2013 | OriginalPaper | Buchkapitel
Disambiguating Wikipedia Articles on the Basis of plWordNet Lexico-semantic Relations
verfasst von : Maciej Piasecki, Michał Kaliński, Agnieszka Indyka-Piasecka
Erschienen in: Advances in Artificial Intelligence and Its Applications
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
A method for mapping Wikipedia articles (treated as a large informal resource describing concepts) to wordnet synsets (relation-based word meaning descriptions) is proposed. In contrast to previous approaches, the method focuses mainly on wordnet relation structure as the main means for meaning description. The description is supplemented by knowledge automatically extracted from a corpus in the form of Measure of Semantic Relatedness. Several different ways of building synset meaning descriptions are analysed. We noticed that the problem is similar to wordnet-to-wordnet mapping, as both resources utilise relation structure to describe meanings and provide partition into lexical meanings. A set of inter-resource semantic relations was introduced as a basis for the manual evaluation of the mapping algorithms. Several algorithms were tested on polysemous and monosemous articles from Polish Wikipedia.