• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to project page

2019 Fiscal Year Final Research Report

Opinion Mining from Texts Including Quoted Others' Opinions

Research Project

  • PDF
Project/Area Number 17K00298
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Research Field Intelligent informatics
Research InstitutionJapan Advanced Institute of Science and Technology

Principal Investigator

Shirai Kiyoaki  北陸先端科学技術大学院大学, 先端科学技術研究科, 准教授 (30302970)

Project Period (FY) 2017-04-01 – 2020-03-31
Keywordsオピニオンマイニング / 極性判定 / 機械学習
Outline of Final Research Achievements

This research project aims at the polarity classification of texts (blog articles) including quotation of other articles toward opinion mining of social issues. First, a quoted text is extracted from a given blog article. Next, the polarity of a text written by a user and a quoted text written by others is identified. Then, a quotation type (“agree”, “disagree”, or “unrelated”) is identified, which stands for relation between the original and quoted texts. Finally, the polarity of the whole blog article is determined by considering the above results. In the experiments, the accuracy of the polarity classification of the proposed method was 0.942, which largely outperformed the baseline (0.893).

Free Research Field

自然言語処理

Academic Significance and Societal Importance of the Research Achievements

不特定多数のユーザが書いたテキストから特定の対象に対するユーザの意見や評判を明らかにするオピニオンマイニングは重要な研究課題である.特に時事問題を対象にしたオピニオンマイニングは,社会情勢を低コストで把握することができるため有用である.本研究課題は,他者の記事の引用を含むテキストの極性判定の精緻化によりオピニオンマイニングの正確性を向上させるものであり,その社会的意義は大きい.
これまでの極性判定の研究では,判定対象となるテキストの部分の性質の違いに着目した研究は少ない.本研究課題は,引用されたテキストとそうでないテキストを分けて処理することで極性判定の性能を向上させる点に学術的意義がある.

URL: 

Published: 2021-02-19  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi