Development of a method for judging the authenticity of content by extracting impressions from texts

2023 Fiscal Year Final Research Report

• Project/Area Number: 20K12028
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 61060:Kansei informatics-related
• Research Institution: The University of Tokushima
• Principal Investigator: MORITA Kazuhiro 徳島大学, 大学院社会産業理工学研究部(理工学域), 准教授 (20325252)
• Project Period (FY): 2020-04-01 – 2024-03-31
• Keywords: 印象知識

Outline of Final Research Achievements

With the spread of SNS, the importance of handling text data overflowing on the Internet is extremely high. On the other hand, phishing e-mails are designed to induce human carelessness. In addition, the spread of hoaxes and fake news has become a social problem every time a major incident such as a natural disaster occurs. In this study, in order to judge the authenticity of textual information alone, we developed and evaluated a technology to extract impressions from documents and judge their authenticity by collating classifications.

Free Research Field

感性情報処理，自然言語処理

Academic Significance and Societal Importance of the Research Achievements

人はある情報「地震で電車が脱線した」を見たときに，「大変だ」（内容を信用した肯定的印象）， 「本当なのか」（疑問を感じた否定的印象）といった何らかの感想（印象）を想起する。この印象に着目し，文章の信憑性を判定する学術的意義がある。また，テキストから受ける印象を起点とすることにより，信憑性辞書の変更のみで結果を制御できるため， SNS拡散への注意喚起，フィッシングメールへの対策のほか，誹謗中傷の検出などにも発展できる社会的意義がある。