A Fuzzy Approach for Opinion Summarization of Product Reviews

In the past few years, there has been tremendous increase in the amount of opinions posted by reviewers on various social networking sites. This explosion on Web has led to the need of opinion mining so that mined information from these unstructured reviews can be provided to the users for effective decision making. Generation of summary from the available reviews on various e-commerce sites like amazon, flipkart, e-bay etc. is a challenging task. This paper proposes an ontology-based approach for product's feature identification and then identified features are scored using fuzzy logic technique to provide a pictorial feature-based summary to the buyers. Further every review is classified as low, medium or high according to the range of computed review score. To evaluate the proposed work, review text of 11 products is extracted from the available review data of amazon site. The performance metrics such as accuracy, precision, recall and f-measure proves that the proposed system is efficient.