & FW 788d1$^8 & FX 78d1$^8  7d1$1vz{9:.235689:<FGIJLMOQTUVWZ[\^defgjklnrtuwxz|ҴҎݎݎݙҎݎݎݙҎݎݎݙҎݎݎҎݎݎhI6cCJhnH tH hI6c5CJhnH tH hI6cB*hnH phtH hI6c5B*hnH phtH hI6c5OJQJhnH tH hI6c5hnH tH hI6chnH tH h, jhI6c0J!UhI6chI6cmH sH 8-.369;5////$IfkdW$$IfTl  \/}MNNN      t(0  4 lap(T;<GJMPVPPPP$Ifkdr$$IfTl  \/}MNNN  t 0  4 lap TPQUW[]VPPPP$IfkdK$$IfTl  \/}MNNN  t 0  4 lap T]^egkmVPPPP$Ifkd$$IfTl  \/}MNNN  t 0  4 lap Tmnrux{VPPPP$Ifkd$$IfTl  \/}MNNN  t 0  4 lap T{|VPPPP$Ifkd$$IfTl  \/}MNNN  t 0  4 lap T OXz3N"/1Z["Fq|½½듌|sh8hnH tH h85hnH tH  hI6ch hI6c5hhI6c6hnH tH hI6c5nH tH h.CU hI6c6hI6c5mH sH hI6c hI6c5hI6cmH sH hI6cCJhnH tH hL+CJhnH tH hI6c5hnH tH hI6chnH tH hI6cCJhnH tH -VPPPP$Ifkdw$$IfTl  \/}MNNN  t 0  4 lap TVTOOMOO$a$kdB$$IfTl  \/}MNNN  t 0  4 lap T !"/01Z[$ & F h^a$$a$ u1$  7d1$[opOt  7d1$$ & F h^a$  !h1$^h$ & FZ h^a$$h^ha$$ & F h^a$$a$$ & F h^a$ u1$ !1$$ & F h^a$ |NO PopPT^`jk   Y \   3     l p v  8<FJ _c̹̤̟̤ h, h hI6chh, h, hnH tH h.CUmH sH h.CUhI6cmH sH h, mH sH hI6c hI6c5 h, 5h8hnH tH hI6c5hnH tH hI6cmH sH hI6chnH tH ;OP^_`  \    & F[ h788d1$^8$a$ u1$  7d1$=>+,&'89: & Ff u1$$a$  7d1$=>LP+SWHghiry@FT] '8:"l覛ulhI6chmH sH hI6cCJ$OJQJ hI6cCJh!1hnH tH hWzhnH tH hI6chI6c5hmH sH hI6c5hI6cCJhnH tH hI6cOJQJhnH tH hI6c6hnH tH h.CUhnH tH h, hnH tH h, mH sH hI6chnH tH hI6cmH sH hI6c5h*XV  c!!!"z#{#|##8&$ & Fo h88^8a$-$^a$$ & FP h88^8a$$a$ $ u1$a$m } F!Y!!!!!""c#x##$7&M&'''(((())**++>-N-------...//!/"/$/'/)/,/-/M/O//////· hI6cCJ h TCJ hI6c5CJhI6c5CJOJQJh T5CJOJQJhI6c5B*CJOJQJphhI6c5B*CJOJQJphh^hI6c h^5 hI6c5@8&()*+?-.........../ /////Ff $$Ifa$gd T$a$$a$$ & Fe8d^8a$$ & Fo h88^8a$//"/#/$/'/(/)/,/-/M/N/O/R////// $$Ifa$gd T//kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp //0 00000000000000000000000011S1T1|1}111111112 2G2J2t2u2|2 3 33U333D4޼ޱޡޑhBqoB*CJhnH phtH hI6cB*CJhnH phtH hI6cCJhmH sH hBqoCJmH sH h-;CJ hJzCJ h[#CJ hBqoCJ hI6cCJhI6c5CJOJQJh T5CJOJQJ hI6cCJ5///00 03040000000000000000000000 $$Ifa$gd T00000000000 $$Ifa$gd T 00kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp 00000011S1T1|1}11111111 $$Ifa$gd T11kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp 111112 2 2 2G2H2I2J2u2 3 333T3U3V3u3v333E4F4$ u$1$Ifa$gd T $$Ifa$gd TD4F4G4H4T4j44444'6(6)6465666:6;6?6A6r6t6w6y6|6~6666667 787h77777788 88 9s9v99999:;;>;кųкŭЗ hI6c5 h, CJ hI6cCJhhI6c hBqoCJ hI6c5CJh T5CJOJQJhI6c5CJOJQJ hI6cCJhI6c6CJhnH tH hBqoCJhnH tH  hI6cCJhI6cCJhnH tH 6F4G4H4q4r444 $$Ifa$gd T44kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp 4444 53545i55555566'6 $$Ifa$gd T'6(6kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp (6)65666:6;6?6@6D6r6s6w6x6|6}66 $$Ifa$gdBqo $$Ifa$gd T66kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp 6666666666666667 7 77787h7i77777 #$$Ifa$gdBqo $$Ifa$gdBqo $$Ifa$gd T77777777778 8 8^8_888899@9A9s9t9u9v9999 $$Ifa$gdBqo999 $$Ifa$gdBqo99kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp 9:+:O:s::::; $$Ifa$gdBqo $$Ifa$gd T;;kd$$Ifl  ִ3 }"!l(/5O  t 0      4 laPp ;;;;; ; ; ; ; ;;;=;>;G;N;X;j;u;;;;;;;; "$$Ifa$gdjFfV $$Ifa$gdj$a$>;;;;;;;;;;;;;<<<<#<$</<<<=9=:=;=B=C========>>>>7>8><>R>V>W>[>|>}>>>>>>>>>>>>'?*?̷̽ƱƤhjCJmHsHhjhI6cCJmHsH h, CJ h-CJhI6chnH tH  hjCJ hI6cCJhI6chjhI6c5CJOJQJhj5CJOJQJhI6c5B*CJOJQJph<;;;;;;;;<<<<<"< $$Ifa$gdj "$$Ifa$gdj "<#<kd$$Ifl  ִ  l(-0D6_  t 0      4 laPp #<$</<V<<<<<<<<<=9= "$$Ifa$gdj $$Ifa$gdj 9=:=kd$$Ifl  ִ  l(-0D6_  t 0      4 laPp :=;=B=C=G===========> "$$Ifa$gdj $$Ifa$gdj>>kd'$$Ifl  ִ  l(-0D6_  t 0      4 laPp >>>>>(>5>6>7>8>R>S>T>W>X>\>|>}>>>>>>>>> "$$Ifa$gdj $$Ifa$gdj>>kdK$$Ifl  ִ  l(-0D6_  t 0      4 laPp >>>>>>>>>>'?+?4?=?J?b?c?g????????@ $$Ifa$gdj*?O?a?b?f?????@@&@I@u@w@x@{@@@@@@@@ADAHAUAXAYAbAdAeAAAAB3B4B9BYB\BsBtBBB C CCzC{C|CCDD#D h 5CJhI6c hM>CJ hI6c5CJhI6c5B*phhI6c5B*CJph hI6c5 h.CJ h_CJhI6c5CJOJQJhj5CJOJQJhhI6cCJmHsH hjCJ h-CJ hI6cCJ7@@kdo$$Ifl  ִ  l(-0D6_  t 0      4 laPp @@&@3@@@I@u@x@|@@@@@@@@ $$Ifa$gdj@@kd$$Ifl  ִ  l(-0D6_  t 0      4 laPp @@@ AAADAIARAYAbAfAAAAAAAAA $$Ifa$gd_ $$Ifa$gdjAAkd$$Ifl  ִ  l(-0D6_  t 0      4 laPp AABB!B*B3B4BDkd$$Ifl  FOY6      t0      4 lap $$Ifa$gd $a$4B9BYBsBtBBB C[U"$Ifkd$$Ifl  FOY6   t 0      4 lap $If $$Ifa$gd  C CCC\ChCzCma[[[[$If $$Ifa$gd kd$$Ifl  FOY6   t 0      4 lap zC{C|CCCCCCDmaa[[[QQ & F_$If$If $$Ifa$gd kd$$Ifl  FOY6   t 0      4 lap DD#D5D`DlDDDma[[[[[$If $$Ifa$gd kdW$$Ifl  FOY6   t 0      4 lap #DDDDDDDDDExEEFLLkNlNNNN?PcPPaSbSTTUUXYYSZUZ~ZZZE[L[[[i\{\]]]^J^\^^___N`_`+aԼԟԍ hI6cH*h^hI6c5 hI6chhI6cCJ$OJQJ%hKhI6cB*OJQJmH phsH hI6cB*OJQJphh5hI6c5h, hI6chI6chmH sH h 5CJ hI6c5CJ hI6c5 h, CJ hI6cCJ7DDDDDDma[[[$If $$Ifa$gd kd$$$Ifl  FOY6   t 0      4 lap DDDEE=EIEfExEmaa[[[[[$If $$Ifa$gd kd$$Ifl  FOY6   t 0      4 lap xEyEzE{EEEEFGmkkkkk^M$ & Fp h88^8a$  7d1$kd$$Ifl  FOY6   t 0      4 lap G>IUKKLkNlNNNNNNN?PcPPP[QQ R'R2R $ & Fda$$a$gd5$ & F n^na$ $ & Fn^na$$ & Fp h8^8a$2R=RSRRRRSaSbSTTTTUUUYYSZTZUZ~ZZZgd^  7d1$ & Ff  !1$gd5 !1$ $ & Fda$ZE[[i\]]J^^_N`+adbebbbcc3e4evkwklll  7d1$ & Fc h88^8$ & Fa h88^8a$$a$+a?aIccccc\d4eeffRg_gchdhjjjPktkukvkwkllll3mymzmmmmm@ngnrnsn oo(ooooooooo.pp)qTqqrrshI6c6CJhnH tH hI6cCJhnH tH h{@h~> hI6c6hI6c6hnH tH hI6chnH tH hI6cCJ$OJQJhI6c5mH sH hI6cCJmH sH hI6c6mH sH hI6cmH sH hI6c hI6c58llllllymzmmmrnsn'o(ooo-p.pUqVqrrhsisStTtt $da$ u1$-$a$sYshsiss"tQtStTtttu&uuuMvvvvvw1w5w6w7w8wwhxxxxyy[yy}u}jc^X hI6c0J8 hI6c5 hI6c0J95hKhI6cmHsHhj{/hj{/6hj{/hI6cCJhnH tH hI6c6CJhnH tH hI6c6B*hnH phtH hI6cCJhnH tH hI6cB*hnH phtH hI6chmH sH hI6c6hnH tH hI6chnH tH hKhI6chmHnH sHtH hI6cmH sH hI6c hI6c6"tt/u0uuuLvMvvv7w8wxxyyyyzzo{p{{|||| !1$$a$gdj{/ u1$yyyyy*zHzzzzzzz?{n{o{p{{{{||||+|-|.|4|8|<|J|M|X|a|||̽|si`W`W`W`W`hbhnH tH h&%hnH tH hI6cCJ$OJQJhI6chmH sH hKhI6cmHsHhKhI6chmHnH sHtH &hKhI6c6CJhmHnH sHtH #hKhI6cCJhmHnH sHtH hI6c6B*hnH phtH hI6cB*hnH phtH hI6chI6chnH tH hI6c6CJhnH tH hI6cCJhnH tH "||+|,|-|.||}H}`}x}}}}%~E~f~|~~рҀ*>f"gdEc"$a$gdb$a$|||||||||||||}}}}G}H}I}J}_}`}a}b}w}x}y}}}}}}}}}}}%~&~C~߽߹tllhI6cB*phjhI6c0J!B*UphhI6cB*hnH phtH hI6cB*CJhnH phtH hI6cB*hnH phtH jhI6c0J!UhI6chbhb0JhnH tH &jhbh`k~UhnH tH jhbUhnH tH hI6chnH tH hbhnH tH (C~D~E~F~G~e~f~g~h~{~|~}~~~~~~~~\]ab΀πʿ}ʿf},jhI6cCJOJQJUhnH tH hI6c0JCJ,jBhI6cCJOJQJUhnH tH hI6cCJOJQJhnH tH &jhI6cCJOJQJUhnH tH hI6cCJhnH tH hI6c5CJhnH tH hI6chnH tH hI6cB*hnH phtH jhI6c0J!UhI6ch(πрҀӀԀ)*+>?@efgȁɁʁف𯛑zlbZjhI6cUhI6cCJOJQJhI6cCJOJQJmH sH  hI6cCJ hbhI6cCJOJQJmH sH hI6cCJOJQJ&jhI6cCJOJQJUmHnHuhI hI6chnH tH hI6cB*hnH phtH jhI6c0J!UhEcB*hnH phtH hEcjhEc0J!UhI6chI6cCJhnH tH #fȁɁ%&'+,?detȂɂʂ˂$a$gdb $']'a$"!"&',=derȂ˂hI6cmH sH hI6c5OJQJmH sH hI hI6chI mHnHujhI6cU. A!"#$%#&P 0A .!"#$%. A!"#$%@Dd\  c 8Aesperontob>ՆPhNlD nd>ՆPhNPNG  IHDR12TPLTE*0wr#mq[U cmPPJCmp0712HsIDATXGmsFoq BE%J KTqN"jB=ZI$3y}a3ڽNwc!gt|ފoo](v4'8O,L 'mAy't68(0 Im/kи (>tNg\Dm9ø̩1Y{KlSP7vqC"IjcC)_o`\P)n $=8}\r=ځS5ʲ%OL[ ӊTJjb'>q.Ij@f^;_|._VMBs GXi.m[N,wmX,Pf+$=2 gԄn7F DźFٛP(Q;>lM!>0g2G ||Zvp&oWoW]u|Pu)~VzOA9fR`=ڷLl; M{u().to/[^F;Q vI,3FsRteQlra]&ReQ޷:tNV(jgYe$ic!k"?흡 0~ÛV8 d!>irxUCkE"'ùrv!~YZf}''NldѡJB#[`}h֏τXy0$|!QXd./F37a~X $H(}'H DJ%[bKl%IfءsUcMN~UZbRЕVB6cJd믈Rɡ5*`QʧGZir$lP+KUP._Hv,BqRIe.^Itt;( Кԃ(8=B;iI?|Pyp-}<rpL͜r8ԈWmfIENDB`$$If!vh5O5J 55#vO#vJ #v#v:V 405O5J 55/ 44 f4$$If!vh5O5#vO#v:V 405O5/ 44 f4DyK www.esperonto.netyK 4http://www.esperonto.net/$$If!vh5O5#vO#v:V 405O5/ 44 f4$$If!vh5O5#vO#v:V 405O5/ 44 f4$$If!vh5O5#vO#v:V 405O5/ 44 f4$$If!vh5O5S5755#vO#vS#v7#v#v:V 0U!5O5S5755/ 44 $$If!vh5O5S5755#vO#vS#v7#v#v:V 0U!5O5S575544 $$If!vh5O5S5755#vO#vS#v7#v#v:V 0U!5O5S575544 $$If!vh5O5555 #vO#v#v#v#v :V 0U!5O5555 44 $$If!vh5O5j 5#vO#vj #v:V 0U!5O5j 544 $$If!vh5O5#vO#v:V 0U!5O544 $$If!vh5O5#vO#v:V 0U!5O544 $$If!vh5O5#vO#v:V 0U!5O544 $$If!vh5O55*#vO#v#v*:V 4J0U!+5O55*44 f4DyK dybkjr@pdc.dkyK ,mailto:dybkjr@pdc.dk$$If!vh5O57555552#vO#v7#v#v#v#v#v2:V 4J0U!+5O5755555244 f4$$If!vh5O5#vO#v:V 0U!5O544 $$If!vh5O5#vO#v:V 0U!5O544 $$If!vh5O5 5 #vO#v :V 0U!5O5 44 $$If!vh5O5 5 #vO#v :V 0U!5O5 44 $$If!vh5O5 5 #vO#v :V 0U!5O5 44 $$If!vh5O5 5 #vO#v :V 0U!5O5 44 $$If!vh5O5 5 #vO#v :V 0U!5O5 44 $$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 Dd 8`  c <Aisoco-logo2RLIE( F IEJFIF``Software: Microsoft OfficeC    #%$""!&+7/&)4)!"0A149;>>>%.DIC;C  ;("(;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;H"F !1AQaq"23T#4BDSrc$6best21Q!"Aaq3BRr( ?TvfE>#xQBP@ˊtt+%l |WJP?wJ~#h"#rq+@V *f6NH)8 WE6pYR!(Wi+8'E/ oObu{-=~;W2a.qʒTl}9ZzcFd[֑K¯mwwt(Yܟy Poy %!c9J9ΤX7WmIPQRH+ا$ /,S g5)Q"C 48Gּ}ATJݕ! 28pK6l4d|i;ϯۿ**4-W":Op$ר~jp>Jʷ(y;6%u>t7?H E\BTN@P6F=Z>&/"+a;[S>$>f5gJ&^Lbsfryfӣ 4LpկZg>wVDJ6kwI;'gjiqGFO[Sk+|ǒӨ\u(` DKIܠ=Pu1;V &/"+ayL+>ti?J*;\IMjiр~ɋ3Vom>n;U]I\.S%*i;z7l}F.8d}t$kOqc* Uз1$}=*3;g j>t#ĘɑЊ~>da`wvQu 9|: \9KSY<59KSY<5Q53"9gr!Ɉ}SG~jLriȡ. 9RsAžgn r|qU)> "[zM|huŰ*kZ\RTa զހ? M$|VM_S-ȇ9aԺpíElol[{w56UQf"׊|R{rG*^~5^H:>{2u[cNUԴw OxNCmd d}֯J$mw뭧oخa i_a.z#f+N)n$*4\ҲG +4tGCkC†1Tq&gKL _+R,P>lzkc1e)NAc녟~v";pgcwp*Aoј (zm]SI qߚ JwUr6}m.yA.;~+D 'O VeGdeq 0bw]]!#99jִڢӄЅYEbQ[I ܑ@޺[Gypg=sv'Q"#)K=qAk\]Q,Opf ijUfb B~?'h₴g%A?kE/ si/` V~掻(6 '> Fĕe]mVIma,L2v4#qճ@ɱQV䕳Ԇ* IZ^NT#+D\p39h^EyO0 IPVqOIEܨJ2`eŭ%!?X$n^zh ,K窷x2J#3'+BUurc!L6v4ʊu(? s+o]XΌ8^~/- H}jChR84v3)SQ}e *PyUP޻'YxvиVZ񌚱;TG%mZH <*]mkCFrd8$edC} Y+Y |ÅڋA%xB4;a^'Zٓ=0җTwHRcjJ Z~CnҹC9i|͋=-4HX^J+rQ-m)"0!n)RI=1I\f>'2 =E-=0]"IHLeΐypԘŒ<8r8W]J|O:]UxNGTuڐz# Qʫ^G*-/;8R('qDPvk 0܎%@$}٧߭we?qDNJ I{jtdRMŌb*.za$晓noXfIE}ʗbҢ9rA=q҈ yri/i3_}YRJyƦ E;bp`/$fvoPåJKdJ'NmZP8I6zfdSI4Emɫ{Z"C)!@{U,a:NqTj #H @g̍M6"_5CBړX9."mn2!- _[dӘywUr4SZzRP],ܐ䀕6j_1N^֫eLbVqJH!@#OjtiQa* Zu;p|( nSa*spf#YZu)7VJ]Hho]\4NM+e`+wi Ǎ].rq:89Q(V":2\R,iՌwIMԾiWGǴ\HI=*ZTpnC6=Z}re&,VqSCܲD( s.WCc܉UKhhHF@@ #W7L>T)4W ?#G uj%FnoHqȨe,89iow(5^mcn1Dg"Zlȉbn- )C%eCP|/179ϰPv$aS JK(L'î*)9GVhn:a(ZЙ(xۂFzfkIm*FI I<'i5-!cOi*)#n9SǶÎ,6t[B1Q[2yJS]Nq]E%pnn) R^Z)(@9H?]1 Nj\9rdץ:Psݑνf}~&ܛdD([TAyۏg9N~HCbz:$ %(D|1DW(oJ]{EX3N}nS;|tI)JG}HL֕*" 6;EN,J$My,W#2ukZQ9+$JCsMRc":5mR]'O$p%]S)KQǴO,`tJc w<$!t$g'6eY_ vzyrJNDo~;##[V޻yu:ryk}뮩KR $:tJM$$FiqRvD1ҥDP %dNHRxzxSdLrZ@並r*"lŴFTxa qNeDG$|j#n"udMY[bJ(8>Q+&2h2}HQ@ ?x4ͻ\u=/)xv8RD^!uR-){Rx :#A1v91 T>s'vH?%}AA}#NxrAw:p|}}EIvcm FH=ngo?q}G_o[vاXz'{;Sl'v{)2d㡣ƯwvGwClh= |';NYt=yE]Hc}A+wxم޼a^W#L35v/~*|kG0!_^%LV+z_dZ$=_,VPo7sgK0A046DfCvw!ԦPC, X|>bNJyċ!ֶlV&_%jʘ6lx0TA*wpju3U)w0G@(!LfnsBNFY3W:l&p:Kj:l63]6#Y%[GǭG.fo|X HN0 Z}{`꤫,P'pqQMgJkl5A;|k4[wUoU Z89GlA?IkLl3[pU= &˿- _^$l2(##gVEEveUyLr*'dX)LWY{7ebmYq=Q(]X dIԈ/vA89MЈ\ɐ>$BM2M @0ZNң{]h 8=L_⇉%pa8[S&|r4lL9އIj&*u#O=QJkЎ?Sv',@|j@*p!06( N@s_,Wiێ8Qᮠe nέ'1(o4yR=+du:Oob\G󅳮6=+[h|XH:vF E|s(YԟڀGrQ{h{ |gs!f_I ^4>f|ڍ\_O3Wl(g0uCÄCRV<&pkB˔ek<}~mIeh30r 5̘TeԩooL%LdNœ[<=9kb ;a!!%)I595 mFS764hYíמ+L &Mgn__,r ~] NrkD[p*_UV@;auwf#0Ə &_O qzB;EĔ74jN}\Q]U.C#+ԵށJ P @ d oM!TN9DQwHڎaWm+R5AqS7Pej*̊  FݎfInU+Gs H4QdOMsשv;3**xdzŪADEK+/Dט413q㙍嵑%2ǂʚf]Ov?%>ލ*DA˨i 11 waɮPt,IhD'p$+>fj t2M!$Ej{QZѠDhԣQ:9ƽ+yvSÄ雭吩1R c9/ROd^J <8+κ7)yW&&Q+TGsӯ˴F_l@nU!~1e2F*a`xQ^H:5:"ʪkQZ WrKnnD(acK1aJBJk/ѤtY|һ|eb֤kM RK9](T)Buk̾DY#6 `{L,CR=(z)ܾ˓'MV]$Zu+E0Vq֥spV$+/&[ M`4&5)Bn2q8SlXsǒWJ류pW^HZ) ¢^eJU:TmD|XSokIe!Ͽtr,IuMZmX{g%1=bX٨VAϘhzSэj'?ZrX/Ƅww8|z }AOƛR34N؞.h@ {RUN ٺ8b2Mf n2AN `Xv dYp#+7[pVIͮDihxaz&,vL賮 j0{ž >"lho1%Z>-؟Yhiû ky5uMfLҫޏO%skI=8RN\sl0DyaC_"D+uhJd|h6 "Mc00jJ HJMKvwg .\\/bS͚f|̄NѠDɩ6wwp6)%IUƏ\c|6?>-C{&2C-m&R==>ڋ@CVԛ&B\sCRlby*i zJߟr8c&\HdY3ǂ_2cI\YWASWh࢞䌟Ӌ<=pmXytR\QK[[yƗ!M$&o{_ y$i#~ڣ|UV,'3g6c8k}8]UE<25yΞ~vڵ~a??l_̜;fWKU*FmqKΤJ/0TU"8B(FSb\7&|_!%=ٗɁij_*tf{;}̥_-xY@!2段L['&Ã$-oGlSi۸*b eaŋbph[Z.[ŗ_f:!yrN=%y( a;w\>|NG $}cbvߗ_%RbY=8?4E qzBrM?ݯLar_WL}_qoii1[mdʃK/#wP7BE|^0RfWpCA$R~?|~}ǙYyO8O"'~lN9'h3Uī[/0Lx+lnA bJ Z<ui F#!GapРU,/s\TEpDhDsA*-,Eı?+#rr|ߪW\љ$x+ҷ3'Ђ/VkU?JJ4z#M33_e&ttÙ(HO"^e2Iq8/^q~bc7#~ko[ݽ0~"D#pIgKx#~ )8FBSoϘP/![NSٟ\-/oοF7hn2Y5uA2|vMFhgUBF_,LHOXI̯rj`iU$K|Zq$BPsAgU"ьNNLB&)Sbp&tYno1ރ)պ6{K/\s9s+sO]LŜGCl˃peeHJMaŒNl1QU"AS*&#ṱ@t qv+݊\LGA{0{ٹ[ www N.t9spks{]}(G& h$;dLS.TI2y`'U&|9^u=H$KNbKrd$x;IVqLE /ԫ< E~m^˧.;\"#Oh_xߝ~m4~8jO)/Fp\A8KRڠm,Q'U5?Kb- 0Dc(G${38ܺfďWVrLLTNANq`g>"=]#]Üz8&zF{E{Eu8vz4; q|\s1oRoV5J(K_x(Lƶ9|`܍Y,ASKY'tZگW(j/ӣ/R>C\Nu jINa .n<]c/\ymS09cڋ0&u$z#2+jR,j1݃ӿ~y,K<64/Etmu=`O>#A#mݷ+s勾}.8:q?sL#2W7"``ux;"Lx @*xjVO7ŖN# `pDnukk[IG*x#)\G<=*p$ q=ҥKng=/88{Ƹyw9}l眊`\àWCEHwsiP$0'I)!8&42UzJل--8EB×?b Gnq Dt us8s%#699-yX-#8a‰PˏԲjm] .+Ea2j~|NN(4 Y׊禔ḟɂcCVE8 8pn>3V,'y׋K͹U K”4B"y ?dɗ\HUTAu2qR_{'Tp' KACs>`K4\N_zWbz+B ##$tzC ?mȃ=ԩ{|$d|{ݺ\,Pu̠4w;SC.98dVv)[ _nOZ}+ȢJUgM-:{ ]2 Jț8MUNGIW=#&ܘ@ yI~{ݹu΢7qv u `_L@fO xM|߃<0DVhkcÈs5-O9;>:S%YW33Ԛ}YaTH^PQ<~'G\;1[ֿ65os^qG{CD?!*7=y|*A|J)/r(Mw#YfćV"H;ѱ8>1$[T8Vb‰,\ xB`zu\2Bpx|C\b8~ihH*jW'}{E >/Ä ~F״J]kFn5B~X]#DZm pΛf˛t@0V-],(51<\@n#O<ѝ>3aK]HƊqrDA`<"5 22^+7|G,J78r#P氠ytX] j9s/s7BuK@J=~X'BFx̐#GtI N?|r"8>up'+_& SLݘR*":^ҷ R4ڬ^ba> :^ 9^֮]<9͠!>ɋ~r`cׁGGf 0Irx!&lځliw&?#ǗTfT1vd?{aQ@=9=:ql؝orz]?N=ńiZ^n*5 (x?haR/<0/{b+bX.Hf=+&DP՛f:5pyKzVe &ء() BH>odC.m|ۛ#_/C ={!Ap&.ѡW t[Xu=+DX@&+nVC* !dԵ`zGbE@M( il8"I`1?ʘx_<1J|Й=৐W!Ua "۵4!ڱ ف# }@ DE0"AN"g> [pP sVV4q G~yڠ3XƃS˒O[CoIgF'82qsOutRnτ3Pz-u/%Kd Fm56E(և\ܑ]oGtC]gI ?jHFn]s@"gׄqkm75j5m:0m`Թ?ks‡̻6x^9Nlˌs?xs6fm[]sk5rvtaJlц$ʛ\·R|WF^]O`J6c\&z*72E HNe_-n>e*۳ [:\Qӕd֛)%ALI׬$C Y®&]zH+>?w/m<;&@%8ɏ.B|I%J]j3qMhnD8K= ò: W@ϓպwISg knA'g\e@6[XVKPʭUm5ӄO=5q'<ǜ zOw)wcl?;gWRW~X)OUiNw$1L8aT;2ʛی3 SJm]kgj 3Xf]ymrHe $MG᜞5|0Pozo6ɉ뜼4eo.EuYw}n;=⭳Ýb&83'ز}g6;mmUD! hMlyTs0 MQTC K/yf^ I40u眰R44hV!usY>b/b( sc}: Swbμ)+7.yc_Ρo?x`u:hM{Q4l-.Hb %jf_cͺV?3u,x%;!$eBC[mv yJbaG9GreGG$Yl@YN|INvxkԕf}/ٺP쓜[O9tqʖ}̦FB~UWQn*b0qZ gF]:e'mZqؐ^3*xbKD⒳qVa÷aoF淄ԥOEݽ$N[jzB0hTWe #$gM.PW*JM:̀Gr :Je7Ժ֌*MT. V^F]j)В;hf:n=ձvXdjh10 $k 5ZEKK BxT!lnْfBb4|d]|s5QdM 1du#٥%Zɏ"nҫjܓ<7M}i=,v4j_$c ^,fB_~%wéG+ kjNVOcL` (Α3%`r@͊Z<GdZê c~eu< M֌Me@"DDf |̼HAg̿ƭQ'lˎ/ռb#w=NրSH}?hp0pesFPm Alq< $=+ѭQ5ihfkkGȏ(ue\Kbav }P@\0CK$zDX*W) RJ\B^,taͤXj~,R: z,-h:-֎g޾.iҥ{2+QEPUzKCKkk]kWTOql)0ӄbW@Ե!%#f}ٟ۳h$Jָ(;nL->Nي3H'WF|lcc5^s/8;2*7-ۘV+U1NԂ9J d81w]ki{WEDyK asun@fi.upm.esyK ,mailto:asun@fi.upm.es$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 IDd y]\<  C A2H9`}L~Ue7OH$^`!H9`}L~Ue7O8 Hx}\/:1쮀 tW /x nL]o(+hItsQW -H4ʙn(HI$P0VR$m Ni"ہ`4>Ilg$?|,4zy~|A-1tQ *Zf>mȅ裏>^} h{U胏':647ľ NzxGС>: Ət0zVq`ܽ`\!=. |t`ܣqw_rP!J ^_ }z~W ^>_ Go 2{\ g4tI4 M7^^M@'Ξ~:9bA4c{8O< =!09,k!]'F ☓0 ̇cG ^MaB#| |/` 08vzP!d}X tXw4Xs"ˀoz3yz(4 dz=5dkm`cࣇ{+X5t=\mf8W ֓f0Wv/W!sA80C3]}X_0S9B s#2+ck= Ʃ#q ƞ> *e p/{}`c`}q=5p>Nt_zчBhDX:,Dֻc9c>7^?`sOxvQ#zƊX^Ȓab]E\ ەWX^/-\m1N^!EMueV8w BVɷG)ajbL-6B*di@V^ 5*̄ nP-'[|Hi|LryIʯ^?y 6_.r(jʭkoXDvS}@ Y΂vImUS. K^ՔH!i힂ҴY}J*n%=lz۠Q>jlJ{[a*K>{c؆*W@_I p5[U)iK*:{S mԮ!-./^g\6䕢6$M5_-Z#i{SdDi[YUEL)ΫȲsfLwi~5`Rpz_D,\\J*p~ST(΁eu5VYLIZ+ͮHRiKM$ϕre܊}e[Wrjo)}u{@ܚy)PڽI.e5iS}ŅٍY I+J+r,q3Y9Ҽ\EyiiIv﫮mU+IbQ%UƔNġ{wWT7eʓ{+Ì.L-\VP"+w_k9Ғu][ꪔ2TS7f5V7B5N &#\TBi1[|)6H)y1@M޺꼦r[{rSRUoݗs:HHVUQX"/TEeKgҒ'._/ b$ρj;Fh,U@&"NK65THjJw+A)+d9%vՃUV1dE˭k 6T^ w{JT7T$_P*+mW[V!$$dP}j/ X̀bJP|s"rms-O6INTXS_7)$ɮ/ǔ'-5;5yIuUCt`5e:‹js$M0V'yC6Xkw:}&(/4%rrrr@)ߠR`z]TSYZU]&dP=7=n$vWbګ$PW-wb(Y>xB TTR U^SuMLQ\)@CY9Fhˋ'yiEnÒ%lmА&UaInzo *m:fS$ҼbqLQX@qVSvcaEq.`L'jON!\ߞ&J#-{OF_ /-,޽[ԹڒҽbaTTJJY^qM ߼:T/OEdNHJ f&D"ISYaRcʲ=0ɫ#f']P\4Ev2lJ&w_2jK5.m5%WI\XW"P \{m h*/ߝ.{-kȧUyy2PUTK@ ; kd)LMt!&&y@e)ޫ$ĺ bT'_VȔ'Š7~$GI^](Yr09wRd喖)Hj,n[VR][/}>i3 innVIkRQlYҔ&ɓKҘ6=ū iDԭr1gܕ_#PWGSUQ:ϥ\M!KeniWVk9$)*ܻ)C>pr3 (vs%U }+IB ҂b k$;B֋V[2ò1XL5\ٺ d%n*,(L݊⯁8Ia* dL֦`$[c)f˻d<ٝLZ[[[' R^kuײ=TDZ,+.(*)j4%YHAR^-kO ad=% %U`Ϡ(i.tVZQ;%@jWsr Vpnnuq}n%$?5J4جUD y]D!ͫU! lIlG(2?;e3*[%րPyΪ4IM^n2dAY5TԘ]P!YpN5R_ԃMTu}KӺzZ(/*.dLxM:ktne7m,𢡄)R8Sgк14Jƥȯ)"(VOqVW\_-7nEYH {I'QYyinrD[$QjBq-ٙ'93H9}Pvi؏RqRSQvO~uP $<,)ha iyR]DUA>(K4$ܢ={Sx%aVe禵|)Y]/6I!,yΛTHZ=6m[װzp IޞFg6{4Ty>Ѽ=] ITP]KfE2|bTe6 К剧.bQ՗6ܕjZ |OHyTHX]Q[r 9mgNIԊա1y.S׸5dEZQ j |$0ZD^* x$bʆr*cR #k+nX76iŞ9{IY5K~dBdV ❢NTTOrz@E-JY颦Zt!M7{.vVr8l+$`12-_vKc%ETLӎBޔ21tMK6EνEP *"I˻>=f_Em͝nJaq][[7ovAk%H)?g)ҵo1n6muSbsv? p=fǫJœe-s [6.YVU]WWl:MT[N? ׊N.[|< `gE8˖.LdWT4/+sA~Edf:GR_"O~qKU/&z`*oJRII-"\E^&$n!??o@$AC ,| ,߬^ŊONħw~zwծ[u^V]R](a͘fM^)ߐ%ԙTv-LNVSM|W%MTOV-\ `HͭM";tÚܛS|$URTrs~-Rת7g$]iNbAݫE즕Ȯ.3c)^[WnHzweܶW7mVV-((K0.K^):bɖJ!J wkkCm3Q-쒦LclPV}%7HnK@R9)l3fEnean'&w"w۱䎨n.7Q<8fڕ\3&!+][Lɫvߤ˽{LuϺ,)_KW߬fT|6mT7k;7}u6OXϿFä4n yo n+M/LPKwUA5ϫμr76n+IP?ӛ=;LIK|WpeCYw:]B[cI5n0}kЍ>ل&s[1E7 V݀)[^/OWni,%΍^֘VD0[;>omu}^ts֠1u&ckbZdp^~n:LcVWMyUmF&uqNv <}ɹ5OnP}Ż LX,8v[7$"?63c v>^̌ԧk[ 0M~sV|3 K`'kwsKkbgQgdU:l;_y)sf:K"O727-m VЗ>%>b=8O 'SKn wA$YV@M = _'_'{@{/ߋmO? skI : `,%,R {,"s=Z,IG`J3Q8/7;vZe&7tm 5\[[ֻi0fg9qoCUnIcA6Ӿwm"h|bEnRoI$|.NO-ɻon);tSVRT+HilZ3ɤa*Ӎ8r7Iފ!. - |bMו3#qm]F`UlM!ꏬ H*ʕv}l)ͨ%҃ά~vޏd>2"(.w~.qbڲ‹I\H$/3'̯S^&YY>s"8(+Gn#"1*-L<nVϚ=~}9=ne޵GG)N w'9˧1i s\EG:-x&Srt&gAI 0:#^){܄ fr (7!d줩`XMYJS#nJ9bS.1 bpǡ#DHj@?J2Sp.܅b=&ރz:,.S'>J'6،Pڨ"i _&2`M@}g`2 <p{mi颋 # c9/`r3$NK 3B-^Rοv[5_C~kQ|A!6d)ڌj Q*+޲<:N7x/)Wh!2:(f6kG`#.FC;[|9:=2s|qy;B[<ΐja'&MУn7 ιǣ},9ЄK@~vs3w]* F{Mj)(Qfn _N \<> ń*2aH.Z> 2SqkJQ ,ܽx`;92pLu1..6ÏXݴq~5аVêj0n2A^W/OK+ E ~⦨ѷ E//y0q j]ό) M?`.*L Cs>]` l {?Kvtcg7 P$̘ LA}VLqiF y!KЕ:oמ.#XbE^TUjsCdg#G2 9Sb,LVU CVHr4~; wfpH0j8vd6K $.+ЎJnT. #(\GSdyLRt&ݖq^kg^?\BS!0շbA۸oĔߢ-XkySOp"g Ã4N;5/P ]j`p74rQ\ vç5uVWM;TāX+[2OdU1זEf;1@Y#0̊Vq.˧{.@H\t l3gCpoomKOQF;źf:w#''fiB?R.ݤvM}*L9܈0YfgBcHKv2^ 1 #/ xÝdž?_2ƌOtPH N/Q_Z>&9k3}?910eq, c3? ڬzF8 W=|K L!㖈;IJ9B]cb *:`պh8sP~Q`[,~#yFcGal ~g/C!QMA ߧQLU.6z=aNOPO(3^`Dޕ=6.3Oa{ۈ()&KZx/SsXx0D.,,a --F cv,+-e zIXQߘQ,½ d5NxNɾ;tQx-}dȤ%^;>]pb% :8Oag7pWcHuc_q?|ܔM_Qܚ + C 滆g !bUev?}"dN 7!ǩ0yB8=4 r|a%P~GIu8sp(GIED~cLFM$M!ËwF7Ez1h5xЄznl_SVqM+2fzdeb[f\xz;{ iT =1OW|@?0 0#aH0Ik /ۙE1 ;.(^>?Ю!SL ]<yţA{!r&.<8BP,aTcCfE+NӖ^|qNza%deG8_Y hWLV̉\ѶE4!'Ϝ>|!B9nK+9)lidUqs.[K﷟ѽ8wEkݤ1oְn4eִZ@Lcd#g{y҈AZ;JEj z x9a(f$s>c4X?;AeM}>v|Wj8P6c̾0-Д5Kq@*Fۜp&s\F ZC4vH3°IpiS_g'hƼE!uZ=xjP#W4 2[tQ*FzV+76ܻh,x3NdYA;'A6R [+1P4 X#u^;8гQ} X7B3٦A(/;~4-]X?ڦ: lfjZ$D|l1S#GiJk'hJ|~ ҄U6Rv CV_  $>e8s; gppP4F1Wq 6aR{м!wPB[0P!]CCK@Daj]nQh#/lcFma,{|llL;@gH;L7#^<f+q8t-f x␓Dž* 12 R3îZNa$* 9F58)s[֪)db”9 l5,p4[1ĸ=X4nYeAR:>592 {QxnUG;CbavzaӾ+Ryn#;s똞W/NQ a4vXj_ X{y}S8Πm:'J':q-xee:0ӋP& )&/#m| :WhHFh"pJ ّ9OCti84*%BJ y'댶cȦlYN9alrǪ-Qs8MQ 1smu<@Fa?e܉}D8<GPI+48< ڋEگLZ8XS76^r {)pZRbgoVit68"R/^6ZDh2NXP[ wTyO1@~rcXvp8ޚIV5&M؄DkϷ;#/p4lz!O:1`d#+ngCJmն(wYeF8?',KV-V~prVP Y{BOm1™*@bl0U rdU]CI3^<]s#8BK--IqH5ڼXBY 6;KP֕Clv_tw 5ǟn{qVǧq<ưQQگ/a݇#΀޵ Bi7a&8wnwCWh8nu443 XiɗjӼdmlݟX,lj9XjAתICsze\ &R=?-&6Ch0]e,Qjr> " }  J ,{[Й_ya d_ ZLM;K?c+8ԻziWqz_aB嶲=LCb2gej>9L(QޟJkD/TLx/5l/Y68 XlV,đ& -IXn[#X6֦u;jX#8QZ6XNd@YʡbGJ&LO&YANތq.gFsbѲdpE*5,/Lg F go.Aϐatzc&:1۟󔹅\R= AϾg>A*1zPKgf&boCAoǖ;{lBeFTxssz34c0S뻧S< 󴑋Ag7sj C4ؑV20]X@8 0qa30gL1Y"FGVԅmd;9cQٵ 8zs:ca=;4 ?"C+K-KK7d9glLBa&[kubq~uJ'r Q&c5Z*>&Jyo)%0Z qIAE[ g!LQ䈟1';QƷzc{ _ڑm6zP!oOj8Bu -=?i;_؇x'L^;n$ rGAޤ0eaqhdqTEs@&m-a=W{D5cbGq1(9˾p'(#tτM}=a0 (>e3cV}}s\ A~fSa;0"q?lAVb#FUs5ἇLWMHQ˶g3Fik䊊 X5m~j) Cㅅ^=Sh8x4i.zЊƕݬ4#:7|@ag;a 3 lD’Fڼ24I#6+Ca:рc-gq , s-!؁9[CFPz S` ?M3aM<)e3VL( JI%X8Ӎvjbq \;h?;mtbEW 'g=b?4/1McÜ#`8ѳOQ&?2)~t<1m;ŏz3sB#_TWt3vWmZzyA`gyɡ&~!(AԔaIM jsxd?ドY]~XqQm짺Ŝ_^OWۦaȓPMOJdBķG S Ar>JVFvp\,`p¡IAU&S#;n_e`ngĻ%Fs+x>x/GSPCL^ˈuh*f?dy0QEM YҌuK'Jepê# c*U$ 7W(((P XaU}a?lgN;ը^~?ϼO&q}.4A |6 é%\ “Tq3Lt4b.H~&a{U Hqs-iPG1kԃ@nUz.Y/:|«-^!-~+65]zxijET*0w9?5zᙿdPS:F^a/gҴy}cHəEm"0iXnjjZ h87G-]X#ݶÄO$=2Up6*7A6-0=z~B3B, h^p^IC̲ըjnq2R uS|Ex\$AX^!2LۓcZZA;at!-NﶀjH8LS͚OL0=\6^x L!ऱϳ6f4 Z9`ADT |LYb!i ɰnֈSTS~/O{T CW5=i3|D犟uzkUUx͡e.拫[]ㆰg-C^pd>v}j숍܂^~vZwZ}85zz!z r<98L}^ #^Naz.ʻ{Mַ@ah|@$Avwq.B,x ԛExFOEH0٪1#gw7§[Cz wNu'<5Tkt{ܠyf|#>1{k _F/D#B5&u<sՈJ܆6ƒVu'&:Ig,SK}]-{YG֣NXo,E(!悜``XX <O GLPC"\LM9B9`5L?љ̥恐[9!GcΟ?wP`5۴f (:yG}s|`J4S;¹gƌ?;?& o;Zt'x'-.cfJŁh;kD$ ӥ(uP$.BqnГN |׿YfѨAi5J oP6\ھaz|Ve$vlj*ihtb(6j ,_MYU6/g[GHIpBXPmBPbm[]ʒƚM;rg\*)i/IH똛y8:wNqы7j\^ّ]Ǿ|`C}u%mMV-I?~gys:"zSwL&Tz}u%M%u{ tdvSS<5aAIAwL0bXʨpG׬**%H2LIufaA `eJ]LTTҚol۲,^XjpzW%r9L'2OWI4nmah49Z!E\^[`qQ- :vn̍VFeU[T33We M5jd\ˤ୰I6P t؂6+n1֭leau,Widx8)51K + iD;0&Ҳwu]j]㸎=s ~Ƕm8hEL+ذUu_lo]Ս@_=Tv]8.9`ʹq?HBtvG?z>1ڍ: yw~,CԾ 6I`X9J,+:izT\*f,s,r=TqG16sc^ | { $CPrVamvXi@XDNR l uLͦQ/T4Ji+QAtyx>9y(t!$-5<%܉a:}wc?JSPJ*{bbҐc'(eS%UlՄmI>nXc ,8}^@FdƂG'<}ǁ5rOAů*{ޝţ C DHK3 *C '*UHHL*_ޒ\G\)-φ^$,n"w0`D qZ#E7g`80l46 &(/֨Z%`F q_mWk3q\-cخJ9Ew[l%>O*bTZXWL0M̒:Q%1GDŽU!.,k$Ba<5Np_p ,>#Ow~i]F sU,p |q) fXa ɰð:tGhRNeht֓o1<K(0 &=),ǺƀZJ$u؉ߢbp}\aG9 ,{: iP;V/x8=mL=?ΰwn-09vw5&;}k[ٶ Xm ? SV%.60MPux *U])&\Z?Q,|+׎$<-`66H ڋ`??!Evϒa9>qb"/ ¯2>J`(u@ث D#џVOI {~&n 1T(!BoB'wL?M܇y%_+ZÂ"L[ghTBc:@U\fąɖ!L6Ƕp،0VÍ! Vv 89G7$Zh d2U1ã.||O0#uy@%22!Ha.:/hUP1*0V&xª#{ ܪ jA X2z%Tj{$UVGf<_qa„mS.>dLuȫ'a%;Y9%d(h\XȄ:Ag*P\=Ҫ|5TC}B=0!)T3xgwCX<Fk¶EUض `Q[;bl5ʴ{Yf0N/'DsjS3W'pR`@ۏ81U!_,^oac\$D 1 \JVD_RReYz<00Kh&8r x? KK<*1! B4x*xXy35DT1IZBm +` +6WVQ$ٙ_uUH k:̭X l#1V#22BLuyb%\ɑ*u_Ƅvv-6KXץ2cVk^EuJY[oAeo= TZesEQ3R1o4AU)芳;vDTW22rr7^aZY(9532ZDm8$( =LWX3*-L=ɴ@&9PIm49RZb\]F9kva`ۘ*8\81&VfLLi0Knʓ%ۡV)"iSdia g\mU`_,8ʘZmupbX%L0 j1 \n0xXLi1dk"gv!t.~_QG- І#2I"F,ȫN$.¥"zZ1Mi pZ y$-[mJ6nZ_-8=JJZ^,]K&2@E8e:t&r* D.Lնg8ٍ! 4K1Z~P#e[-]H^!e sog`ŭ2Ua%5cÒZmRaY7LDJ^"%P-آ ϰi=(SALEi߄6L? `ZIbp=Xc eQ|EG-z/LkR"aq()ږqp*INa~⦽tgk8ޒuӍrD*hcYdԖ/ PzyfhQbY?ǛokAqd`zh>Dq~ ~w,J【)BTÈؕ&;XI<ܔ.Oߡie%M,)<~3J$q`#aW>|nuh]+_ ^:~C9l$O-l71:&$!!]OB֛iEy j)Qy̫sB@i*tqe]ìV_L܇~Ga$e ?PEH+LG)m&uIBȼh8ͅL(@8ԳЊޅ4ֱ2em?'(+ϿAe: 2. z\g+t>JB"/ |sIJ\H: ^ړ̂j E:Z,̳t If+5P3 Tx~;Eb~{Q([&d[}@I D<\~f˗ UOӑN"St>/ _O=m#n9KEH[!jfr`[Gqg[ ,UZ?pg"c\5X} [w^h1/zoD),lcmsMn>v8E1g{"kiŮ?ZbͻtҒ2."@t]N1$V*j 9Ndw٫y]y;̥DxN&(Lq|jw23;cbF+W^:1}PUV/&AX V1Cif`Ƹr"'S QD 9M~0W6 "'/I43DZ&!Mq$mew}%VC%Оnωfw(jϼ ͿpyvvʐjT0xm  158=i4Q]{K_iqa^.-.^???=?87jmtz|sy"ȭgKxbN;y5 ^ǼKaPIENDB`DyK uszkoreit@coli.uni-sb.deyK @mailto:uszkoreit@coli.uni-sb.de$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 IDd M<  C AbIox+]"W HI nHox+]"W PNG  IHDR![sRGBPLTE ("5)3++1,IFZ3TT-g1k9c9k1c9c)c1c)c=cBc5-[!Bc%Jc5RZ9RcI; HA;NPD6Ve39@#;!!) . '<&&B>CXN8SZKpZ2xpEHVYVXThZZnmYoAWFZ/[Q3Tў%ϦL4o ;n;qDuAt)Ep P},Py7N{2ZJPzO`~JiMiRsZtkPnlonpl  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvxyz{|}~ Root Entry F0&}re@Data wWordDocumentaObjectPoolIre0&}re_1096711384  FIreIreOle CompObjMObjInfo  FPBrushPBrushPBrush9q   FMicrosoft Word Picture MSWordDocWord.Picture.89qOle10Native ĺOle10ItemName_1097834869  FIreIre1Table      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~BM6(p "()5++3I,1ZFTT3g-k1c9k9c1c9c)c1c)c=cB[-5cB!cJ%ZR5cR9 ;I;AHPNV6D93e@#;!!). ' &<B&C>8NXKZS2ZpEpxYVHTXVZZhYmnAoFW/ZQ[3T%Lo4n; q;uDtApE)}P yP,{N7Z2zPJ~`OiJiMsRtZnPknol}lpnqt~quzj|Z}iuzzwuëƹενι÷8`JcJkJc!})! !!!!1B!!!1))&75FG51B=RgXwNRk^ZZVgcmwwmx{z 6~66! kkZ i ! ksT ~66k~Q ksrskskskskskz}z}z~zzksz}~z}sr6zzzz}}z}}z}z}z}}z}z}z}z}sk}z}zsssz}sw}kw}~~}zskzkQ 4Zssa! Qizzksrsksksksszz}z~zU}k k k Qz  ! 5 iZ  6665 5rk6 4666! T~!UkQ~Z6Ykk}zzkZU kZs k !Zz~gU ssgQ5z6 ~Y U~z5!wz! aZ Qz QczU 6s6! 4a!g5~k!U55ZkZgzY6Qg zg i6UZQ gsZU!kcs z!~ kT~  gzkYzQ5 Q s !kQY Q!zT! U !z}UQk5 w}Z~!5 kQ!5Q c5!!z zg !!~cT 5 Z5kc!!5 Z ZUk~T!!~Z6Q 6Zcizg ZZcc6 5 ! U~zZ ZQ!}zk5 Qzzi!!Y ZcZ}5 s6z6 z!g UskZUc UZ~krkZ Zkc }5 !z zr k! UwUiZ~c Zk6 { Z6Ykrs5 }TZ6  U~! Q ZkU5Q Z Qk U5 g U 5Q k Z 5~! Zoa~s ZrQk665 4 k k T U!kTUs~ Zk k Z~ Qaz5!5 ~! z~k U UkYkQ~k ZksU5 UzwkQZk5  Yw!!zk!Q U 5 ! Zr!zw  Zgc5 Z5 Z  Z!!}zk!Q ZkkQZ Zk6 5 Zc Zkz }5 !!! sk T U Tk} Zkksc Z!Qkk5 ~QQ c !a T Uz~4gi Zo! Y ZzQk Q5 5}zT 6! s Uc! ! Zk}Z! Z QoQ5 6Z}6ZZ s ZZZ{Z! ZkQ~kk Z!QkUi5 U UY ~{Z5 Uk!ws c5!! !!!5 Zc 6! Z}}kQc5 c kQZk k UT !z~U ZcZk5~ Z6zrQc5 T ko! ! }5 U~~!6 ZZ ca Zrk!g5 6ZQ!ss}kQ YU UsUs5 Zikzkz5 !{!Z Ts rz6} QaT!ZsZ kgk!Q! rU k5Us! z Z QU!ck ZZ5kUg!  k QQw6a s !iZ! 5kg6!gkzc1aisrg! Tigc15gcicicc! krgQ!gchQ  czka QchZ! 6~}Q c~kQ Zk!!oc  ksiU cis !g !!54!! kQ !!6!! !k !5!! w           k !!45!  6Yk5 !!6!!  !TcU!  !isQ   Qs! !45!! 5!~!sccccz~kcccikc~ !Qc}~U6 QU!!~~ZZZZZZccZZZZZZzzQQQ  65gkk !izzzzzzz~zzz~z !!QQ6!4 Qg !4QQ5! ~4 !!6Q64 !~ !k6  5zZ 4Q64! !ci !46Q6! akZ QQ 45QQ!! !5 4Q!! ~U  4wz 4!QQ!4 s! !!5QQ!5 !s !!QQ6!! QZQ!6Tk{cQ !TgkZ56ai6Q  6QQQ5 ! 6QQQQ55kQ! QckkgQ! Qa}6crQQ !6QQQ! !kcQ QZk6 Qg!Q!c~~kQ5{cQ !TccQ55QgUQ4 QQkk } !Zz Q~Zz iZ Q5 k! ~!kU~Q }UZ6i i}k !k }U 6cZQ 5 k~ ! ~w iUZa!UU zs6g{kz!kQa~6Yck!T6!}Z~ U 6kka c!w  {!ik! Uzkk 6Zcz 5krz~ Qc Zi}Zk !5sc! ~!5g!Qgg}~s 6cc Tkk~z}i !kiZ  rQs~~5 !!icT6z~~s 6cc~ Uzkki 5! !~k ia Q 6Qwg5 azgQ z~s 6cc!U}k6 5 5 U iZi}cZs5 ksi z~s 6cc!U}kk U 5!! iZ ! aZsk5 cZ4z~s 6cc~ U}kz ! c 6Q ia}iQ Zwg 5 QzUz~s 6cc!U}kg c 5Z z iZ6 {kZsZ5 5QU}z~sz6cc~!U}k!! c 6Zz kZZ!!Zs 5 !Qk!Zz~s cc!U}r}! a 6}` iaYszZws 5 !i 6z~ss6!5cc U}!4 ! c 5!5 iZ 6 QZ!4 5QQ6!6ai Qz~s!krccTZZZZZZZZZZZZZZ U}6Q!  c 55! iZkkcZ6Q! 5 ~6T6 sz~sgU! cc~!U}ks!! a 6U~ ia5 aZwk 5 U az~s zZcc!U}k c 6kz iZU!!Zs 5 !Q6!iz~sQcQzcc!U}kg ! a 5kz gZQ}kZsZ 5 ks}Uz~sz! cc~!U}kz c 6sg Za ZZwc5 ~QU!zz~s!ssZag!U}k!! a 6 ! ZZcUQ Ysw!5 ~ Q!z~ss55csc Uzk U 5Q QZ6zkzQs5 ~ 5zc4zz~sz ~Y6!c~zc~ Ukk! T 6i} Ta 6QwiT5 z Q !!s~z~s~~i!QQ!ck~g} 5gkZ ! 6 oT}ZgZ 5s65! QQc!k!g}}~~sz~ z 6Qc!ci Qk 6~ kz6gZ ~5U!s 6 ~z !Qz}c~si~Z66cQQai YQ 5kg  wkUc TUa5{ wQ6 6QgU!!k}6{w ~ cQ{Q6w6c{c 6gzrU ccU!k6~g g{! gz}56w s wa!w }5sUUk c5 zQ   za5 QZgZ!!Za!  5554!  gZ! UgcQ QczT! T  ZZQ 5YzU 5ZgZ!5ki! !5554! 6ia!  a66i~ka s6 Z! QZz  kT 5Z~Q 4kYZ`ZZZZZZZZZZZZZZZZZZZZZZcsUYZZZZZZZZUZcUYZZZZZZZZUaUYZZZZZZZUY`ZZZZZZZZZcUYZ`ZZZZZZUaZUZYZZZZUazUYZZZZZkZUZYZZZZUUYZUZYZZZZUZsZUZYZZZZZZUZcUYZ`ZZZUisUZYZZZZZUY`ZUYZZZZZZigZUZZZZZZZZZZZZZczQ 5cgUYZ`ZZZZZZUYZZZZZZZZZZZZZZZZZZZZZZZZZZkUYZZZZZZZUZzUZZZZZZUb+.;'***&.;+**&*&**&+;+&&*+:-&*&*'b+*&*,-(&*&.b+*&*->+&**'*&*&-<43&***.'***+<43-+&**byvvv{b+**+:<43-**&.xvwx{y\IFLX]\xyof]VOOMHMIM^x.***-<43,***;^VIJIMMOV]]nvl]\MIFMltFMXynfVO]]]VVOSMMHHHDBAI]{;+&&,1<-*&*+vMFCCDIHMJNOSV]]SVV^f{~tMIAItIILlxy{nOVMJNOOSV]]]SOMMJMHIDBACIIIIIIIIIIF&&*+>18-&*&"BGIIIIIIIIIFBFHHHMMMOOV]]]SOOOMJOV]y{xxXIIFGGFFHIGDHHIMMMMOSS]]SSOOMJIHHFBB"*&&-3<+**%ABDFHHMJMOSV]]^SOSOMMJIHFFIIGFFkAABDDGIHMJMMOVVV]]]SOOMMHHHFBA%**+>7-&*&AABGGHMMMOSV]]]VOVOMMHIHGHDCBAAeABBFHHHIMMMOSO]]]VSOOMJIHHDBA&*&,3FLFFGFLFFGFIC>:**&%ABFGHHMJMOSV^]VVSOOMJMHHGGDCAAB<=ABABFBGHHMJMMMOVS]]VVVVOIJHHFBA"&&*,IyF\N3*&*%BFDGIHMVVVVV]VSOOOMMJIHIDDBBABEDDAAAACAAAAABGGHHIMMMMOS]n{x^MGC@A&**-8NFLMFIAMIuNF>,**%DHVfvfVVOOMMJIHHDDBAAAACAACDC=?EFDFfy^OCBCGDGHMJMOfy{_OFCCCFO]"*&+xuGy3:&*'P\LCCCCI]tvVMMJIHIDFCACI]lyXDGBQeHGIvm_IFFDGHHM]x/**,>LxOm3***{nOJGGGDCACFOmxVHFQ?IHHn{xmMMLVxyvf^]\V^y.**+<FCvv:&*&vff^fff{yfMLL^xx{OHHKՕ?IJIVMFIIFNvym]V]VOMJHFAFItv{.*&+8lAx_LAuu>:&**xvXMHHMMV]]SOOMVxx_IIIGIOMIHU?SIJMHA@FXmtVMOSV]VOMMHDBCFF%**+7{IA__FILFl]F<:(*&EFCCCDHMJNVV]VVOMMHIVyxlIFKKMJ`㖙`SMMMKC@AF\l_VHHMMOSV]SOMJHFB"**,1yXM]]v^]M{Ny>:&*'BDHIJOS]]VOOMJIHBCF]uy^MAGMOMMRj?2eSNOSOMD@AAAAABDIMJNOS]VSVMMHDB"**,1{MyyG<&**BFHHMOV]]VSNMMHGFBAAAA@BIOVSNMZ22,**%BDGIMOfntvnfMJIHDC@ACIV^f]]]^c8<<Gka^^ffnfVMCABDFHIMNf{t^MCC@%*&:<LyO>+&*"CFL^ntMGDB@AIO^fnff^fgpe <=xIA=8 0=qnnftntnnVMDDDGHJ^/&*,<M\&**ovIBACGV^nttnnfnkqĹe2 022=@y^BE?0p30rwvmnmfff]V]y&*+>MMOIC1*&.^VV^ffgnntwrsĹe e082?UBDFJe rkff^]]SVS]]v{++*C_lMNC<-**z^]SOS]]f^fnksļj=38 ?=FB]fCGQ8 3jj!Ujļqkc^WVOVMMO]t{**-3IMx>,&+xfOMMOVVVVgij¼pU!3 =p ?=GBHODGQ= U7!dqľĹjj[SLMJIMOf'*+>Lx\\F*&;x]JIMMMOWdjľ˼qq ddj j eHDGyMGHT = jj !0ppeRKIHGGOb**M]x{n8,*'[KGHIMK`j= 1 j e 2==eJFGnMHHMj2 e = e?RHGDD#-8IOnC>%BDGDK`?e=d!!U= j `JHHlOHMJd =j 3j e23 08?`FB4AvlLy]]]F71MlMmm<U<3 82e Udj `Ue2jZMMJVOMOM[p0q dĸ!Ĺ32 848ItlLm_tMX?<222<2e `` ɼ 0ļcVMMMnnOOOV^ļ0  j0e== >I\ICIlvMy\3< 883e jd ! `ľkSVVM]^VVSVfpq1! j2= =AufO\FMt\\v<8vDFAllIIu\N{C <<8? 0?e|wnnffffvxtnnvvnnnoqq<FtLuFC]I\IvvC2? 314444441444414441414141441414%*****&++,,--/-/:999;bSSVOOMU< 2?pZVV]]]acb99:/-/-/,,+****&%4<2?qc^^ffmgg;///:/--,,,+&+,%>83U3U3U8>>><7l{]IXl{Y3>>><3U3U3U3<>"-+&+),,--/-9/9;dgnff^f^k=8>蛙2 e=pqinnnnnncb9;9;;;/9/**3>7ny{{{{yxuVQTU=?=TQ]xyn\\O]]]{yvTT=???=TTQ^xx{{{{{x\773>+*./9;;;999;bgnttnniqq 8 28O^y{x^\\\vy{n\l]_u{\t]\x{lX]Ov{{v\\]jd37>,*.|f^fffmowľj3 < 2~{7?223 `of^]VV^y&*17>xuv_^uvx__]\O__lxm]ytl]ux^_\O]__lyxvl\l]\TK>>-*.m]]]^fisqU d =2  2O\@?2 03 e jqUqļqg[VVOOn+*41>\\Xyyml\lm{{tl\ltv__lxxll\lt_y^IdT>>-&.{^OOVVakqU 7= ej2 0==< 7DMA4<?= 3qU `ppaOMMJ%*Y^vyvuvuyQuzvvuvulyyTd7<<:'JMMMWjq!!Ӹp2=e?3 2=CDBFK203U !dd`3dMJIJQ_xmscUn6 a~wV^uIQ7<38HJISj3q 0ľ 8?DCOmFDK? eed`q=˸03==UKG6<<8xx_Q!rsgg}riZ U rtz!6!TigizUis lXhjZ<14=GHTeej7 dp`20 2`HDIlGHHee0 q`0 d33 QED3<_X~5 ~k6}{!kus~}1 ZQZ!}ssxtXXUCDHe23 e e70UUĸjRIHLmHIM`= d==e2220UA1}\l gY gzT~!rnzZciZQ!4nuOpT<B50=e2e26pd!7e0jRIHMwMJMOje` 0 =0 <2<=@hXtx!!!Z gY5 !kiYzQ~ZZs]u{YZ4<<  j3! ռ j=jdMMJ]VMOO[p5 =e?227amlka6ksU5 !rQ_s5 Z ~]]Yd78 <8 j=e2 !0d˾gVMMMntVOVVgqdjļp0= ? >A<2 8 38 jU`skf]S]ny{f^fffn|Ľ=0 2>4vNmzkZkUMv~~6QUkzfy_M38 <? 3evmfff^vvmmntnorľ32 >@ALIyIlLII<<<<< = ?nntnnfny{xvnmfffkp 8> >FACFL_nyxlLC\tyxn\GFAAC>82<28kgffmnnvwyvff^]]Sad822 3ACCCCACCC22 jd^]]^^fnxyf]]VVOOSZe<4> dUOOOSVVfvqkWOMMJIJKQ?1>>>?UKKHIMMMWfqpĸ̽e7RIHGGBD4<4>>7BDHHIKT=̼fnf]Sd'%"HBB@>A>7AFDJP'e`Vfnf|]gf]MF>&***%"<GmnxllxXFFnxnumx_LNlmv_vvINvvutnulL<1@#%'+**-3BHN^t__g''''PG@:&*&****%y_yxttOLu^\{NXI\x^xIN]nlv\F4"&*&****+7BKb'/.b/,,,,,'#@%,,,,,+,,+OAFxNM',,,,,,,,,C6,,,,,,,-,--,-$/---,--,--. CFFF\O\t'--,-,-,-,-%'-,-,--/-/--/-/--/-/-/--/-/yFICFxfm---/---/--/---/--/-:/9-/-/9-9:/-/:/-/\llFACLA{IvAGF#/-/:/-9:99:9-9-9-/Ǡ9999999999996OFFluF]Nv_n{_NCtF/99999999999999Ţ;3M\CLC]yNtC]LuN]nFANCnnXtFOl\tA>;9;;;;;999.-7M]yvXv^lNAnnux\x\x]x\n>/9;+*-3myM^Am]{I>**;+*:3vC\vx>,*;+*-7A{MLF>,*;+*:3lNA{F>+*;+*:7N\]XLln]N\F>,*;+*:3^mt^{_{LNL>,*;+*:3~tNyxM\]_>,*;++7ylM\yvvC>-*;+*3xvOl_X\{]\_lI>,+/*+3{l_F\N{>-+/+*7{MyMIFIL{IIF>-*/*+7\MFIx\yCyyGC_\{m>-*.+*1xlG_lmMFLF{y__vF<-*.&*1_nCFyAAyOuMIuNO\O_{{tF<-*.**1Ayv\MC\lltOyIOMxXx{wL]]<-*++*NN\lXy]tnN_FF\MLCCF\MmC<-&.&&Fxx]CFyLIllyxN\nvlN]vlMlmmI8-&+&*C^m{lLX{ym\L^{_^{MIly]AA_xlMNlXl8-*+&*FGL{_mF\]IIMv]A{yXImMI8-*+&*AlIGLtI\FC]M\A_]Fvm8-*+&*IMnClxI{y\_GvuML_C{O\^_OMM3-*+&&NullxCxuNt\v_nml_L]lXFmX3-*+&&ANX^y\vvIAyvFxvFN\_^{]FL]AM\3-*+&*\O\IIIIOICCIFILF\XIIIMNF]CAC3-*+&*CxIL3-&+&*AxGFx3-*+&*AlF]8-*+&*3>-*+&*>31@41<>-*+&*>>837A133<>-&+&*++--><3831A1838<::,+**+&*****&*+-,:>><<<<<834413<<<<<<>--,***&*+++*/.*&*+**&*&**&,,,-,,,+&*******+*&+/;âb;.*+&*&*+******&+,+,,::::::-,,+,+*&**&***&&*&++.b;b...++&*&*+++*******&+++++++)++,,,,,+)+++++++++&**&***&**+*&*&+.+.;;/........+++*&*&****&******&*&***&++.........;b2  !  #$%&'()+,-./013456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqstuvwxyz{|}~}qnq~tujz|Z}iuzwzuȰ÷ھĬΥέ̷`8cJkJcJ}!)! !!!!1B!!!1))&75FG51B=gRXw딟NRk뽂^ZZVgcmƈwwmxΈֈ{؆Ж֔ތޔӧۡޥ掠ֹޭ޵޵޵޵޽޽@ pHYsodE^IDATx^}{PTwFn{2i3(GxLC+ H3tk@ #3XMd0l"U#!2V7ޭ[5Iٺa?f;Vq~s 6~ŬG;su\ _P7l>cѫ{_kdZǧիG8 !䠃_m\v, WSkl,;nG܎}W^}O]ó{-蝝;X 3gz\o={__PΝ냯5[N#?*k̫,{4;Bff:onɻ_3ϳZ_ٳg㷳u^ܹF?sd'%SODGtQ(R"y_o]v41zIA\ts!\A,Y09I;:}!0WKsKbeL qbI4sI(NDgc\֊|bբ"V|4?M,'8sE$`zƨ(c Bk%T#)7f$veGE*i1 ?X,ui!kvE):).tn2X(#(Q5yh:dX.20C&rmھ!0WKMP;jw5Tw̚Zcuήr6O=݈3+&Xl0tfr `t]FZD:|Xt r2fD$&ON78skܚ\}3]deP \TS+=e*EcNX.)ę;> uW;`"v9]`(k#ʖb? +MCHt[@,D+6椬.W`5&Dk3T"ht8w4w G!e^A,"T=kw|` jٍ}Օn[*x`#:Dlpx~tX)5 fl<ۆc2V}.WL /s+~Τ8/D`{Y[` s~0|xL>~6 La*pшB[tJe|-^lܙ$f# | x}Ƿ".;>.2N8qŞI2.I␟%`Zh!0WKӸ3N\]:0i=\J!>rdLd)Kff6@Qٰ ݄[,Idc"YcA~'hf%֒5o:=;b7K\CQU\\innsZ!03ku0wM^޹gi)84^ZN#x Xs sm\[a['ͳfPL`ܷg7[1 ZOCF~KAU<;Rm]諩,{f Ƃb-Y Od~&'x$c\;q濿ybiqs֫]ff@dnck-Ijme5#Ã{{z΂ 5*kCՖbh^+;&r֖lTyȍč?oϓ*;<5F}{]bb+׆GF~>z./mv < hn;6vgt֍i27n` +]333ЛLxf|xȼb溴&vq{]mmSs`Ԯd0ݎO<942|NĒcӌ "\bl!ToST1#&֝ݛxA3٥t< lGm9dyrrÉ{wHQ Q2yXꢓ*bn4ǏQ& fP>x-kc&oIYy1]JGuD cjJn}jjS$ "8io*n36vGj*ڨ͸5@idퟧ{/|9iu2zJg2;eN4w23fj9\η><|]=2zMtW?9<W7d4 vB6:&T5:VFswÉh(X3ʸY385K]}O؄ljT%T'?e3fYU61]]m3J8sR_IYXJL=69YMfxMX[ S4)ӮU5nU{cw~V5^1T+="Tg*iTyeS)j3w-)HjVxa6Dn&ݟ*'Ǽ868k4);먬26RI'Z(Mc>vOľ_|V3 ghsfXnzԎ1)Xv3lqKw, pqnۏ% 7`Qc3T5YI?8p9#7OF>ue`px$IHzKM6ruœwb_g˴XC.zhmlrw/$7Y@QQM V]*%~ H)(/7f҇hwghk22s^ԱRk7PNǓb*NN'/gTs=Ω6QQpwu9+*lbhksWXm𩫋v:Q7֒ (lohÇ-11.f$j1/;Vq&Se;w(h:'/Sϙi]Sgju<̗_GVeamY ^g!h81?1}t>®ߕ&sĬ;w>^õMuuoÝxtn;oyڻ".;Y+-3/EP/fn5vN<7}]{c+z.WFnp{v9fwo͸~cUUZf")G/ceUGOÑ}/n {+'^lLӯ;zKM kֽm^˗/8 ;wc)3ajgz~#p|K!O\@D> o[eDrXb_Σ/[`>\mTy`2`緐vx 4p&b.~:,/\pi7w_3Y#7T^{[ i3 @! -9-pQ)v8Ll&=ؕ۷O=/ eZvxB)K,L-,# {dGζq&9x=>.nK^3jV%vHje\o{t9-ۈ;w|GOD~%.ed &/vֿeϠ1 :ʝ{~A!/'gSnq*ȃO둺f.ǟ/1<|qPܱ}֭[{O`z~U}x^xɸ?oOJ`,4hNF on>V "?}5+@AA>PLϿ(tuGGb#Mg3gm-,c ~ kW{ 93鞝uV›VM'\fp_GbowoPYrX %_ 3p 1KK[V]3422|c_ ,f+J&.w*@*rl^Gl/_vPu҃fg)_v9a To!QؚwcΚgAk^ N-JФ|b4dP뫉1iY5zͷ5aO4;\ ?٧P0i'#:拰n7o\-z[? m]$HYd/4rȕմ f_uYӳ2~1zu/BWK͊>yH0;`WsxG>0,o k^F|[&"4lBJ/F0C6nИWbTiYڍ Y5ݻćN>IVP8Bɂ&\}=oL@0) 6#Y'#Pnsz ^+TeuYy(. 4&)U⍟ee45c24I7a>Mi>h8A:i)3]]^g =>lR D&RQlмJo \k Yd1zpLAd}zȥ43|kV 74tR?c%PPS`?ߥרcN4P(fof xDIre1YO6jOۼ9 k[n%Dn6{VV>(BF80bTT \^ <kޯP8>f݉15ofP볈y偁2 b se27G[^ ӓb1o`ho5`X6P&+)XMbwnA90j7ꚻC7_P yyߎbM/!}$rwɆ 3n;P&`5z7AwgUO>|_ѿ)8Ai76ކ= +aGx!징"#xW&a6Ѡ'yW8VLX|7_$vMJY(]&X4 ?sp.añ 2:&p.MP5 QwoO`cfq^y&0i)}<~8yȉ64|`!4~|-6N>S^JH1;Zw nӆk_V7|ԽB0hUT5ަ46n +GQ]ȍp@{IpOe;7B_H'˚F 4(rj.ޗo_AAh@gc}z}1[F* 2e(ƦUta&UE^Lć|0уv6bAh=B{*iS[ߣp } rbh K-!D!80>6 > 0q!/-,NP_t+ZwCHp֭tn9ez!"U&x%[p䣕h /כ A [.=N%ٕ`4[ōTrygIAᶕ DE= [_IK{K^L˞tA7AbF0اAH߻9[#Y*K}%5c^d1>Кmq&4e9yF~ pGjَhUa:XLPa]z. 30ᦔH8, 3IQϱ7hrIT&+KT%B n?JQy vBU85i^[E)r&~HAhgfG쎢іv}pѶº ~tH\-pfNMqW)΀?amVn g%Fpy&sRAPYp2ҭjwBR,˙A@vY;*<cT/M 77X8iX*H=;l4_8rxx``w"M>2|W,&h(r&p7Qgucc|N='s K`pNM" x2vSnث_0ϜC7=Yqgigh7kd +'W2d`ƕ,cAhOh2 Kۇ-Sr܉d"efs%.>厞n:OzWϴ݀EAqjMjFtV+@~IA*&$)7LO5UA* Ē)J+/<3hپMB3{/>uWkꐉz:t~^n6J&'s3+\Q8)?!P;5E`ʋs(.'7е.GuQQ m{$Kz.CE(UgI8Ӄ|yk!^gcp&.Mck= tƌe? /7ͻ@Wd( ǾLƶ"ڗnsfﹳ+xRe JkhɆvI%rXAA^^H_4p$͸ԯ#Jgdwr&uRls7tB,ha/-/]E Rfo%LsC G;O8 RCJLW :I9 (C.~ ҋ%\al!9E@\S;ebn@8(6eqғJky><IK㴻mfg$>Fd4ΒĈ p8:/QЈq{! N0|ƭ; B{Ez `n#v-'ː96 f[JKA:3d*vwF'uUU.XY56%L՞ ɞ]_U vS YK* έ. )h R1 2eS}TP8oy&}hn5+&L_quyjjYWiD_|\.)d t)d t)d t)d t)d t)d t)d t)d t)d t)d t)d t) {ak.,-UV[,XʸUώ2|r_|Ǻ-G=e2Ҫ_;U/=hJÃZ*}Ba0޶-S-{e8pd@5*}Og~X(3E~_ۏDp5q+ӳQ2?zM8Pגi;~uaPaSe's?`{ ۨ;-v :)3*#RsQ z +U0QjGST#ZJے#(|RUAeK!UP1X;d{ m&ԓ,Q *&A^%/Pu+'f=? lr[&"sdipU@xѣW[HI]!(K.4/NS45X2{U-4) P0P*4I.Wy P·N|׉G22wS#0*e1Et$9%\DKDބ+[H?)&J  )@dIأ o"&*%',BE4?{^(&$<ƛn lT=!9U g4TџMk>BPF;Aʙ<0ՂId+ &ӂ@p ș11O2wsrX6B/6xu u3[h-^+OS(A3"8$q7yF_!w#Ř~k)ߘ ,gI}3 -Rkpf[I&wf.3kbV3y]8:3 grf_'i^S^̲^Fh;vRU(^7cp&;#ca)y8Ӣd\v:}Dk90ŬRIpbǮH8"r{Z67g L1Ո[]>L]0L\yL50G"8Ĭwt[nK:CwF'MZ6F0UKjazD d PJW2^n"T2;g $juzڬR>촛aoKV$vsĥT)f;I0'}Ԙޡs`k78Qj@a8o$fF ~;SON~+͗TD&EViЦ^#b*P̂C yYeSv|ukFMNގxRgzƧPpS&YҌdy#qkX ޮ-)0y.N~Nc }z=`*` Pva4Kih=`?A@̠su+ j!0Jխ,? *9WK3\B`.z̠su+ j!0Jխ,? *9WK3\B`.r9jCS`yC!IJQ`YLRܰkt'mx_L31Η3E2 ap$/#R `Frs.s'z!@8%iI%fi̖AS(qY0#E iԠ"bN}n|GCo4y>}03.{ Śk SA=s?"|&b $-OYG͢@^`04iȕ` Nwm2t UX)H:}O"K恵ENUWޅ |tN[t\o3L;+Vez5"SJSJS7Py|2tE_Dq]:`JK!HSb@>d0s6M,< ^d#u&K%^$\r#kҀHz 6{¬OcJrfI=G%sxc.x T9Ie#|dxL8ʢ$XęLᅓ"M$!3:H8w$c |F`oLKR\O ĐxܐY)ܹoΔK=-rOy+ Ab:$p{Kfe(8Z8H\"Đ$ "$%Z0.N490ǂ$rҽJÁ\\«޽^TlYsAv}F+\ȉS(:_'ۉ(be`%ZȞ!`4-LșIc%P@2! ¯>U*LbC 8~#y]wpZr_q$$OZ$.k!F+8s'W[gʃI}:B㐼(sS>͉^~XH?Pb${ZB~N GB,HeY Yf S&,ϋv1X?/|Z_2S|$434-K)c\A,#XXȚӫ&$^c}pDB- ȂYPK SR)8r'$qȤCgzHK녪 G[~T2Y08=h1~,LRX߭ IRp3XX&#TZs,:,6W)*cI \m>\Iv%X\@>g 2u-+wQtL#<"N*L;Gt"#T$a`"F,jd,gN{U@3yv`q&IK5Fd92;k&Pț RdCsp&wCdl]5@F*\fMLz襔] Lo" ɊYe- &PV&<0Жv.)E:o{Q4.Njy!N;ł ˰.1?;J.t0%{ |z @.< Y[FdRg ;=gI:]I ؇c$~HJx[ r8~.䲖k/죳-G؊\/v?UZ):7 L8n~)H13h^zM wf- LPB?q t  L0Л&J{+L5U=.$8י8Lvalq_*a<.҆v.R)hZtqIENDB`DyK valli@csc.liv.ac.ukyK 6mailto:valli@csc.liv.ac.uk$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 DdA # T  c 0A cidemb/pOz qv[ _  npOz qv[PNG  IHDRLPLTE˿ʾȼǻƺ·ｲｱＰ챦찥쯥ꫠꪟ馛饚餚飙蟕螔睓晏昏明唊哉咈䐇䎅㌃㋂~~}yxw}t|s{szrumtlphlekdkcjbf^e^bZaZ`Y_X[TZSXQWPVPUORLQKPJPIMGKEICHBE?D>C>B=A;@:=7<7;6938372402-1,0,/*.)+'+&*%($'#'"&!%!#""! Z9tRNS0JbKGDH cmPPJCmp0712HszIDATx^\k{E .!$vƬDFF7j@]&^؈ $\4$sS}~y82O:?W=]}4syx.1jV\z3g`~E8=P?{F{W+gV[FW.UO:Ջ4x'g-˳xhɟDO>6^`W~'g=mĭor]Fk*'Z=ua챢>6<ܶy;x/%w-+cŗE4Wo=o0^/jJ;,^?Q` 2[_=dɍWߣ; K˴7g4RKTokCý ?;ɗ.x[$py+xOFTAgݞE+4ff}Fi\^u\#OsٻI`c2..xƷyOA"]#;Dj忏+hSa (dG7yqGgcŀd'̡|XI`WB^v^S|NXb+r"+_|l2.YIq½3ryJPSU/׷XYQQ$,9K^ӮUDQGyǴaqHMi' 4{n/0s`t,/V5|'p_>ڪ)O+ш8^ŰS4=g=ś|BI} Ӽa^/KDl€cfޏ%BP)榯eҳe9˔r./RHDÔЗTkpQvنgαdWaH!LêFfEk>0UdMM=4?0K!Buނ:#'4"9~ K*Ys,(䮯&0k䃁Oއ^v= `ÍWV!aozYkQ_DU<XQ D?Q⎦5x47^܊ecƖ˰30q9DG_A,VG`x7m /`n$ qm@-׈2U^W>ZD A!I9Ei-1E\x=8$G$K8)OIi?GnPt6 kYÞVGg:Lr  #\xe[B5&Ċay6kʧ|BG'`ZL8e< Z,W"D  陵btȲ^nqqO^/\P?DKԴAɒKaFD"!~Ao?[_n6]TM I%J9/Gb W_'~9<`t Hr Yq`^,}!qƊ%'-d&G ]{!DKqgế#PDtNxg>I*7$3(>jN {Ҥ,+2~k+H™!llXbpvnĜrZ?w@x`3~u.w7$43;&B<^g!M3ȴ4^\ظ ^оr7 |Z42#lm{ sCf>Kqy*؉[]i83 ^(_^XvWd\"¤[5ՒFb>bcQX>}P|c65JOD/F(ttJvS)#L (wf:FhKQҙX8JJ\w}F8:9_T/WR ņ'MK< b+x!gK𲟃ѿZ(1LIP_d{Nx<(c,b4o>8iE xԈaō|kU)u; Ŧv,[xqӹlJ~Vr5pJǙf돚'pqVeݰ )2l5=r~]}88D2{m}Ųnт7E:"*~hL^ytXXtΝ+n*+;++XH3y:N_.xbq &pJg ] \]P]MN[M5LvmsUVe0(QE)kM;gm<اӖ4 r&.BħE W6i}2nuޜDOGZ>fj N:[qFb ő8nv g^f0FxK?`V i,!<)v`=Q\<>N;+#~Xy{W+&=yCfX`/'('j)@by A5:~)Ym2ot!\(Nل._K .印/'ܹˀ5O nӗRX=eFra{;z5V/~^{5\`j34܊0 #4M&GU)~L/uE"yWYԚfХ3RQ:Jo7\l1wq,gOUхPR&7sߥfsJrv"54u|k[}ZUSܜVSK $ ̈ҷ%hp,.$$?2:,vcAw [7| o#\H;ol):<pD}"!E.[bGȳ@](c?6opȍ`GPvzu _㌒6td'3>PIo$KbMYクAq:+cR"=G%XRj=\yDm>Ma^X;A> CV Qf%Rm<ݞhI<„$_T3Tjgk4Fu3+#e}S G!/vI?蛕_b.K.R=2M6ހg\.G9&/tAh4&huCǞ!dŖ̪I~Jӏ׾*MJjcJvAK=W6b "ȢT냭QϹ  &݂yЄOK]*4C&ufȮrth6xQ(q4'wgL& (wyKΩcۣsk-HD=,"frWomy&23]"QmAcmu7zؔ7X$ ]1vcg~Tj\"Ef |WB)њNX>~}"." v*jH") 3(_qQ`Yz1p/8kұ))1,7?(6MݘxT?w*_kMRxñ{}Jo`~Yq$ HvOwS\ޣ 鵅cYEgm(^zDp-+VG.KyTF}q) R}^P8o:K(.m2]*I_{4/T SpN"풯넘cLwKވw;6.T^l,WA]1=C|w;JĝnM O!acfF28='\TO7p{r[. =bYج9G!Ar*%:H: W]^ZMe cQ{z(1hpWɭBINFN:)n;/[:0;ER^fYu0gj Vt2=e 5r<1Sc#9XЕ(ë33byeF*#Pp:d$cdKD Շ/H6 Gח3$x xQgjѯʥ.r>\t/̟8z|hz|F/MKdt fʯtuOl|(vkXPGSbKc4rD-ty {ftëO;2-AȀg3[*e]=#Dq/$:g6";^ ;|rYK%mVT3[k0cOw[j&? 3ſB(]~6^ihL^A=\*>\BɁ׫L/Rݘ+&Sk훷WCf&Ę/݊IaW!'^*;i$ۚA3lw}ٓ#tLo+xx"b{m9,N9SwBoEAK,/Ֆ+'^yy>r8wכ-R/*׵󽸛ۗI@tm3abn>\V#zҊ*M1z4Kur߿L:QY+WwD[kϾ ^W'od}V`N`zvE|M s(pROoRw_,EK߭H%+L F=]]xgO NVFJ\EъRQ2<ԲSTY*o459D>Rx }Њ%*E@1t<3גxθL]M=2./w}(wmY\Y.W絓ZmU _p 6L={ A>,[T#n-Bo[.O$7&ȣ=%9+Y?oC/j7ne]o_Hw+/%AnmuߌL+uč}*6d mqi#=ex bVopP&+p/8PZuԍ_ 5:=8Do>+zFHW Q/;9޲+79-˱rʎ~߲+7YobLIENDB`1%Dd$^  c :AresidenciaR$SŎ]rn,NJ[$6+ FS$SŎ]rn,NJJFIFHH Photoshop 3.08BIMHH8BIM 8BIM8BIM 8BIM 8BIM' 8BIMH/fflff/ff2Z5-8BIMp8BIM@@8BIM8BIMq#anima_00#8BIM8BIM8BIM |pP`JFIFHHAdobed            p"?   3!1AQa"q2B#$Rb34rC%Scs5&DTdE£t6UeuF'Vfv7GWgw5!1AQaq"2B#R3$brCScs4%&5DTdEU6teuFVfv'7GWgw ?>}d}'UN&=Y8~VK\Kl[>v{Zѽj*\'}zj]0>el*ymn¾ױuFii\1zӋs1zo=)_fn冀P^n?1}em= G}9sk*k% ݻcUuQuի/qexx'X?je:eӖ&a1{<_݉:2!_YpzfW:v^>9'F!^\oZ/Rܺ5]wSc[~?Z6fkn /p,u1=E'7CFϑNYGx YA$W&xI/<b 1Ye4?gȧ~WXӶǚwh\ Z=Ap[Xީ[X0+-h(Nή2gkFONkZ-rN:FG"54ZzKʫ{anm\mwƙ,w^wn8+u:Jlzz?[p.}j0rMY-Wߞ\'C$WW9r㌤g',83L+=X,;׿Ѭ}Gu\ˆ%Bo5ݳ~"D3o??/?>ctՠgcﳏ` DoRqwF v Ѯ&o/xI?ɲ޾,_~Q-<[>mG}2[M!߮zXa]SOKڿJI/pv|_8 ?rϛlv=>;F77K)Zߨ2g?f G{z^ٿ24{6_ٱߨ@s$d}w}{+=>=R~պ7[/I.*/G??@fYgPK}ȯm'_O/y^߫v_wXIz޷`~o)'G']fK??8BIM!UAdobe PhotoshopAdobe Photoshop 6.08BIMAdobed@      #  !1A"a#Q24uqw8 R35ƇHx !1AQ"2aqBstRr3b#45ESF ?Vۘޓ]2VV-LbGmd{P_t-YUO3wBָUā8*'gEgLuZ?3d#UP}u%DU^S9Na$1A =]lZ]pX*8 ŸXVZtRn%"ڳTzfDŽHD^9QJZ8TΝ|0!]tO eujM4r-w%(ڡ}\щ YR`pir(;SR ‡9!fk К" H LvTrcyi=U⼓|mכÒeW1lT["^8DOݤʻ3#B cǙJ+iw~gCKmE )5ݐir2։/!zVQh#Ʈ.+Z*@CntP{dnWWeZ+Ymeie~$~dt^@MxB"sz8y@փg黍'46ٺݨHG;Q28S+?UxyI %6cOJsaM#͋m8%T^"/[{EhׂT. El:kow%haCf -u:p|?֙1x 4@$Opٸ(H S&F܍gw=MKi.ŤE# Q64rnI&*`6Fap/I848#~׵!!u5GiNb;Y$s 'k@mQIA;B}8eI-.ÐS<+Qtbcl$N,h=#j8siS0Zƚjbq["10UOB>(Զwb0=AI^:G5hZp/ )UK]Fd5$`W$JxŶNTQ檽x8 sI9h袎+To_cU6O6ZE# 74Y7|/Tqhc`}X \Gsp9gMZ*QG6ዑ&aW?2W@h{p/_xmHpnMPy+{c/p(u82!2VZ xMܖoUdTN;!l!pe}k( 0DKvKH*x8.Bnܯ=2mO ƵW߷VR5^DQ}R%DHwF;>ߒկn+ 7ţG}Iw;s̬iuUjBˊ\puf8# |#|/EʧMuOƚXUO ZKhBFyd];{wcHq52b~"I'6>Jf !;*+ljN, 4',iˀK x$q(2L?HS2rdj 9z{(˱";FM2ZUL#Gf.$2$wOf6v]_k{E 5(MqvhTEƒ4 IV5=ͨ}0 gz "rFJyNwZ+#ZhFמq65G,sp8*~"VK.q WӬnb/akutIxQM77k8͈u礜͗ fX'kd>}4wu DqT w,K2wJK~\T͋qPh$%x(/:v {<֥: Ž= F~Ok}uSk('6,˂ۄ$),DP{UUU:e70ā[.O7psu.B;\Ubn%Ka?ȸUy+g> ɐ)vAQAWgc$ 'c[]iڮ -o KqCk>iP)az.]6eGobI5h~(t;1ʷAqÐ']ӡlg;ÄlLWcB=Unv~Qnq*"`9a!XȨKi"?RQ_pW2inDqo'km7MYlۢ%3qĹAU_;}Ǎ?{]AXUm{(nÉ *@(Dʑ ٿv!)a;)72٬'IsWr5M~DJ)yN,mКAlD]GGNfv1Wߖ.zd软[+IRHwʟǙ`n m4- )(w0\Og&'Gn61Dj XZэ؞vI]{,*T3di j@8ӥvD-OWnr倃 E3TV{z/els#^Z(TcF\jm蹓PEFŢ;u%{^S2K\$'jMX_wpQc1:ꎗ6[.{B0'WւY_To_cUcrIrM&,Y2t4;$#E(] }WҺFaE|3n(²  [(@(NormalCJmH <A@<Default Paragraph Font  l,b$j}l2kixG @n( QQQQ 6   AB S  ?Y at@ Hb@GzTimes New Roman5Symbol3& zArial"hUkUk!r0 Gebruiker GebruikerCompObj hObjInfoObjectPool IreIreWordDocument e)q bjbjt+t+ e)AA] YYYmoooooo$^RNY7"YYYmMmmmY mYmmmmm 0#3c m  jUmHN N!9"9#/$0%n?j}l2kixPNG  IHDR`gAMAPLTEҤ7֥@ ` @ @@@`@@@@@` `@``````` @` @` @` @`@ @@@`@@@@@ @ @@ @` @ @ @ @ @@@ @@@@@`@@@@@@@@@@`@ `@@`@``@`@`@`@`@@ @@@`@@@@@@ @@@`@@@@@@ @@@`@@@@@@ @@@`@@@@@ @` @ ` @ @@@`@@@@@` `@``````` @` @`ࠀ @` @` @` @ ` @ @@@`@@@@@` `@``````` @` @` @`𠠤_R2 pHYsB@#txIDATx흋r(EjݨOA}=Uq=Qt4Hf\=fzP P D\A0I?{h,5 kNy5/`殂%Bo킕I@0_>,"ՓB"8A05YI%I* 8]TB0 VXQ%&XA0jPM023,ϕGCLr9YYzO0>-Ke%66d }c0VBTwMAWgb]}lJ` a` 6'}grʹvwm"YfW&ls¨ c3&91l`р`# ACDfv d`3* X`4`= J_%-X gٛ~c` E X iM&*v;S[?t,iJDg qE`K{-y ؋wV;`aMOlcUo},:{6 [h쿿Yr6ɵ(`BA6Pl1MyJgٛp`mqQ)ǒz-*ϰ;VԂ`\.Xyv3d5s`(`b6 F '߅$^ }i} F _;Z;n,[ަI`r?JM0{F2y+oS0P6 GfT0=^zp Meuv =U$;ʥ"KuHq@GE0+p{4``d`, 'Zc#.5зٍox5 X` N @[`1@åmjMO.Wq#Vus >а X "4t/8 Z!-A*ܴcxd hG}G &}+.uU5BWHbϙ˓;^E~&TQbc)^F`?v@0>O\0"C`}v`mX1X[~&XZWȍV:cc2\I~ .& oSￊ`ʍVzkI _ $߱Z@t%'-$Ral!D 61]d~sov#9EFE"4By)B(0?,߃m $p^/غ|ۊC혂~+f A FE -A> wL/`̢68YpY4hUSA0*C/  N`'[vځ`;`@@v  X;lU؇`,ځ`;`@Td{'I`5*SNB0`4V@0FjU*4r"#!kzI m6 Av `zr>'g^>5шT'a5\4Eޫ\*l4rV *Ky^9f~,-5xw\˩_^_gZ ~=wm^UV!ժIencnc0v!wMK_5:Ov)W ȫ :j_KGՎj)K=ibN%Hjq} `Gvm'n-ؼd!R4黬J?IMc{`sK5}`Ufjerj!ƆUVrj!؇ƇZբVrJ!v   X( k_@=`J4 ^P& X`i97B#o$Wxljj#c{z X( 6[ U$`Y q`e i`+Eٰ< kA0vj%؎#XzLrMp6E/ɝyMv +` X?L\nS@x9= $%cZ- ?]0>ك)~*c1K7J(\`C>첳T(z#j;\-lIڂݧ)-Sd d+SѼAߦhLR0 c^EkRH'Kd yO8e^S maz fׅ`!Wx{MII>5Tg@0 `@03lk@ՕuQ*˫օ`օ`'XI)T]jXcVc`ڻSv<&"]L*xQ]`g 8`?4A>8C6~F&Z+onx2;EB0Ps  ILC0@( @( @(LosǷ~&XK' p m\ȃ+ln lF O{`W$y}`Jm|SJd%N )JIy NGidZ'Հ\N@ Ƈ9{J X rxRErn$`'[ZLX;W:*%Q ml`/L 7{91kl-J0v ^.sv3` 1oh`IԜ([q F&C KC(q<#B*XݦxY֍GE0cV 9qlZ%fFvĽuᙧ3`]_ @( @( @( @( @( @(*؁;` B` B` B` B` B` hv=mXeV}t]Tu[` ޮw xOl~lv[*F|vl*n/X}u 6@|a 9$|dyn;I4[N+r“)V<7'ݍ![-).bncC6@˺^C04:nKT^bҝh$JerR+)YC[aTM0z mq*"04~Tl={V U%B0]QYZY A.Z&аi /E zߦ /%,STJiA7K=p֯:3yQ/(wZ~ɍKIobqQa[\.CGD]0D~a-U;.Jt'[_1U0wL27vL0ƭS0rpbYa❔Ut`CՂ],vʮ%YEeH׬b@7-,Vn.F֒B$7h/H()/W6uщ\U0 *\^3e%ZOfTo}0r.IE@$VlwaK~eS0AnBڭ !G0?s־vp#Js#U&vKfAt*ؐwڛh"\v.k^o\.mH/6 v/(@0#@@0@0 @0 @0 @0 @0 ,Ș'l=` B` B` B` B` B` Lܓ[6A{Io3,2/9,ظ$t'[UY7[)yB?{r\%L}=&*Us9“`gf5xMqi0T.[y"{JrlNMG_aS K""F?e&:7@ݮA`oxWEGB,$2G^(?p-^@e4?m%.eYS,>HSԼSd^NKR׮]G!; Zv"Jw"I]L[Z5R^n"|G\O"4 H!6bo^n!93 drio'ވMfwK M6E H#NAk 7Ff7 \Ι>B` B# B` B` B` B` B` fL_D@( @( @( @( @( @( @( dq,̒ʳNq枬0O,4On5/{/>Z/\tFK`lc)R$'GKCL׼+Hs E-EZ+p.l_̻ 6]K(D"NM-G@>R0rL` KH&z-N3L̦Fg Ňj+27"ўj,N{[<\q{Fߗrn2MVOD|}{t9lLVX =ZG 1`I8EdSߑp4ټWm/R0kPN\2)ڒpO<,Xd.'{r6o|r:#jT#[,̥%慢$/c": _PID_GUIDAN{D2EB26E1-0902-4573-AAC0-E8792A275583}y[iaJ[ɑ>Kb£ب *"*m{Q=N+KAꨶv4`p iAd•EkJ\YQ.},Wt?Nőx.͏jrshsG""*/9__D"'Nכ Gm?շbjLrj kHц16Ms<)v*"u-$tm(0'ͅ!f-%a{gR *qPT3qTà|nΆ%ETf3h".`ch*ǧut=8SVSL9 Tu+UyN,7±DWi~9*O.I6ͯ/JNSྊҖpx<~ϩ5[HUo2Q!lI= զX/w/o5K|֟bUZo&ݿϫqS_tvQy)eo]k ˋڕTY2,H\HD\vr&b1(ƹO*:h]qqc]C1ѨĢ)72lBحʝObR|lw 11CA5QRN~(ޣ]DÓzg>>e`ч<.J˂Q"ƿ%v~A9 k)3L \a6<yD|cwouN>HnMW$I7OK~A7z{ˊO 5ETW jĵEOΝVٟ ϭz:3f?%FH5 U >UϢWC2;:(9TݿϫqX\ڟ:{VoӬh _ d-S[!IG NUAETbK#iEyaY{zjfj O@M@SRf-Aǿ1J }OYmI/zQF,u?77N|kiT;V۟:$ݚFͭvёjG) |>Ot}:@*?#Zފߠ7s8s|Ig\χexќBA?cb&Z=%ʡp(b1{LT].pWHGkŅZx,+F޶6"$-y*/Ư#4Y¥d7Qj NB{/d S}Ful}Kv^h<`7Srp\J4mᗅWeXN[ vKQ "]C'ӁAARU^sHyn>BٖAR7\\n 9rJm 6WmyvQGؖQ2SNYnrۈ 䘚򄂩ʪ{4iq_OȚ[aZZ7iB a+Jfm6<ۺ':U!RDJ %ِҫ=`rt3dl;xE:n}%N:59IGBR)uem +zb<䙍⡨xZQFSeY.j9xUݏ][1wuPffe~2;.>y%q(Ћ`ñǏBT␺fW1!E/mC#U5Yϰ}o:%t%nD.+ % v-񪜋ke|VES1y(.="l6|+\p i h#[\Oy*ɿ$r}xn5WqC-Xo.K(sڜe7630T#H #@ V;]ZO h$g*qUTÃx+Uo Kb-{C~㓆,Sb%|aF{˔몏ek5 -r|o)^f /u"ٴu` ϴkUGlKB=IȲ!7 %=Udp q{U__las偤F UPQk1{Kn9ֹ=;a4gsZXuf$_TMnV2k#KWWmEבH*|j?e빕WNIzth?6 F% 빤`|=P٧bk!@̏;+m˼_1XgG7–6Ir\0ti$<@KȨi*p&UGqmύ{2Q45c! 1PcPmyfnvԦ\b+ ]LL&# M* uUP{~}AoϨM$z:Yv i̕\?@7O"£ViqS'`?j-iwE yGRmD.PB8[58k<yq jNu,RS1ԁskbf@sS3q2]iPn$PD%O={+ .$3kΨ-v]낇VMsw!xUA_,EtQGEtQGEtQGEtQGEtQGEtQGEtQGEtQGEtQGEtQGEtQGEDyK cwert@fundacionginer.orgyK @mailto:cwert@fundacionginer.org$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 <Dd4^  c :Alogo-cidemR6Y.bZ߸`fQ F^6Y.bZ߸`JFIFHHC    $.' ",#(7),01444'9=82<.342C  2!!22222222222222222222222222222222222222222222222222q" }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz w!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz ?'DuĶk5Zm6I,g#vݪ67dY4P_յkm!a apPzWH4"666 p\ 0X׫+e,Cs+4}]5M1YkC2ƌ*īNtּ3Qȱ23\zޓ}Yi_i!{kcsjVUHFn%H D~]S_LE4;ZR$߰ pp2A4Yo/>$Q+qo$2CKF0\\j66Zid3 |PCk MF$Wڦe±2F(yqu%o0 -gLl@'*O- u,MԢH>Y|ͷ 7u u♯[)<'$pR ʟq go^octkkrNϣ I%c&`/,6O5s V.b?@>9⢳~t[]0G1aA%**l׍umFǚq7 hSk3\׺|ٺuEma\Dz2*і |+uv絎5itH] [8^/&6R n C_ˤk>-e@f\`p}++O}ѵ ^'ku2&KXcj;]%}L-].`sİ,*X=r6iEɩhl#|&PV9]%F cɼt tsOּe,e c$dnEV՗j57?JCՎz=cTդX$`X>gLq&I$ <]toi"?3N6X2B$Y$ޟ)>%u;5 WNsyS|Y8g9=D{8eg-e\ͷ 7g'j&][QӬt Ӊ.0>ؠPKrF71ڠg<&s-f3t 6U-͹,\ (\w~5Da!OAvlk K`Abmw#UI~TG<85uIc~ߡڛ8#]9LxzM.˩5w=KO}:OXVPH )u 03tv]/]<7Ҫ6!!rqkU%7^JSmv(9&?aWK!SF5oQ| ]f-ǡyI:&s0$FqFFr:@yDӖAs(c{iU'2S&?!54^! Zs0Fݑ69_'9=EnWw^LbZS}kO#UV]v#ؿ?s5]̱dXU|? 4n.(5hrZC@{Y@Gՠ+qeEtm 0e=?B@i 9Ozo#M_"ipJœc|M {-(K>d~:~5 +ĎKy.Ki 0HdnIdU{mpzӮuc$9q2x5k_ERM1 ZJKJ+>ZW?H%giip;6;-Sڼ~}p[0;b:t>V7>.O u9)l,rf>&O!T^(Nd8-ٱ:g>ƒӾ/DWGYXdR -sj]jTd@9(0 _ ƹt?㎱mOum*,IO,޹6&DoGBMڗMƀ=GZ7)lVAltҵW=T,yD1Sy$n(O_xcGcOa!Zuq~Z?~cB<=s|(ǩ !<R麄gZ6p$%C \?xkV)k(((((( ; JQV?:fcI"4ܜ j@Q@|?4xP]Q%.cv`7S $&^%͝4w ~UG98P&(ҾKE0XXnu=:9|W!bYOlu>(}۟ihl4넹|Ež(nuJ**EQ@sv֚Uq#Yp@=1gP9{hZuyom̴޼1 1k6Ҫ|]7nһ*(VzmaoiAc(TQ@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@Q@~:+`D7lV sWt)^TQEQEQEQEQEQEQE:__u? z&ܒv08`OϸBd>_9xUW_]o eR9C dž]e$n;ڎkQres<-c j::I-=0 1_ }ï^!Ԯ_Rͱw X(*z-Ǵ#<3}Λw Ρw-KFB[әf?ԾzwyMS⦵Y2OˏZ?CHRmx5) >NJ/]hA鰤~;:z.x=K~Yq;3,Eڟ ^_s=Rޟo'ѭk3lE ݰs ]o9sҼѯt-2@͕edIu ygڥ} .]~aKSd?gv.i~ D@ԯgKc&z'TcKi6%վ"8W%6,E+8Rј.[ZVY듘#&T/F t5ZvLظETKWU oZw% U`/RHQܰMV?""|5iڤ{~pKj}4~ =?LS\qI'gCD#@%}~gACb4mϯ*bp/o5 -f& 3pmoBH%OZQH-jqI0Q0^#jjٟ2vzR.oGvకaV0npz.[;#nmkov[qg <߈ck%vځ g]ek FKw+-;ې 7FqӇƞtYu]^?!AGaq8*/_p(&m_te~ qzFu,2J7!!>B:\z&>o&n ̒RyQԲ^8&l. 1k3R-Ǎ-5 g:enMdmZ[{|=zͿ}᛭WGXlF[S|c%}j{oC1k:3CmwA${ǏXS񝡖XP,`>`vzSo<_/Z^ݢ6˟8ϸiѿABdJTvZ^)Xhٵic*|nwS~KSkڤ;aYi#QU9pGsKz攖= Es wzf{c0cfzr=R15$z$Jv˝myq A$7gAk߫4c⽦+ h{mmyBRu>.U^EWYEPEPEPEPEPEP_4xFgэEÎQ/j2 ( ( ( ( (:q}E`|GC? (O ( ( ( (?DyK cgomara@cidem.gencat.esyK >mailto:cgomara@cidem.gencat.es$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 g Dd d  c @Alogo_biovistab stI_-*O k n stI_-*OPNG  IHDR0?͆PLTEֲ̿޾ţսǶ̙sɋǎwX}mm_{yue__UrrqKs{fffgbY??4isZZXSRQ33"doKKK<<<333$$$?A ooo gNqo\Aoocoo sHBl C:\WINDOWS\Desktop\n-logo2.gif<1Dihwr 7zIJ056AY9 T=7V̞YqO~Ry?|//p/n1{  `™rSv~йjSYA= sɱi$(_7̳c (~sOX$Y>`!"DxfYxf0c &uDf8뱇4INѢխ)إY2l7N%/Mf(^p8ϸ:T֨/N)Km^z>ND& m/#I"MdېMM3x0t',$KN'AsMUK[Hk"=.p9 Wo˼QmqhWt87SAHGb*Dy4|{+ppb@oArB4x@xtIyu wr8gp*rE#hJK袝ReYdi45K`/ 1R" {qe P0HgrrࠚVAG17`#b|{Rl.b sT^Ӛ$*z4uֺNP\QB#]/#@lC#Jjkj@ADh5."S 0o".?闹QgI }0sx#X:}Pqv2Q+xu>$B, % "nl&-~ȏZ zxՅ":6<*&@$:aQZO .@aFfcm o$9AAZe* @/%vruN p5Q]1V%i8Eɡ*jCw%[ɫ{Y\Zό`9&^aNDF jwna5W2n @s=V% I@+[}הxP.8IaQՕv;XY#O.,)[c+`tCũгvJ\D \S 2A$a+R[>P6. Z9zK8$Ė.#v2?@xn¶L||o2"I*N<;;p8T8v'@1# $d|yn_ ^LMg`fea5nȦfp;KZ[5fQWPv\Z-.8!ƕTd]3jP-u#fC`:V+ye1 GZr4̢zNF婐;p^. 9/mi:z_^@XJ FsH]% RIA9̫mO]6tkNX7̊uA[+v0%N/]x4XŃk35L"zzeO2:QJn\ƎXeyk5pdCoElWNNw 4/ܵmU=?^ 7mϏSL> Mf&15$lNڼuڦܸaiP.>/"+f8!qZTF}Š?m1Ku `qDoPrw(`? cN[0-uo%}O hrңKzA|mlĊXz޻2mWXct/!'D'ytFhLi d!UfyQbxSh7i>JF@y«^'[oQF1 {+nԅXr=y[nKEY oOQeqRIENDB`DyK biovista@ath.forthnet.gryK @mailto:biovista@ath.forthnet.gr$$If!vh5\ 5$ 5 #v\ #v$ #v :V 05\ 5$ 5 44 {DyK  _Toc33275002{DyK  _Toc33275003{DyK  _Toc33275004{DyK  _Toc33275005{DyK  _Toc33275006{DyK  _Toc33275007{DyK  _Toc33275008{DyK  _Toc33275009{DyK  _Toc33275010{DyK  _Toc33275011{DyK  _Toc33275012{DyK  _Toc33275013{DyK  _Toc33275014{DyK  _Toc33275015{DyK  _Toc33275016{DyK  _Toc33275017{DyK  _Toc33275018{DyK  _Toc33275019{DyK  _Toc33275020{DyK  _Toc33275021{DyK  _Toc33275022{DyK  _Toc33275023{DyK  _Toc33275024{DyK  _Toc33275025{DyK  _Toc33275026{DyK  _Toc33275027{DyK  _Toc33275028{DyK  _Toc33275029{DyK  _Toc33275030{DyK  _Toc33275031{DyK  _Toc33275032{DyK  _Toc33275033{DyK  _Toc33275034{DyK  _Toc33275035{DyK  _Toc33275036{DyK  _Toc33275037{DyK  _Toc33275038{DyK  _Toc33275039{DyK  _Toc33275040{DyK  _Toc33275041Dd` Y aB   S A ?  b[iX 3oi7n/iX 3oiPNG  IHDR`gAMAPLTE@ @ @@@@`@``@@@@ @ @ @ @ @@ @ ` ` @` ` @ @ @ @ @@@@@ @ @@ @ @@@@@@@@@@`@`@@`@`@@@@@@@@@@@@@@@@@@@@@``@`` ` `@ ` `@`@`@@`@`````@``````@````@````@````@``@ @ @@@@@``@``@@@@@ @ @@@@@``@``@@@@ࠀҤ7@ @ @@@@@``@``@@@@֥@ @ @@@@``@`@@@@v pHYsBA:o9IDATxkv(FYA6L73QJۂN#nFnF'  P LąFRZ^ϰWbh*X` TY9.XI \ XedVO d'$,`/tS-O"_B`R2AKt !+@ +`  iM5h`o!Z@G@(4?^5vBg7sw) ukf5`Y@0a`O,%M?\ L-X/a hR> ؝x`Y(G0Tu](3y`%Xbsa#3Ko\y4<!+h,Q^`/,)&Sbc$~u,+<%܇`+Lv,p 0o}qzZN9αMY$*dmNG:&aIzN!Xd4 "udPP'{&gY]0 3`D JC03* X`4`=W f=c :G&@X `@C`s`ȿv;S[?t,iJDc E*Ϗ䉧! M0>%4`q` x NADMkK|Z@_Ol ~UREaFZC`UV A0' NZê7>UV}dokh54 , sr`Q_VR:-f)X#{8 =K1X\/EvZ+n&LCf\,Ya߈!<P#3oS@0_0OwuX_W|ך_k]&m@0푔H:&c7n$ײ6 `e`ydF㵭7Z `[F_L0.g`XT9yI V=hs n``wl 외%z`O062]`}x(W_ <|qɸ  *A k`1@k`1\*؆K}7lU7W3; ` E X(+MC ?"~=_ep"R9Ia{(?<=]|M[?O`V}?)Gp{`ҷ ZqM0YY I,r<ȏ@$XbUXqX c" ]{h-ub5y V` Vw`F+1{`b#.& oSￊ`ʍVzkI _ $߱Z@t%'-o \K0u.X27 `;""X~q`<`AP>J0  p>I0h .߶l;`Y~+f A Fy`m.vh-c1cll_̲,tvJ^T K%hR1Lм%f[n,"C{C/D,Ss`BZTjd,`APTi cw~&` rO`vӋ hM!c"ܦ`7ZT/`̀ 2q(؉s`;`@@v  X;lk``'!@v  X;lY>!I X*)딓 m&2MAu #T`L5*``9HZ^Eҡ=`[iM!`m1i^=`;hf|I/r+wP=a$b4 kGKc?7K;O[OvZkZY5iփ{ 6Dj֑O,4On % eJ @ ,M$̐3v`SWMC0B`,12 nV"!;ʍV#+/dH`X,̆y5_P Aq{ > -;U|G#!RTft+؁P;T_A./C=s`KrhD^D7B+ _(2u /sy%ޫL\nS@xN?= $%cZ- ?]0>ك)q+b1KwJ?dav.QU!y%,)Gy&7 ["R[|`7"b"cG0u&XJm杗 w6E`^薂Vv*R^SEG:Y '[<.!q9˼,% ܻ^\ 5%-^$aPKoV X(  X(8 UTTW>{E!v/.5`uu]jxtyB @ k Vg  TY4|~ N$U @/\ L`,&1Y{hRަψD@+suy'.m `NO:@0@迭A04 @( @( Ɵ=w|k"}`.\ ypmV٭`MP5$&:IDǃ`$R<ߔ/lRD<gZ"a +ANO(x^-27LsK)N.ŠJ F--s셂b9ƶcW>0 ư:VjD͉U`i2T0>3RX.mgl`xLPV`f,1 ߺol7O<':l X}` B` B` B` B` B` B PT@0 @0 @0 @0 @0 \,XOD;k܆MV[a呦\@ŪnM7a._S_:vޖ#b`ZX_xo:\ @0s}i><7;IT[r“d=RTmx*t3ĺ/YR\D}#$dk"~m0`4ht 3UR{YK](ՖH)9vLUK"*'CN;N~ ,`%=2!嶬MH{6vj M ̰ `b^ZGom R2uJxtg<5jrvܹU3 !et"%5 6ѓo)QLQ߯wX cUA$vWsvYEǘ4O*a;&];&wX V)99,հpFUt`CՂC,ʡ%]YEeHӬb@w-,]A]0{&'WI.H()/W6֛mщ~U0T.uD@jEɲ%ZOfTlo}0r.I"]| K+g;0ҥ>m)o 4!OIEJ}Z#jh~Xw`9Sb/#҆Zt 6CW`M{nM~6m.Uw_n6VXinn  X P P P P P P dL@0 @0 @0 @0 @0 @0AQ&- $tql\]-*Ӭ-<=9.> }I@3?Zc`՚{`Kiָ4JE*˭C`%^9E^UQ6#/`0)%2:w]uXoxWEGB,$2Gw-'im%.eX|HSy9]#]d^:ϥPY^Yg/wl v"Jwu$Pޮ &ߖ֚K)^n"]?v)!Yg/wl$.ѩy5>sAP P4P P P P P Q@0 @0 @0 @0 @0 @0 @0 W5/V-Y=}C, 'ldZQD(%Z͋l|`.LZf|y,U?Ռ+27"Q,N{[<\qGFK90&rS0Q)]γ(SU0x`Vтq'_$t$ٯP.Xy&_#Vl}xYTIO\ C=鍾_| XjU )KKRۓ`yc`Bl$/c": =`.E BP =`.E BP =`.E BP =`.E Kc(rxGu,hOK{F8[elX`I䑂tXlF~4[/m' 6{Y۝u}`]z;*rl|뱬!,g2bҍϋ,/꣣ܬ0Zm~5<' 6m4o=у"FXԆb\%\就e ^HvWh&CrGo~CxNz$ȶ69]K425=O*^f򭒦RYbs7B*&(MK [0ZuSPZCJxL4"Nɤ|6k,G77/Ib葱)ls_YzNқ P>F%W99|fN]17X`|~2FUjZmm~+sbg =3tEr32(6Y&^Yu7D ; fGBE-S,%};0ZHjOR>YN(r%C^2Ne"ٟlK[)vrzBY},[i~Sl,Obcҡmlh5ݱ6]}h£Y E zp`?' $@0sA(z\@0sA(z\@0sA(z\@0_5%\IENDB`DyK http://www.w3.org/XML/yK .http://www.w3.org/XML/DyK http://www.w3.org/XML/Core/yK 8http://www.w3.org/XML/Core/DyK http://www.w3.org/XML/LinkingyK <http://www.w3.org/XML/LinkingDyK http://www.w3.org/XML/QueryyK 8http://www.w3.org/XML/QueryDyK http://www.w3.org/XML/SchemayK :http://www.w3.org/XML/SchemaDyK http://www.w3.org/Style/XSL/yK :http://www.w3.org/Style/XSL/DyK http://www.w3.org/RDF/yK .http://www.w3.org/RDF/ADyK 3http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/yK fhttp://www.w3.org/TR/1999/REC-rdf-syntax-19990222/DyK !http://www.w3.org/TR/rdf-schema/yK Bhttp://www.w3.org/TR/rdf-schema/DyK http://www.w3.org/TR/rdf-mt/yK :http://www.w3.org/TR/rdf-mt/DyK )http://www.w3.org/TR/rdf-syntax-grammar/yK Rhttp://www.w3.org/TR/rdf-syntax-grammar/DyK $http://www.w3.org/TR/rdf-testcases/yK Hhttp://www.w3.org/TR/rdf-testcases/DyK !http://www.w3.org/TR/rdf-primer/yK Bhttp://www.w3.org/TR/rdf-primer/DyK #http://www.w3.org/TR/rdf-concepts/yK Fhttp://www.w3.org/TR/rdf-concepts/ Dd f   c BA picture-layersbS^}>1&./] n'^}>1&.PNG  IHDR !PLTEM.M>M>.h>)MM.hS)hS>|J1hh>|c1|cJ||JT8\=p8{=pT{\dBkGqKwO}ST\BGKdkqdkqǟOقVǟwЦS٭VЦ}Z]`fZ]w}`f٭ᴇ麋ق̙([~bKGDHIDATx [8@9M)m(GB.icIV9ٍe]f$(O@|PBIP%A!PBIP%A!5SP)o/,=5/@gyqP`p/ i .!GpzB(ჱr+l-P}IT^Uer>J`,(<:$$t'dM_zB(4f]%6Y>DX BL%;bߒ.4Ƅ,*񏼨1ӁΏD,YnJ 4Ē T?ՒJ%`L(ĮClu /- mGY|pK/V@Oޓdf ы߃X%7 rkB֭اE}Rg,hޡ,<KL e'JB( $A Vň۾">[Pz*ɦjܢ#-:~Σ ȋAyiTW< Zj o%0;QvQŒ@%LWe`wg2i gP7T%JE(F=+C}Zm_%픟RN`&(74C1.5)k z,*xbiM}Ai>BY}Aq;>]r_f9w_Q?ﳇPi^X;)Eu \$(JB( $(J@i P2x!PBIP%A!PBIP%A!PL@1]5d`x̸EҬCYl y>x?8&8eO$5I@yg5, , !O{O4X"2 9{$q:=<1u Cpʉ85eK.fyD 2c 8'j/2W!i)V:ux FFx C!V@2PEUlfl{snPYn,z9%9S]O3~H#A>[do, eCŔg)Mxb 'gOc؝SU8E&uC>ɘHR)ܧ4eo&"b 13_}zPM ,c?Du)cF}FSh9$+LBɞh`ߣimNNl͝r;^3SṢJB( $(JB( $B-oe<**V庼X~l«4_Pԥ}!7i{qC\\- AS..BV5X5szҧQ7dEe -WϫU>%Ju>YeOyڜC}X 9^3EESq_6-viUAV>EPBɄ4#46Т*}JV4%c+C(4PRmu $(Hy* +J ƋQPB(q4(f3XP0%Lm%oW) e+pÆSJQbzRTs( f>JyJ~̂PB( $(JB( $Pml>,}/,Jxe]Ũ=0#&bY}ѿSb}Gea}JQ-CуT-}d9K/-bt#sgqzBT:E[ bs PjHj@Ǡ覲Lo9A"/~PD F)B))Bm@[2R;Qܒ%̟6V&h⍑:Z6>%5~G(jJc*HAp O}KC _9&W3*Wn^  )Gk7P|jTܡSA(r6u8@_wmJ WwmB 蚿Jl4KP\\]RA(H_y9, _Y9 B񕧑C _ oSA( =/⫐Ք'⫠⫰=:B|!_BhP|zA 0(>l<TQN;BSJԩ(W%A!PBIP%A!u+JTϔ6YC$1P@8-iUFX-\㠀KnΔ*(܂Pb HeB0SދP*ܗT)3U:(KN^}!1 N%A!PBIP%A!PT^/IBSQ2UJB( ((?MŪF17tEXtOFIENDB`DyK http://www.w3.org/2002/07/owlyK <http://www.w3.org/2002/07/owl$$If!vh5I 5#vI #v:V l    t0  5I 5/ 4pT$$If!vh5I 5#vI #v:V l  t 0  5I 5/ 4p T$$If!vh5I 5#vI #v:V l  t 0  5I 54p T$$If!vh5I 5#vI #v:V l  t 0  5I 54p T$$If!vh5I 5#vI #v:V l  t 0  5I 54p TDyK "http://www.ontoknowledge.org/oil/yK Dhttp://www.ontoknowledge.org/oil/5DyK 0http://www.daml.org/2001/03/daml+oil-index.htmlyK `http://www.daml.org/2001/03/daml+oil-index.htmlDyK (http://www.w3.org/TR/daml+oil-referenceyK Phttp://www.w3.org/TR/daml+oil-referenceDyK "http://www.w3.org/2001/sw/WebOnt/yK Dhttp://www.w3.org/2001/sw/WebOnt/EDyK 4http://www.w3.org/TR/2002/WD-owl-features-20020729/yK hhttp://www.w3.org/TR/2002/WD-owl-features-20020729/9DyK 1http://www.w3.org/TR/2002/WD-owl-guide-20021104/yK bhttp://www.w3.org/TR/2002/WD-owl-guide-20021104/1DyK /http://www.w3.org/TR/2002/WD-owl-ref-20021112/yK ^http://www.w3.org/TR/2002/WD-owl-ref-20021112/IDyK 5http://www.w3.org/TR/2002/WD-owl-semantics-20021108/yK jhttp://www.w3.org/TR/2002/WD-owl-semantics-20021108/5DyK 0http://www.w3.org/TR/2002/WD-owl-test-20021024/yK `http://www.w3.org/TR/2002/WD-owl-test-20021024/ DyK &http://www.mindswap.org/2002/owl.htmlyK Lhttp://www.mindswap.org/2002/owl.htmlDyK http://www.daml.org/language/yK <http://www.daml.org/language/$$If!vh5M5N5N5N#vM#vN:V l      t(0  5M5N/ 4p(T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N/ 4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$If!vh5M5N5N5N#vM#vN:V l  t 0  5M5N4p T$$IfP!vh5O5555555#vO#v:V l          tP0  5O5/ 4aPpPykd $$Ifl  ִ3 }"!l(/5O          tP0      4 laPpP$$IfP!vh5O5555555#vO#v:V l  t 0  5O5/ 4aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5O5555555#vO#v:V l  t 0  5O54aPp $$IfP!vh5555_5555#v#v_#v#v#v#v:V l          tP0  55_5555/ 4aPpPykd$$Ifl  ִ  l(-0D6_          tP0      4 laPpP0$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_5555/ 4aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp "$$IfP!vh5555_5555#v#v_#v#v#v#v:V l  t 0  55_55554aPp $$If!vh55 5#v#v #v:V l     t0  55 5/ 4p$$If!vh55 5#v#v #v:V l  t 0  55 5/ 4p $$If!vh55 5#v#v #v:V l  t 0  55 54p $$If!vh55 5#v#v #v:V l  t 0  55 54p $$If!vh55 5#v#v #v:V l  t 0  55 54p $$If!vh55 5#v#v #v:V l  t 0  55 54p $$If!vh55 5#v#v #v:V l  t 0  55 54p $$If!vh55 5#v#v #v:V l  t 0  55 54p DyK yK Fhttp://www.w3.org/TR/owl-features/=DyK 2http://www.w3.org/TR/2002/WD-owl-absyn-20020729/)yK dhttp://www.w3.org/TR/2002/WD-owl-absyn-20020729/)5DyK 0http://www.w3.org/TR/2002/WD-owl-ref-20020729/)yK `http://www.w3.org/TR/2002/WD-owl-ref-20020729/)V<@< NormalCJ_HmH sH tH8@8 Heading 1$@&5@@@ Heading 2$@& CJ$OJQJD@D Heading 3$<@&OJQJP@P Heading 4$$@&a$56CJmH sH u@@@ Heading 5$$@&a$5hD@D Heading 6$@&56htH uH@H Heading 7$@&5>*CJhtH uD@D Heading 8$@&5>*htH uL @L Heading 9 $$@&a$5CJmH sH uDA@D Default Paragraph FontVi@V  Table Normal :V 44 la (k@(No List ROR 0berschrift 1$ uhmH sH tH u\O\ Autor&$ 1 1$^` a$hmH sH tH uBOB CellBody1$hmH sH tH u O! Bold0O10 HochgestelltPOP H1$1$dd@&5CJ0KH$hmH sH tH uLOL H2$1$dd@&5CJ$hmH sH tH u0U@a0 Hyperlink>*B*N @rN Footer$ !1$a$hmH sH tH ubP@b Body Text 2 $ 7dx1$a$hmH sH tH u^^  Comment Text$ 71$a$CJhmH sH tH u4@4 Header  9r DOD List1$ & Fa$CJmHsHuBB@B Body Text$a$ B*htH uNL@N Date$ 7<<1$a$hmH sH uJOJ Normal (Web)1 dd mH sH uHOH H3$1$dd@&5CJhmH sH uX"X Caption $ 7xx1$a$5hmH sH u@&@@ Footnote ReferenceH*:@":  Footnote Text"CJ@Q@2@ Body Text 3#CJhtH u: : Index 1$^`: : Index 2%^`: : Index 3&^`: : Index 4'^`:: Index 5(^`:: Index 6)^`:: Index 7*^`:: Index 8+^`:: Index 9,p^p`6!@B6  Index Heading-8@8 TOC 1 .xx 5;CJ6@6 TOC 2 /^:CJ6@6 TOC 3 0^6CJ22 TOC 4 1^CJ22 TOC 5 2^CJ22 TOC 6 3^CJ22 TOC 7 4^CJ22 TOC 8 5^CJ22 TOC 9 6^CJZC@rZ Body Text Indent7$^a$CJmH sH u,X@, Emphasis6(W@( Strong5ZOZ ECAI A level:$ & Fu1$hx 5CJPOP ECAI text;$1$`a$CJOJQJh:O: FirstPara <1$`JOJ ECAI B= & Fu1$hx 5CJ^O^ ECAI C level!> & Fu hh^h` 56CJNON author?$`a$OJQJmH sH uPOP address@$`a$CJOJQJmH sH uLO"L emailA$`a$CJOJQJmH sH uHO1BH abstractB77Xx]7^7CJ<O< p1aC$a$OJQJmH sH ubO2b heading1!D$$$ @*$a$5OJQJmH sH u^O2^ heading3E$$$ @@*$a$5OJQJmH sH uOqb Bullet ItemYF>TO ItemG$ & F >T-^`a$OJQJmH sH ubO2b heading2!H$$$ @*$a$5OJQJmH sH u>O> Bibliografa6mH sH uPO2P heading4 J$@a$56OJQJmH sH u^O^ table titleK$$$xa$CJOJQJmHsHuFOF ContenidoTablaL5mH sH udOd referenceitemM$^`a$CJOJQJmH sH u.)@. Page NumberV>@V TitleO$<@&a$5CJ KHOJQJmH sH uROR tableP$<<a$B*CJOJQJhmH sH uFV@F FollowedHyperlink >*B* ph\O\ Esperonto Title LeftRCJOJ QJ mH sH ubO!b *Esperonto Title MiddleS$a$5CJ$OJQJaJ$PO1BP Esperonto Subtile MiddleTB*Ng@QN ijWHTML TypewriterCJOJ PJQJ ^J aJs))?rR)Z^Wm qЃa̎2~&zz ,Zr 7Wx<Px>hz  >h&z=zC%&')*-.:;<@ABDEwyz{|} @AFMNSTUbiklmn;<N/ &'0@JKVest)Lm 3 U ) A R l  ! 6 V  ! A Y v   + 6 : I n x 3:mo}DENWYn'f <j4k0_Q Q,XEv 68FGHI23mn[!\!##########%%%%%%T'U'V'y'z'{')))))$+%+,,.......11111222334466777777999999+;,;>>>>>>@@@@@;Bd1qr1v9:#-.369;<GJMPQUW[]^egkmnrux{| !"/01Z[opOP^_`\     = > + , &'89:XVcz{|8 !"#?%&&&&&&&&&&&' ''''''"'#'$'''(')','-'M'N'O'R'''''''''(( (3(4(((((((((((((((((((((((((((((((((((((())S)T)|)})))))))))))))* * * *G*H*I*J*u* + +++T+U+V+u+v+++E,F,G,H,q,r,,,,,,, -3-4-i------..'.(.).5.6.:.;.?.@.D.r.s.w.x.|.}................./ / /7/8/h/i//////////////0 0 0^0_000011@1A1s1t1u1v11111112+2O2s2222333333 3 3 3 3 333=3>3G3N3X3j3u33333333333333344444"4#4$4/4V444444444595:5;5B5C5G55555555555666666(656667686R6S6T6W6X6\6|6}6666666666666666666'7+747=7J7b7c7g77777777888&838@8I8u8x8|88888888888 999D9I9R9Y9b9f99999999999::!:*:3:4:9:Y:s:t::: ; ;;;\;h;z;{;|;;;;;;<<#<5<`<l<<<<<<<<<<====I=f=x=y=z={====>?>AUCCDkFlFFFFFFF?HcHHH[II J'J2J=JSJJJJKaKbKLLLLMMMQQSRTRUR~RRRESSiTUUJVVWNX+YdZeZZZ[[3]4]vcwcddddddddyezeeerfsf'g(ggg-h.hUiVijjhkikSlTlll/m0mmmLnMnnn7o8oppqqqqrrospssttttt+t,t-t.ttuHu`uxuuuu%vEvfv|vvwwxxx*y>yfyyyyyyzz%z&z'z+z,z?zdzeztzzzzzzz000S0T0000000T000S00000S0D0D0D0DS0DS0D0DS0DS0DS0DS0D0D-0DS0DS0DS0D0D0D0D0D0D0D0DR0D0D0D0D0DS0D0DP0DP0DP0DP0D0DP0DP0D0DP0DP0D0DP0DP0D0DP0DP0D0D0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0D0DP0DP0DP0D0DP0DP0D0DP0DP0D0DP0DP0D0DP0DP0DP0D0D P0DP0DP0DP0DP0DP0DP0D0DP0DP0D0DP0DP0D0DP0DP0DP0D0DP0DP0DP0D0DP0DP0DP0D0DP0DP0DP0D0DP0DP0DP0D0D0DR0D0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DP0DP0DP0DP0DP0DP0DP0DP0DP0D0DR0D.0Dp.0Dp/0Dp00Dp00Dp00Dp/0D00D00Dp00D/0Dp00Dp00D/0D00D00Dp00Dp/0Dp/0Dp/0D/0D.0D/0D00D00D00D/0D00D00D00D00D00D00D00D 00D00D0.0D .0D .0D.0D 0Df 00808080808080808080808080808080808f 00#0#0#0#0#0#0#0#0#0#0#0#0#0V'0V'0V'0V'(0V'0)0)0)0)0)0)0)0)0)0)(0V'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.(0V'09090909090900>0>0>0>0>(0>0@0@0@0@(0>00D0D0D0D0D80D0vE0vE0vE0vE0vE0vE0vE80D0(I0(I0(I0(I0(I0(I0(I80D0J0J0J0J0J0J0J0J0J0J80D0DL0DL0DL0DL0DL0DL0DL0>0P0P0P0P(0P0Q0Q0Q8 0Q8 0Q8 0Q0Q0Q0Q0Q0Q0Q0Q0Q0Q(0P(0P0q]0q]0q]0q]0q]0q]0q]0q]0q]80q]0b0b0b0b0b0b0b0b0b0b0b0b80q]0Ge0Ge0Ge0Ge0Ge0Ge0Ge0Ge0Ge0>p0k0k0k0k0k(0k0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m0 m> 0 m0 m; 0 m0 m< 0 m0 m0 m= 0 m= 0 m= 0 m0 m0 m0 m0 m80 m0/z0/z0/z0/z0/z0/z0/z0/z? 0/z? 0/z? 0/z@ 0/z0/z0(0ނ0߂0߂0߂0߂0߂0߂0߂80߂0J0J0J0J0J80߂0ڊ0ڊ0ڊ0ڊ0ڊ0ڊ0ڊ80߂0W0W0W0W0W0W(0ނ00000800000000& 0' 0& 00000800*0*0*0*0*0*0*800y0y0y0y0y0y0y80000000000000H000d0d3 0d3 0d3 0d3 0d3 0d3 0d3 0d0d0d800000F 00B 0G 0G 0G 0G 0G 000H 00000000000000000000I 00B 0008000000000000000080000000ނ0$0$0$0$0$0ނ00q0qJ 0qJ 0qJ 0q0q0q0q0q0q0q0q0q0q000i 0j 0k 0m 000n 0n 0n 000000q00000000000000f 0000000000000(0000T 0T 0T 0U 0T 0T 0(0(0000V 0V 0V 0V 0V 00(00r0r0rX 0rW 0rY 0rW 0rW 0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r0r00000(00000(00000(00000(00"0"0"0" 0"0" 0"0" 0"0" 0"0" 0"0"Z 0"0" 0"0"0"0"(00P0P0P0P(00000[ 0[ 0000(00000(00 0 0 0 0 0 0 0 0 0 (00'0'0'0'0'0'0'0'0'0'0'0'0'f 000000000000P 0P 0P 0P 0P 0P 0P 0P 00-00o 0o 0o 0o 0o 0e 0o 0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000#0000000000000000000000000000000000000000000000000000000000000000000000"0"00000"000000000000000000000"0000000     ' !"#$%&()*+,-./0123456789:;<=>?@ABCDEGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~00000000"0000000000000000000000000"000000000000000000000000000000000000000000000000000000000000000000000000000000"00000000000000_ 0_ 0 000000000000000000000000000p 0p 0 0p 0p 0 000000000 0 0 0 0 0 0 0 0 0 0  0  0  0  0  00000wMf 00yMp0yMp0yMp0yMp0yMp0yMp0yM0yM0yM0yM0yM0yMpa 0yMpa 0yMa 0yMpa 0yMpa 0yMa 0yMpc 0yMpc 0yMPc 0yMpc 0yMp0yMp0yMp0yMp0yM0yM0yM0yM0yM0yMp0yMp0yMp0yMp0yMp0p01ep01ep-01ep01ep01ep01ep01e01ep01e01ep01ep01ep01ep01e01ep01ep01e01ep01ep01ep01ep01ep01ep01ep01ep01ep01ep01ep01eh01ep01ep01ep01ep01ep01ep01e01ep01ep01e01e01ep01e01ep01ep01ep01ep01ep01ep01ep01ep01e00tp0tp0tp0tx@"0@0@"0p@"0@"0p@"0@"0@"0@"0p@"0@"0@"0p@0p@"0@0p@"0@"0@"0@"0@"0@"0@"0@"00,@0@0@0@0@00X+ 0H00p0p00p0p00p0p00;<JKstDEzOy00Oy00 Oy00@0Oy00@0Oy00@0K K Oy0 0\UKLK<K K K K $K Oy0 0G|KLK<K K K K $K Oy00A@0x M M M Oy00̃U@0x M M M Oy00AOLM<M 1Oy00U@0R Oy00U@0RO900>>]]]]]]]]]]]]]]]` lKyK x6 s1;GRXgxls yߊ-E$x1|/D4>;*?#D+asy|C~π˂   !%,146:AOYdkmoqrsu | @ T m  ; &JsD4+6AKRYjsvט(P7Yy1-;P]m{[O8&///00011F444'6(6667999;;;"<#<9=:=>>>>@@@@AA4B CzCDDDxEG2RZlt|f˂   "#$&'()*+-./0235789;<=>?@BCDEFGHIJKLMNPQRSTUVWXZ[\]^_`abcefghijlnptʂpX } o Y   '6Gbd 7:Jeh/2Kfiu+.?Z]i1LO_z} 1LOe '*8SVe%@CVqt 146%%%JJJJKKKIKgKiKKKKKKK#L@LbbbbcKcMc{ccccccdFdHdydddddd eCe)oĺ%UW J~-\^ս׽DFy̾tttzXXXXXX:XXXX 4%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%:XXXXXXXXXXXXXXXXXXXXXXXXXXXXNUX`!nsXXt  ,b$7Kn#.! `@3f~  (  Xf   @& S"`2  C 3f `  C ?  @ `   C  ? `   C  ?p @& fB   s *D)fB  s *D)@dfB B s *D)0`dbB  c $DfjJ"bB  c $DfjJ"TB  c $D)TB  c $D) B S  ?(  VB  C D"*+BwnzW;+tTt"Wtm#tm#t$8t`+K!+u9 _Toc26033749 _Toc26034205 _Toc26087448 _Hlt22969509 _Hlt22969510 _Toc25907589 _Toc33275002 _Toc33275003 _Toc33275004 _Toc33275005 _Toc33275006 _Toc33275007 _Toc33275008 _Toc33275009 _Toc33275010 _Toc33275011 _Toc33275012 _Toc33275013 _Toc33275014 _Toc33275015 _Toc33275016 _Toc33275017 _Toc33275018 _Toc33275019 _Toc33275020 _Toc33275021 _Toc33275022 _Toc33275023 _Toc33275024 _Toc33275025 _Toc33275026 _Toc33275027 _Toc33275028 _Toc33275029 _Toc33275030 _Toc33275031 _Toc33275032 OLE_LINK1 _Toc33275033 _Toc33275034 _Toc33275035 _Toc33275036 _Toc33275037 _Toc33275038 _Toc26033785 _Toc26034241 _Toc26087484 _Toc26033786 _Toc26034242 _Toc26087485 _Toc33275039 _Toc26033788 _Toc26034244 _Toc26087487 _Toc3327504040811 _Toc33275041EEE 9#V').9>@@SBDPQ]km.pG.Oh+'0   < H T `lt| ss Gebruiker ebrebr Normal.dotnext web generation771Microsoft Word 10.0@*{@@qP@re3A  FMicrosoft Word Document MSWoDocumentSummaryInformation8(CompObjj?] 7FFFLLLL}R}R}Rd8o*tz=nxoL#pth#q<#rs#s"t<>u$#vԉ#w<x4!y"zt!{$#|d#}l"~"##M#DM##\##ܜ#,##<}"|"L|}" f"d\o"o"OPDP33<4|4qDqqq,A#lA#A#,B#>#>#?#D?#?#3T33'Kt A A O d d m nn++,w-//!/,/,/HHdee)i)i3i;i@i~i=nWn]q]qgqnqsssz      !"#$%&'()*+,-0./123457689:;< /U~ K X X l u u /+,{-/*/*/3/3/OOdee1i9i>i>iSiiInYneqlqlqqsssz    !"#$%&'()*+,-/0.123457689:;< 9 *urn:schemas-microsoft-com:office:smarttagsState>(*urn:schemas-microsoft-com:office:smarttags PostalCode:"*urn:schemas-microsoft-com:office:smarttagsStreet;#*urn:schemas-microsoft-com:office:smarttagsaddress8:*urn:schemas-microsoft-com:office:smarttagsdate8)*urn:schemas-microsoft-com:office:smarttagsCity=;*urn:schemas-microsoft-com:office:smarttags PlaceName=<*urn:schemas-microsoft-com:office:smarttags PlaceTypeB'*urn:schemas-microsoft-com:office:smarttagscountry-region9=*urn:schemas-microsoft-com:office:smarttagsplace 11012171998199922020022003222533078DayMonthYear=<;:  : :  : ::  : =<;=<;=;<=)('=)'#"====<;='='='=='==)=) ': =:==)':=):  Zdgl  ! - / 1 3 U V W ~ ) l m n V W X          * + 5 : < D H  &'2;ABMNOP_`kv|} Demq{E!I!["d"##"'*'****++ ++++!.$.!/*/0055՜.+,D՜.+,H hp  Vrije Universiteit, AmsterdamR0t{  Title 8@ _PID_HLINKSAzf#http://www.w3.org/TR/owl-features/EShttp://www.daml.org/language/RM&http://www.mindswap.org/2002/owl.html";0http://www.w3.org/TR/2002/WD-owl-test-20021024/Q5http://www.w3.org/TR/2002/WD-owl-semantics-20021108/{./http://www.w3.org/TR/2002/WD-owl-ref-20021112/ P1http://www.w3.org/TR/2002/WD-owl-guide-20021104/"!4http://www.w3.org/TR/2002/WD-owl-features-20020729/J "http://www.w3.org/2001/sw/WebOnt/6=(http://www.w3.org/TR/daml+oil-reference|10http://www.daml.org/2001/03/daml+oil-index.htmlPR"http://www.ontoknowledge.org/oil/Ehttp://www.w3.org/2002/07/owlzs1http://www.w3.org/TR/2002/WD-owl-guide-20021104/$DescriptionLogics#DescriptionLogicsvz#http://www.w3.org/TR/rdf-concepts/!http://www.w3.org/TR/rdf-primer/l-$http://www.w3.org/TR/rdf-testcases/I)http://www.w3.org/TR/rdf-syntax-grammar/ http://www.w3.org/TR/rdf-mt/!http://www.w3.org/TR/rdf-schema/d+3http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/dhttp://www.w3.org/RDF/Dhttp://www.w3.org/Style/XSL/ http://www.w3.org/XML/Schemakjhttp://www.w3.org/XML/Queryhttp://www.w3.org/XML/Linkingnghttp://www.w3.org/XML/Core/mhttp://www.w3.org/XML/y mailto:biovista@ath.forthnet.grG1mailto:cgomara@cidem.gencat.es. mailto:cwert@fundacionginer.org /mailto:valli@csc.liv.ac.ukNw mailto:uszkoreit@coli.uni-sb.deEi  mailto:dieter.fensel@uibk.ac.atA" mailto:asun@fi.upm.es(mailto:richard@isoco.com mailto:dybkjr@pdc.dkWhttp://www.esperonto.net/v,0http://www.w3.org/TR/2002/WD-owl-ref-20020729/)D2http://www.w3.org/TR/2002/WD-owl-absyn-20020729/)7777NBRBBB$H(HHHHH>JDJFJKJ``q qorsr4= $&35CEQSce|~ 1LV[f JNʔΔ5?@Afj.2#D%&VY̫Ϋ%&L&35=٭',;dkmyݯ ȴʹߴ4?DNߵ!08EFXZfgxzŸƸܸ ι߹ (.`l  COINhm8<7>*.Y]0BY 4?NV  kq@ET\ "'emr!x!%%&&u*{*F4U4445#57799CCLDYDcHHbKfKUR]RRRSST#TYYZZddddeesf|fffffffff(g.ggggg:h@h_hfhnhshyhhi#ifiniti|iii!j&j0j8jPjWjijpjvjjjjjjjjjjikk5l?lcljlll?mFmmmMnUnnnnnnnTo\oeomopp]qeqqqqrr rrrrrrrr?sAsnsssty&z'zz3 T U ) @ A Q R k l   ! 5 6 U V  C3&&{''))))..C0K0L1W1337 7@@cAzABBCCCCADDDFJKJPPBTFTkknnv vI$&@B|_g}xzY_۞ܞ||;M5='dkݯѱرȴ8? ҹ߹u{#Ze&1ptFK  {}R[&&''''3399@A~AACCDDcHHIISJYJbKKMMYPcPQRVV [[\\aaddddzeeeeeee?fsffg&g>ggggghhTiViiikkll'l(lRlTlvlwlllllll mm.mmmmnn*nfnsnnnnnn6oppppqqqqq)rErrrrpszstttyyy&z'zz3333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333333'}A/U l V  6 EW 9I##99==UUq]]GeWekk}pp߂y08hz*$1Y} 9m 1r1 P`&''"'$''')','O'R'( ((((((((()*, ---.(.).5.6.:.;.?.@.D.s.w.x.|.}..13333334444#4$4/44445;5G5555566686W6X6\6}666666'7+7J7g7778&8u8|88888b9f99999: :!:X:Y:: ;;|;;<#<<<<=f=x=y==L>>?HcHHHISJJKbKKLMaUUtuuyyzz"z$z&z'z*z,z?zeztzzzzbKfKty&z'zz Gebruiker Gebruiker Gebruiker Gebruiker Gebruiker Gebruiker Gebruiker UNI Innsbruck Sinuhe Arroyonext web generation~'1   ?B'j^c  =LF(0Ejw  1G N d:jq S  [ {F~  %G Hw|"Q  o;  lqP jMi ~ Y_j18   BE 1f:`'  h F J|  Jwppa kp$ΙH} p E!j_(@" v" s" F"B'jsk# tY$ ;$F V%j%j%  %& zn&j:3& Gz@'j1%K'Zʌ$j( a*m(=s(tD)4&X++ , { - o.  . ;/ kc/Y.20U2hd0H#M3 M 4[5nU4 R5 \&>6 6d7` 8 !8 9.z{9 qq9 %:.(;P-Eo; Ej wLl +l 0mm >n`yDn o j&X++^](FZ$j(o; Qn_$e VM+l:3&Dn;/M[vXy'I5nU49H_^cp 1GwtZc?Jo{~H}h Fdg K&pSLTrEU];bC,bR [|zg3}-1iKXV88g =L<;$\&>6FKnOl.D=T!8Q [iwL(F8ve{ -'"] .YK8[M1G x__kp'"@a%G >nOIhd0kc/D) Dzn& JGz@'2SJJ-[|% x1'k1i4M>~>{@3G3N3X3j3u333333333344"4#4/4V44444595:5B5G55555566676R6W6|6666666'7b777788&8I8u8x88888889D9b999999::!:*:3:4:9:Y:s:t::: ; ;;\;z;{;;;<<#<<<<<<<<=f=x=y=FLL}Rd*tttHu'z+z?ztzzz=mJmJmJmJmJ?3: @bKbKH/bKbK@{zp@Unknown Gz Times New Roman5Symbol3& z Arial;WingdingsKTimesNewRomanPSMTM Arial-BoldItalicMTO1 CourierCourier NewWTms RmnTimes New Roman3Times?& Arial Black?5 z Courier New;& Helvetica"qhrkx&Mqg3A3A!24d0t0t 3qH(? Gebruikernext web generation                           ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~  rdDocWord.Document.89q