Extracting Product Features from Online Consumer Reviews

The exponential growth of user-generated content in online environment calls for techniques that can help to make sense of the content. Despite of a host of research on online consumer reviews, there is still a great demand for research to improve the techniques for feature extraction. To this end, we proposed extraction methods based on detailed categorization of review features. By taking into account of the characteristics and patterns of different types of features, the proposed methods not only identify new features but also filter irrelevant features. The results of an experiment demonstrate that our proposed methods outperform the state-of-the-art techniques.