User-based Network Embedding for Collective Opinion Spammer Detection

TL;DR

This paper introduces an unsupervised network embedding method that effectively detects collective opinion spammers by leveraging multiple user relation types, significantly outperforming existing solutions on Amazon and Yelp datasets.

Contribution

The paper proposes a novel unsupervised network embedding approach that jointly exploits different user relations for improved spammer detection.

Findings

Average AP improvements of 14.09% and 16.25% on datasets.

Average AUC improvements of 12.04% and 12.78%.

Effective detection of collaborative opinion spam campaigns.

Abstract

Due to the huge commercial interests behind online reviews, a tremendousamount of spammers manufacture spam reviews for product reputation manipulation. To further enhance the influence of spam reviews, spammers often collaboratively post spam reviewers within a short period of time, the activities of whom are called collective opinion spam campaign. As the goals and members of the spam campaign activities change frequently, and some spammers also imitate normal purchases to conceal identity, which makes the spammer detection challenging. In this paper, we propose an unsupervised network embedding-based approach to jointly exploiting different types of relations, e.g., direct common behaviour relation and indirect co-reviewed relation to effectively represent the relevances of users for detecting the collective opinion spammers. The average improvements of our method over the state-of-the-art solutions on dataset AmazonCn and YelpHotel are [14.09%,12.04%] and [16.25%,12.78%] in terms of AP and AUC, respectively.