SoftMax based User Attitude Detection Algorithm for Sentimental Analysis

The social microblogging sites empower users to more boldly express their views and reviews on various topics. This has led to the development of a new subdomain in the field of sentiment analysis which can be used for the benefit of various business. The business values associated with mining the user sentiments havefueled a profusion of researchers, companies etc. to get involved in the process to build, update and even promote their products. A majority of the existing methods utilize only the semantics in the user comments to mine the sentiments. The reliability of the comment is not taken into account in the previously existing methods which can severely undermine the sole purpose of sentiment analysis. In this work, SoftMax based attitude detection algorithm is proposed to identify the user nature efficiently. The reliability and the accuracy of the sentiment prediction can be substantially increased with taking user attitude into account. The proposed algorithm is evaluated on tweets fetched from micro blogging website twitter.