Detecting Deceptive Reviews Utilizing Review Group Model

Li, Yuejun and Wang, Fangxin and Zhang, Shuwu and Niu, Xiaofei (2019) Detecting Deceptive Reviews Utilizing Review Group Model. In: Mobimedia 2019, 29-30 Oct 2019, Wehai, China.

[thumbnail of PDF]
Text (PDF)
eai.29-6-2019.2282593.pdf - Published Version
Available under License Creative Commons Attribution No Derivatives.

Download (502kB) | Preview


Online product and store reviews play an important role in product and service recommendation for new customers. However, due to economic or fame reasons, dishonest people are employed to write fake reviews which is also called “opinion spamming” to promote or demote target products and services. Previous research has used text similarity, linguistics, rating patterns, graph relation and other behavior for spammer detection. It is difficult to find fake reviews by a glance of product reviews in time-descending order while It’s more easy to identify fraudulent reviews by checking the list of reviews of reviewers. We propose sieries of novel review grouping models to identify both positive and negative deceptive reviews. The review grouping algorithm can effectively split reviews of reviewer into groups which participate in building new model of review spamming detection. Several new features which are language independent based on group model are constructed. Additionally, we explore the collusion behavior between reviewers to build group collusion model. Experiments and evaluations show that the review group method and relevant models can effectivly improve the precision of 4%-7% in deceptive reviews detection task especially those posted by professional review spammers.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: deceptive review detection opinion spamming review group detection reviewer collusion
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
QA75 Electronic computers. Computer science
Depositing User: EAI Editor I.
Date Deposited: 10 Sep 2020 15:07
Last Modified: 10 Sep 2020 15:07

Actions (login required)

View Item
View Item