A Lexicon Based Method for Opinion Mining

Nowadays every user wants to know about particular product before buying, movie reviews before watching to confirm whether it is good or bad. The developers are also interested to know about their products or movies based on the user reviews. For that purpose the sentiment analysis is very useful.

The sentiment analysis is very important to know about the product reviews, movie reviews, tweets and so on. Based on that reviews, the user can classify whether it is good or bad (positive opinion or negative opinion). So the sentiment analysis plays an important role in human life.

This paper describes the simple lexicon based approach for classifying the sentence to positive or negative or neutral. Our aim in sentiment analysis is to produce summary of opinion based on product features and reviews. For this process we create a lexicon. The lexicon contains positive, negative, negation, blind negation words and emoticon list. The product reviews, movie reviews, tweets contain word variations, emoticons, hash tags etc. We perform some steps to process the sentence that has hash tags and exaggerated word shortening. After the pre-processing our lexicon based approach tells that the sentence is positive or negative or neutral based on the adjectives present in the sentence.