An Entity Disambiguation Approach Based on Wikipedia for Entity Linking in Microblogs

The opportunity to read articles and microblogs on the Web to get information is more and more increasing. However, hyperlinks to entities do not often exist in such articles, and it is a troublesome task for the reader to look it up online. In this paper, in order to make it easy to look up entity information in microblog articles, we propose a method to extract entities in Japanese microblog, and to perform entity linking which links to entity information automatically. The method consists of three phases. First, we extract named entities, such as personal names, place names, organization names, etc. from a microblog article. Next, we disambiguate the extracted entities in order to make links to correct entity information. We use Wikipedia as the source of entity information to verify the usefulness of the proposed method. In our method, we extract some Wikipedia articles related to ambiguous entities from microblog articles. Then, we extract some related entities with the ambiguous entity using word2vec. We compare Wikipedia articles of related entities and the Wikipedia articles of ambiguous entities. Finally, we get the correct Wikipedia article for each entity in the microblog article.