Skip to main content
WorldCist'18 - 6th World Conference on Information Systems and Technologies

Full Program »

A candidate generation algorithm for Named Entities Disambiguation using DBpedia

Word Sense Disambiguation is the process of choosing one sense to an ambiguous word in a context. Ambiguity refers to the fact that a word can have different meanings. One form of the lexical ambiguity is polysemy (Apple is the company and eventually the fruit). The state-of-art approaches generally extract named entities (NE), generate candidate entities from a Knowledge Base, and apply a comparison method to select the correct one. As a complement to the majority of those approaches which do not use the NE categories, we propose a disambiguation algorithm that uses those categories to reduce the number of the candidates. For instance, categories include person, location, organization, etc. we will show that considering them will considerably reduce the number of the resulting candidates. In this paper, we will focus on the step of generating the candidate entities from a KB, thus we will propose an algorithm that will use DBpedia to link NE categories to the values of rdf:type property. The obtained results are very promising.

Wissem Bouarroudj
LIRE Laboratory, Abdelhamid Mehri Constantine 2 University
Algeria

Zizette Boufaida
LIRE Laboratory, Abdelhamid Mehri Constantine 2 University
Algeria

 

Powered by OpenConf®
Copyright ©2002-2017 Zakon Group LLC