Potential issues and prospective solution for sentiment techniques

Abstract

With fast growing internet feedbacks regarding products, policies, services, people are being generated in large volume via different social networking sites or online portals. Data generated through these sites have great significance to numerous applications. But this bulk of content generated need to be assessed using appropriate mining techniques to extract information which can then proven to be useful. Sentiment analysis is the process of analyzing a text to know attitude of writer towards a text to be positive, negative or neutral. At one side sentiments are proven to be beneficial, on the other hand evaluation process of sentiment analysis is vulnerable to various challenges. These challenges act like a hurdle in interpreting the accurate intimation of sentiments and identifying their suitable polarity. This paper has put forward an overview of challenges faced in sentiment analysis and relevant techniques of some researchers acknowledging those issues. Probable solutions for various issues given by varied analysts have also been discussed in the paper.