Aspect based Sentiment Analysis using support vector machine classifier

Abstract

Sentiment Analysis involves the process of identifying the polarity of opinionated texts. Lots of social networking sites are being used for expressing thoughts and opinions by users to rate products. These user opinionated text is highly unstructured in nature and thus involves the application of various natural language processing techniques. In aspect based sentiment analysis, the various features of a product is identified through the training process. For e.g. the aspects of a camera are picture quality, size, resolution etc. The quantitative analysis of each aspect is done using support vector machine classifier. In most of the previous works, a product review is analysed as a whole rather than considering each aspect of it. Aspect based opinion mining is tedious since the identification of individual features is in itself a challenging task.