[4Rin1] Interactive 2

[4Rin1-07] News Article Retrieval Related to Stock Price Change Using Polarity Information

〇Kazuto Tanaka¹, Rie Yatabe¹, Minoru Sasaki¹, Tomoya Suzuki^1,2 (1.Ibaraki University, 2.Daiwa Asset Management Corporation)

Keywords:News Article Retrieval, Stock Price Change, Polarity Information

In recent years, various economic reports are published and they are important tools in all markets. However, it is necessary to consult a lot of news articles in order to write an economic report. In this paper, we propose a method for effectively extracting news articles including important events related to the fluctuation of the Nikkei Stock Average by using Japanese sentiment lexicons. The results of the experiments show that the proposed method reduces the number of articles by about half and retrieves relevant documents better than the method using only stock price fluctuations. Therefore, the Japanese sentiment lexicons is effective for extracting news articles including important events.