Abstract
With the rapid development of e-commerce, most customers express their opinions in the internet. The number of reviews can be in hundreds or even thousands. This makes the customers to read all the reviews. A novel approach proposed to perform aspect based opinion mining in online product reviews. We are using word alignment model with rule based approach (Part-Of-Speech rules and syntactic rules) to identify opinion relations (relationships and association between opinion words and its opinion targets) as an alignment process. Then each candidate’s confidence is estimated through a graph-based co-ranking algorithm. Candidates with higher confidence are extracted as opinion targets or opinion words. Naive Bayes algorithm is used to identify the polarity the opinion words into positive, negative and neutral categories.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
