HOMS: Hindi opinion mining system

With the increasing popularity of the Web 2.0, we are provided with more documents which express opinions on different issues. Online posting reviews has become an increasingly preferred way for people to express opinions and sentiments towards the products bought/used or services received. Analysing the large volume of online review data available, would produce useful knowledge, which could be of economic values to vendors and other interested parties. A lot of work in Opinion Mining exists for English language. In the last few years, web contents are increasing in other languages also at a faster rate and hence there is a requirement to execute opinion mining in other languages. In this paper, a Hindi Opinion Mining System (HOMS) is proposed for movie review data. It performs the task of opinion mining at the document level and classifies the documents as positive, negative and neutral using two different methods: Machine learning technique and Part-Of-Speech (POS) tagging. We have used Naive Bayes Classifier for Machine learning and in POS tagging, we have considered adjectives as opinion words. Extensive simulations conducted on a large movie data set confirms the effectiveness of the proposed approach.