作者: Alireza Bitarafan , Chitra Dadkhah
DOI: 10.1109/ICWR.2019.8765274
关键词:
摘要: Online stores and e-commerce platforms have become increasingly popular in recent years, a reasonable approach to compare the available products is use comments or feedbacks written by other online users for each product. Therefore, these can be great opportunity spammers promote demote their target with fake reviews. So far, there plenty of studies done purpose distinguishing spam reviews from genuine ones, but it should not neglected that often work collusion control rating score product more naturally. Hence, this article focuses on latter aspect i.e., review spammer group detection. In most previous works, Frequent Item set Mining (FIM) applied early stage find candidate groups then an unsupervised ranking procedure based some predefined features. Although, FIM methods mostly suffer threshold setting, using low support values causes inefficiency high ignore useful patterns. Furthermore, instead methods, semi-supervised ones which don’t need many labeled data, improve accuracy detection greatly. article, we tackle above-mentioned challenges taking advantage instances Heterogeneous Information Network (HIN). Using HIN preserve semantics between different kinds nodes network. Also, extract behaviors relations makes robust when decide intelligent. Experiments real-life Yelp dataset show efficiency our approach.