ColluEagle: collusive review spammer detection using Markov random fields
- PDF / 1,132,584 Bytes
- 21 Pages / 439.37 x 666.142 pts Page_size
- 30 Downloads / 235 Views
ColluEagle: collusive review spammer detection using Markov random fields Zhuo Wang1
· Runlong Hu1 · Qian Chen1 · Pei Gao1 · Xiaowei Xu2,3
Received: 10 December 2018 / Accepted: 16 May 2020 © The Author(s), under exclusive licence to Springer Science+Business Media LLC, part of Springer Nature 2020
Abstract Product reviews are extremely valuable for online shoppers in providing purchase decisions. Driven by immense profit incentives, fraudsters deliberately fabricate untruthful reviews to distort the reputation of online products. As online reviews become more and more important, group spamming, i.e., a team of fraudsters working collaboratively to attack a set of target products, becomes a new fashion. Previous works use review network effects, i.e. the relationships among reviewers, reviews, and products, to detect fake reviews or review spammers, but ignore time effects, which are critical in characterizing group spamming. In this paper, we propose a novel Markov random field (MRF)-based method (ColluEagle) to detect collusive review spammers, as well as review spam campaigns, considering both network effects and time effects. First we identify co-review pairs, a review phenomenon that happens between two reviewers who review a common product in a similar way, and then model reviewers and their co-review pairs as a pairwise-MRF, and use loopy belief propagation to evaluate the suspiciousness of reviewers. We further design a high quality yet easyto-compute node prior for ColluEagle, through which the review spammer groups can also be subsequently identified. Experiments show that ColluEagle can not only detect collusive spammers with high precision, significantly outperforming state-ofthe-art baselines—FraudEagle and SpEagle, but also identify highly suspicious review spammer campaigns. Keywords Fake review detection · Review spammer detection · Group spamming · Markov random field · Loopy belief propagation
Responsible editor: G. Karypis
B
Zhuo Wang [email protected]
1
School of Information Science and Engineering, Shenyang Ligong University, Shenyang, China
2
Shenyang Ligong University, Shenyang, China
3
University of Arkansas at Little Rock, Little Rock, USA
123
Z. Wang et al.
1 Introduction Online product reviews are increasingly influencing customers’ purchase decisions, and thereby influencing product sales. To promote or demote product reputations, review spammers try to game the review websites by posting untruthful review content and/or rating stars. Ordinary customers have much difficulties in distinguishing fake reviews from genuine ones, as a result, are vulnerable to review spamming. Nowadays, as the word-of-mouth marketing prevails, group spamming, i.e., a group of review spammers working together to promote or demote a set of target products, is becoming the new form of review spamming. Over the years, researchers proposed various techniques to detect spam reviews, review spammers, or spammer groups. However, the problem is far from being solved because the underlying mechanism
Data Loading...
