Two Fold Clustering Schema (TFCS) for Acquisition of Authentic Reviews in Web Crawlers

Nowadays, people will not stop their activity once they purchase the goods on a website. They will post their experience and emotions on a website as a review. The majority of customers search for product ratings on websites; they focus only on filtered products to save time, and as a result, many spammers post fake reviews online to increase the sales of their product. For this, we propose a method known as the Two-Fold Clustering Schema (TFCS). In this method, the first fold indices the web sites using a partition clustering algorithm based upon the personalized framework to produce the customized crawler, where the websites are filtered to produce only the product review web pages posted by the customers on the e-commerce website, and then in the second fold of agglomerative clustering, reviews are grouped based on their properties, relevant features are extracted using the classifier module, where the authentic review is filtered and the remaining reviews are discarded as they are fake. In the first clustering activity, the reviews are selected, and in the second clustering, the prepared dataset of reviews produces the genuine review. Existing scraping methods have dependency issues, but in this system there are no dependency issues because it retrieves the data dynamically from search engines, so there are no dependencies and it is very efficient and accurate in terms of human generated and machine generated reviews.