摘要
词语的情感倾向判别是文章语义情感倾向研究的基础工作。利用中文情感词建立一个基础情感词典,为专一领域情感词识别提供一个核心子集,能够有效地在语料库中识别及扩展情感词集,并提高分类效果。在中文词语相似度计算方法的基础上,提出了一种中文情感词语的情感权值的计算方法,并以HOWNET情感词语集为基准,构建了中文基础情感词典。利用该词典结合TF-IDF特征权值计算方法,对中文文本情感倾向进行判别,实验结果表明,该方法取得了不错的分类效果。
Judging the emotional tendencies of Chinese words is the basic work of the semantic emotional tendency study of text. Building a basic emotional lexicon with Chinese emotional words will provide a core subset for identifying emotional words in a special area. It is able to identify and enlarge emotional word set effectively in corpus and also improve the efficiency of classification. A method of calculating the emotional value of Chinese emotional words on the basis of the similarity of Chinese words was provided. And also a Chinese basic emotional lexicon dictionary was constructed based on the HOWNET emotional word set. The emotional tendencies of Chinese texts were judged through the dictionary together with TFIDF. Experiments show that this method has achieved a satisfying result.
出处
《计算机应用》
CSCD
北大核心
2009年第10期2875-2877,共3页
journal of Computer Applications
基金
湖南省自然科学基金资助项目(05JJ30122)
中国包装总公司科研资助项目(2008-XK13)
湖南省教育厅科研资助项目(07B014)
湖南工业大学研究生创新基金资助项目(CX0812)
关键词
基础情感词词典
倾向性分析
情感权值
种子词
basic semantic lexicon
orientation analysis
semantic weight
seed word