Working with Ontologies Introduction to DOGMA and related research

Outline Ontology DOGMA Semantic Web Issues

Ontology Definition “Classical” definition: “Specification of a conceptualization” Keyword: Agreement Semantic consistency Unambiguous communication

Ontology Paradigms Logic A priori specification Formal logic Necessarily Small-scale Modeling Focus on application Formal basis Potentially large-scale

Ontology Paradigms Extensional vs. Intensional Intensional Strongly based on axioms and rules Hard agreement Extensional Large collections of facts Scalablility

Ontology and IS Semantics Conceptual Schema agreement ONTOLOGY designer domain expert user Any Design Tool Implementation Information System (including the WWW) interpretation Data “World”

Ontology Grail “specification of interface, communication and documentation for any module in any software system is mapable to a common ontology” [Meersman 2000]

Outline Ontology DOGMA Semantic Web Issues

DOGMA Purpose STARLab Ontology experimentation platform Flexible, modular architecture Lexon-based metamodel Ontology Server generator

DOGMA Architecture

DOGMA Metamodel Lexons: elements of form  where  is a context; t 0, t are terms and r is a role

DOGMA metamodel Example: (#my_company) employee is_a (#living_being) person is_a contract_party WITH first_name WITH last_name WITH empl-id has_birth date has_start date has salary works_in department

DOGMA metamodel

DOGMA Syntax XML-based representation of the model. Bulk conversion of ontologies: Conversion of existing ontology to DOGMA syntax Bulk insertion in a separate context (Semi-)Manual alignment

DOGMA API Programmatic access to the ontology for clients Java 2 API Direct support of the metamodel Basic operations support

DOGMA Content Incorporation of well-known thesaurus WordNet Project-specific content] EuroWordnet base types IPTC Category System ….

DOGMA Applications Generation of application-specific “views” on the global ontology Delivery of support applications (Tailored) Browsers/Editors DOGMA Projects: Hypermuseum NAMIC

DOGMA Applications: HM Hypermuseum project Purpose: To create a tool for the creation of websites to browse of museum information Ontology-supported navigation and searching of appropriate museum data Ontology sources: Models from museums Data from museums WordNet

DOGMA Applications: NAMIC News processing project Purpose: Support of journalists in news agencies Project-wide ontology-based semantics Ontology service User profiling

DOGMA Applications: NAMIC

Merged ontological resources News categories (IPTC) Lexical resources EuroWordNet Named Entities User profiling Determine the user’s information needs Provide a consistent view of the system for developers and users

Outline Ontology DOGMA Semantic Web Issues

Semantic Web Introduction “The Web was designed as an information space, with the goal that it should be useful not only for human-human communication, but also that machines would be able to participate and help. One of the major obstacles to this has been the fact that most information on the Web is designed for human consumption […] the Semantic Web approach instead develops languages for expressing information in a machine processable form.”

Semantic Web Syntactic level XML: General syntactic infrastructure Arbitrary document types defined by DTD (or XML Schema) Related standards Namespaces Linking ….

Semantic Web Vocabulary level RDF(S) Topic Maps

Semantic Web Vocabulary level

Semantic Web Vocabulary level RDF Schema Classes and properties Constrains Extensibility

Semantic Web Vocabulary level

Semantic Web Logical level Very much in progress Some prototype languages and systems Fundamental scalability problems

Semantic Web and DOGMA Similar assertion-based metamodels Possibility of using DOGMA as a repository for Ontologies in the Semantic Web

Outline Ontology DOGMA Semantic Web Issues

Future work Alignment Visualization Mining Semantic Web Convergence

Alignment concepts Merging: To create a single coherent ontology that includes all the information form all sources Alignment: To make the all sources consistent and coherent with one another but keep them separate

Alignment algorithms PROMPT: Semiautomatic, semantic-based algorithm Simple frame-based knowledge model: Classes Slots Facets Instances

Alignment algorithms: PROMPT Make initial suggestions Select next operation Perform automatic updates Find conflicts Make suggestions

Alignment algorithms: PROMPT

Mining Content availability is a major issue Sources: Conceptual schemas Database schemas XML DTD’s and schemas Semantic web ….

Issues and DOGMA Aligment: Direct support (and better algorithms) needed Mining: DOGMA model allows quick incorporation of new ontology data Visualization: Potential large-scale ontologies may require new techniques

Projects available!