Detta är ett uppsatsförslag hämtat från Nationella Exjobb-poolen. Klicka här för att komma tillbaka till samtliga exjobbsförslag.
Ontology Learning from Swedish Texts
Ontology Learning is the process of trying to automatically derive a formal model from indications of classes, properties and formal axioms that can be found in natural language texts or other sources. The aim is to reduce the initial manual effort when constructing an ontology (formal model of a domain), e.g. for the Semantic Web.
During the past decade several such systems have emerged, as part of ontology engineering environments, e.g. Text2Onto from the University of Karlsruhe. However, since most of the techniques that are used are language dependent, few OL system for creating OWL/RDF ontologies so far has the capability to treat texts in Swedish. The aim of this master thesis is to analyze existing OL techniques, select the most suitable ones, transfer them to apply to Swedish (e.g. develop new linguistic patterns, translate existing rules etc.), and implement a prototype system that can treat both English and Swedish texts as a plugin to an ontology engineering environment.
Recommended skills to attempt this thesis work include: good knowledge of Java programming, some experience with language technologies; ideally previous experience of frameworks such as GATE, but this is not mandatory. Basic knowledge of Semantic Web standards such as RDF and OWL is an advantage, however, knowledge of logical languages and XML is also sufficient.
Informationen om uppsatsförslag är hämtad från Nationella Exjobb-poolen.