A Parameterized Approach to Deal with Review Spammers

Today, E-commerce web sites are providing huge number of platforms for users in which they can express their views, their opinions and post their reviews about the products on the web. Such content contributed by users is available for other customers and manufacturers as a valuable source of information. Though these reviews are important source of information but quality control on this user generated data is not assured. As the popularity of Ecommerce sites are immensely increasing, quality of the reviews is getting worse day by day thereby affecting customers' buying decisions. Spammers may either post positive comments to promote a product/brand or negative comments having intention of demoting a product/brand. Unfortunately, many organizations are making money by doing such activities. Spammers are also heavily paid by such organizations for writing fake reviews for the target products on web. In this work, we take a different approach to detect suspicious review, suspicious reviewers and suspicious reviewers' group considering geographical statistics and networking parameters.