A NEW METHOD TO FIND SCORE VALUE FOR ONLINE OPINIONS
International Journal of Computational Science and Information Technology (IJCSITY), 1 (1):
1- 9 (February 2013)Abstract
In today’s world, we need information, not data to take a decision about a product. Opinions are data and become information after filtering. Also organizations products superiority is dependent on customer feedback of their products. Web is a platform providing facilities to post opinions for any product. A customer gives feedback when he uses a product and feedback is what customer feels after using it called sentiments/opinions. Opinions are subjective sentences and not objective as facts. People uses natural language to express sentiments and way of expressing each customer is different. Now it’s inevitable in current and future era to do sentiment analysis called opinion mining. So that, what customer is saying about a product whether enhancing reputation or not, will be posted on web. Opinion Mining and scoring of opinions in fast competitive world is required as opinions posted for a trendy product can be in hundreds. If these opinions not analyzed properly and have been thrown as is it is on web then it’s not possible for anyone to read all opinion sentences. So, a system is needed which automatically mines online opinions from web and score them (score value) properly correspond to positive and negative opinion words.