情報処理学会 第77回全国大会講演要旨
3C-03
Entity-oriented materials and opinions extraction from news articles
For detecting the inconsistency and alerting the bias in news contents, we propose a method of extracting materials and opinions from news articles.
We extract materials and opinions by 1) classifying entity-related phrases into objective and subjective descriptions, and 2) discovering opinion holders in subjective descriptions. The extracted materials and opinions help us construct the Material-opinion Model for difference analysis of news contents.
A prototype system is implemented and we also show some experimental results to validate the proposed methods.
We extract materials and opinions by 1) classifying entity-related phrases into objective and subjective descriptions, and 2) discovering opinion holders in subjective descriptions. The extracted materials and opinions help us construct the Material-opinion Model for difference analysis of news contents.
A prototype system is implemented and we also show some experimental results to validate the proposed methods.
