Modified aspect/feature based opinion mining for a product ranking system

The extensive and tremendous use of the Internet has led to the trust in word-of-mouth through social networking sites. A product that has best reviews are preferred to the similar kind of product with few favorable reviews. These consumer reviews are rich in knowledge but are unstructured since they are typed by common people using these sites. Aspect/Feature based Opinion Mining is modified in this developed system to help consumers and entrepreneurs in the analysis on these user reviews by determining the strengths and weaknesses of any products. The main motto of the project is ranking the products and its important aspects which would eventually gear up a faster decision-making. The developed system includes four important phases: Pre-processing, Enhanced Aspect Identification and Opinion Word Extraction with modified naïve bayes model, Aspect Polarity Identification, Products and Aspects Ranking in sequence. The developed system will be beneficial for the consumers and business analysts/entrepreneurs who seek customers' online opinions for analyzing purchase decisions and market statistics and strategies.