저자 | Kyoungmin Kim, Jaedong Lee, and Jee-Hyong Lee |
---|---|
학회명 | International Symposium on Advanced Intelligent Systems |
학회명 (약자) | ISIS 2013 |
pp. | 36-43 |
학회시작일 | 2013-11-13 |
학회종료일 | 2013-11-16 |
비고 |
Abstract.
Sentiment classification is to identify whether the opinion expressed in a document is positive or negative. In case of the method using machine learning algorithms for sentiment classification, it requires gold standard annotations or opinion lexicons. In this paper we explore methods to automatically extract rationale words for sentiment classification via Genetic Algorithm. Experiments show that our method improves the performance of sentiment classification. Furthermore, we employ rationale words to identify most subjective sentences in the document. The result maintains the same level of performance for sentiment classification while retaining only subjective portion of the document.