Sentiment Analysis Using Harn Algorithm

Sentiment Analysis, is the field which has been looked into a great depth recently. There has been a lot of work in identification of polarities. There are many techniques developed by researchers to classify the opinionated data. During our research we have found out that there is no technique which could give 100% accuracy. There are many challenges to Sentiment Analysis like semantic ambiguity statements, comparison sentences, domain specific adaptation, sarcastic statements etc. We have found a solution to one of the limitations, thereby contributing to the elevation of accuracy level. The basic problem that we have identified is polarity switching of a word based on the domain in which it is being used (Domain Specific Adaptation). To solve this problem we have developed an algorithm which could correctly classify the statement based on the domain in which the word is being used. The proposed algorithm is HARN algorithm. It is an unsupervised learning method which uses basic structure of the sentences, domain dictionaries and pre-defined polarities to classify the given sentence. The below report discusses the present existing approaches to sentiment classification, HARN algorithm and its implementation details. Keywords: sentiment analysis; Stanford parser, NV dictionaries, data mining.