Research Article Open Access

Survey of Various Opinion Mining Approaches

Abstract

Opinion mining or sentiment analysis extract specified information from a large amount of text or reviews given by the internet users. Opinion mining classifies the large text of opinions as positive (good), negative (bad) or neutral. According to the number of positive, negative and neutral reviews, the product or service will be rated. Sometimes an overall rating for a review cannot be helpful to identify various features of a product or service. For example, a camera may come with excellent battery life but poor image quality. Hence more sophisticated aspect level opinion mining approaches have been proposed to extract information from online reviews. In this paper, we are discussing various approaches used for opinion mining. They are frequency-based approach, relation-based approach, supervised learning and topic modelling.

Gayathri R Krishna, Jothi S, Minojini N, Sowmiyaa P

To read the full article Download Full Article | Visit Full Article