Abstract

Sentiment analysis andopinion mining is an emerging area of research for analysing web data andcapturing the sentiment of the users. Sentimentanalysis is the calculative analysis of views, sentiments, opinions andpositivity or negativity of a text. This paper identifies the factors that areresponsible for the different sentiments of a person regarding a particularentity. In this work, the objective is to categorize the factors that influencethe system of sentiment analysis due to varied sentiments of an individual. The evaluation of theproposed model ISM, using the performance evaluation measures prove thatsentiment expressed in a post should be considered as an important factor foridentifying influential bloggers. The findings also indicate that ISM showsbetter results to find the influential bloggers as compared to the existingmethods.

Keywords:Sentiment Analysis, Level of SentimentAnalysis, Feature Extraction and Selection, ISM (Interpretive StructureModelling)