情报杂志Issue(2)：136-139,85,5.DOI:10.3969/j.issn.1002-1965.2014.02.026
中文微博的情绪识别与分类研究
Study on Sentiments Recognition and Classification of Chinese Micro-blog
摘要
Abstract
This paper uses API data port of micro-blogging platform to collect data, which is then preprocessed by participle techniques as well as getting out stop words, thereby the vector space model is built based on TF-IDF. The paper uses document frequency and informa-tion gain to reduce the dimension of feature vectors, and then builds the optimal mood classifier by a comparative study of a variety of text categorization methods. Results show that text categorization methods based on machine learning algorithms are applicable for micro-blog research with large corpus, while SVM performs well in classifying fine-grained sentiments.关键词
微博/情绪识别/情绪分类/机器学习/支持向量机Key words
micro-blog/sentiment recognition/sentiment classification/machine learning/SVM分类
社会科学引用本文复制引用
何跃,邓唯茹,张丹..中文微博的情绪识别与分类研究[J].情报杂志,2014,(2):136-139,85,5.
