Ontology Learning from Open Linked Data and Web Snippets

Ilaria Tiddi (1) , Nesrine Ben Mustapha (1) , Yves Vanrompay (1) , Marie-Aude Aufaure (1)

1 MAS - Mathématiques Appliquées aux Systèmes - EA 4037

The Web of Open Linked Data (OLD) is a recommended best practice for exposing, sharing, and connecting pieces of data, information, and knowledge on the Semantic Web using URIs and RDF. Such data can be used as a training source for ontology learning from web textual contents in order to bridge the gap between structured data and the Web. In this paper, we propose a new method of ontology learning that consists in learning linguistic patterns related to OLD entities attributes from web snippets. Our insight is to use the Linked Data as a skeleton for ontology construction and for pattern learning from texts. The contribution resides on learning patterns for relations existing in the Web of Linked Data from Web content. These patterns are used to populate the ontology core schema with new entities and attributes values. The experiments of the proposal have shown promising results in precision.

Domaines: Autre [cs.OH]

Ilaria Tiddi, Nesrine Ben Mustapha, Yves Vanrompay, Marie-Aude Aufaure. Ontology Learning from Open Linked Data and Web Snippets. Confederated International Workshops: OTM Academy, Industry Case Studies Program, EI2N, INBAST, META4eS, OnToContent, ORM, SeDeS, SINCOM, and SOMOCO 2012, Sep 2012, rome, Italy. pp.434-443, ⟨10.1007/978-3-642-33618-8_59⟩. ⟨hal-00831694⟩