期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2016
卷号：5
期号：5
页码：7392
DOI：10.15680/IJIRSET.2016.0505134
出版社：S&S Publications
摘要：Internet is accessible to almost everyone these days and therefore people spend most of their time insurfing the internet whether it is for education, entertainment, online shopping, etc. Not only do they use the availableinformation but they can also give feedbacks or suggestions. As a result there are various reviews available onlinewhich can benefit both, the customers and the manufacturers. Customers can use it for making smart purchasedecisions and the manufacturers can take its help for locating the areas of their products which require improvements.These reviews are basically the opinions of various individuals. Opinion mining is a discipline that requires naturallanguage processing, information retrieval and data mining. It aims at classifying the opinions available on the web aspositive, negative or neutral. It is done at three different levels- sentence, document and feature. Feature based opinionmining aims at summarizing the opinions for each feature of a product and then classifying it as positive, negative orneutral. It helps in determining which features of the products do customers like or dislike. This paper discusses variousmethods proposed for the feature based opinion mining.
关键词：Feature based opinion mining; opinion mining; polarity detection; sentiment classification.
