[4Xin2] Poster session 2

[4Xin2-84] Classification of articles using Wikipedia categories and definition sentences

〇Nozomi Suzuki¹, Masaharu Yoshioka¹ (1.Hokkaido University)

Keywords:Wikipedia, Classification, Structured Knowledge, Category

There have been several attempts to extract knowledge from Wikipedia. One of the most important information to be extracted is the classification of articles. The Wikipedia category, which classifies the group of articles for ease of navigation, seems to be a good resource for this task. However, since Wikipedia categories are also used for different purposes and many categories are added to an article, it is necessary to select the representative Wikipedia category for better classification. In this paper, we propose to use the definition sentence, which is the first sentence of the article, to select the representative category among them. In this method, we extract the definition word, which is for representing the class of the article, from the definition sentence. We also propose a method to classify the article using this information and categories, and evaluate the method using the SHINRA dataset.