A Review on Different Opinion and Aspect Mining Techniques

With the rising popularity of internet, online drug reviews have been proved to be extremely helpful for patients suffering from chronic diseases. Most of the patients search upon online reviews before taking any medicine. Online reviews, blogs, and discussion forums such as WebMD on chronic diseases and medicines are becoming important supporting resources for patients. Extracting useful information from these reviews is very difficult and challenging. Opinion mining or aspect mining involves the extraction of useful information (e.g. positive or negative sentiments of a product) from a large quantity of text opinions or reviews given by Internet users. Various algorithms had been proposed to extract information from the opinion of web users. Some of the algorithms are LDA, sLDA, NMF, SSNMF, DiscLDA and PAAM. A detailed review of the most important opinion mining algorithms is presented and a comparison among the discussed techniques is given.