A Rule Based Sentiment Analysis System for Hindi Language | Original Article

ABSTRACT:

Hindi is one of the most spoken languages of the world. Today Hindi language users has better input mechanism to express their sentiments easily on Social Media, so a large volume of User Generated Content in Hindi are digitally store on the internet. It is being seen as an important source of information, but no computational system has yet been available to analyzing these contents. A Sentiment Analysis system has been proposed to solve this problem that analyzes these Hindi contents automatically. The basic principles of Software Engineering and Natural Language Processing have been implemented to design this system. It is a rule based system that follows some linguistics rules to classify any input text into Positive, Negative or Neutral. To evaluate this system, a dataset of 4000 sentences has been created by compiling User Generated Content from Twitter and e-Newspapers. The accuracy to Polarity Classification of the system for Known and Unknown dataset has been measured about 69% and 52%, respectively