EDL: A Framework for Entity Disambiguation and Linking to Knowledgebases

Abstract

Extraction and integration of entities from textual data and linking them to knowledgebases (for further information or processing) is useful for many applications in natural language processing. However, a major problem in this process is disambiguation, named entities might refer to different things. In this work, we propose a novel method to disambiguate named entities. Our method is a combination of search engine results and knowledgebase repository mining results. We obtained 84% correct disambiguation rates with the data sets we used.