期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：9
DOI：10.15680/IJIRCCE.2015. 0309076
出版社：S&S Publications
摘要：Due to the absolute amount of opinion loaded web assets such as debate meeting, analysis site, blogsand reports corpora existing in digital structure greatly of the existing study is focus on the region of sentiment analysis.People are planned to extend a structure that can classify and organize opinion or sentiment as represent in an electroniccontent. In this we tested on movie reviews, product reviews and Facebook comments. The consequences illustrate thata fusion arrangement can recover the organization value. In this paper, we calculate the pros and cons of the existingtechnique and improve the existing techniques with POS tagging and generate the score of the comments or words andpredict the sentiment of the comment and then the conclusion is devised.
关键词：Natural language processing; sentiment analysis; Linguistic rules; POS tagging; opinion mining.
