Using Function Tags to Recognize Opinion Elements

カテゴリ: 部門大会

論文No: GS16-5

グループ名: 【C】平成21年電気学会電子・情報・システム部門大会講演論文集

発行日: 2009/09/03

タイトル(英語): Using Function Tags to Recognize Opinion Elements

著者名: 袁彩霞 (徳島大学)

著者名(英語): Caixia Yuan(Tokushima University)

キーワード: Function Tags|Opinion Mining|Opinion Elements

要約(日本語): The challenge of automatically identifying opinions in text has been the focus of attention in recent years in many different domains such as news articles and product reviews. Various approaches have been adopted in subjectivity detection, semantic orientation detection and review mining. Despite the successes in identifying opinion expressions and subjective words/phrases, there has been less achievement on the factors closely related to subjectivity and polarity, such as opinion holder, topic of opinion, and inter-topic/inter-opinion relationships. This study addresses the problem of identifying not only opinions but also opinion holders from news articles. We apply our function parser to this problem and show its applicability in an important subcomponent of text opinion mining. When testing in the MOAT dataset of NTCIR-6, our method get an accuracy of 75.9% in opinion element recognization, which is comparable with the best result ever reported.

PDFファイルサイズ: 4,892 Kバイト