Feature Based Sentiment Analysis of Compound Sentences

Sentiment analysis is the study that analyzes people's opinions, sentiments, attitudes and emotions towards entities such as products, services, organizations, individuals and events. The rise of blogs and social networks has fueled bull market in personal opinions, reviews, ratings, recommendations and other forms of online expressions. Sentiment analysis is accomplished by examining this large amount of data on the basis of positive, negative and neutral views of the end user. Feature based sentiment analysis focuses on different aspects or features of any entity under analysis. The reviews, attitudes, opinions and emotions are available on the web in the form of simple and compound sentences. Feature based sentiment analysis of simple sentences is easy compared to the analysis of compound sentences which contain multiple clauses. In this paper, we propose rules for analyzing different conjunctions and propose an algorithm to perform feature based sentiment analysis of compound sentences. Rules are defined for the different conjunctions used to form compound sentences. The algorithm analyses different conjunctions used in the compound sentences based on the rules and provides a feature based summary of different attributes of the entity under analysis.