An Efficient Network-Based Spam Detection Structure for Reviews In Online Social Media

Abstract

Todays, a major part of everyone trusts on content in social media like opinions and feedbacks of a topic or a product. The liability that anyone can take off a survey give a brilliant chance to spammers to compose spam surveys about products and services for various interests. Recognizing these spammers and the spam content is a wildly debated issue of research and in spite of the fact that an impressive number of studies have been done as of late toward this end, yet so far the procedures set forth still scarcely distinguish spam reviews, and none of them demonstrate the significance of each extracted feature type. In this investigation, we propose a novel structure, named NetSpam, which uses spam highlights for demonstrating review datasets as heterogeneous information networks to design spam detection method into a classification issue in such networks. Utilizing the significance of spam features help us to acquire better outcomes regarding different metrics on review datasets. The outcomes demonstrate that NetSpam results the existing methods and among four categories of features; including review-behavioral, user-behavioral, review-linguistic, user-linguistic, the first type of features performs better than the other categories. The contribution work is when user search query it will display all top-k products as well as recommendation of the product