APPLICATION OF MACHINE LEARING IN SENTIMENT ANALYSIS

Abstract

There is ample amount of statements on social sites which can be inferred with the help of sentiment analysis. It is very beneficial to find the public opinion. Sentiment Analysis involves capturing of user’s behavior, likes and dislikes of an individual from the generated web content. There is no concrete definition of “Sentiments”, but in general they are considered as thoughts, views and attitude of a person arising mainly based on the emotion instead of a reason. Millions of users use social sites to express their sentiment about brands, services, political and religious views, emotions, beliefs or opinions about things, personalities or places and people they interact with. This data is mostly unorganized, slangs, etc. and therefore, text analytics and natural language processing are used to extract and classify this data. Any Non-contextual and irrelevant contents are identified and discarded. The classification of sentiments will be performed on this data, which goes as follows: a training data set is created manually and based on this training data set sentiment analysis is performed on the twitter comments. Machine learning such as a hybrid Naive Bayesian classifier is utilised with the lexical dictionary and natural language processing for the sentiment classification