Research Article
Detecting Spammer Communities Using Network Structural Features
@INPROCEEDINGS{10.1007/978-3-030-00916-8_61, author={Wen Zhou and Meng Liu and Yajun Zhang}, title={Detecting Spammer Communities Using Network Structural Features}, proceedings={Collaborative Computing: Networking, Applications and Worksharing. 13th International Conference, CollaborateCom 2017, Edinburgh, UK, December 11--13, 2017, Proceedings}, proceedings_a={COLLABORATECOM}, year={2018}, month={9}, keywords={Fake reviews Spammer community The comment-based reviewer network Network structural features}, doi={10.1007/978-3-030-00916-8_61} }- Wen Zhou
Meng Liu
Yajun Zhang
Year: 2018
Detecting Spammer Communities Using Network Structural Features
COLLABORATECOM
Springer
DOI: 10.1007/978-3-030-00916-8_61
Abstract
Spammers generate fake reviews to influence the reputation of products. By grouping together, spammers can dramatically alter how products are perceived. Different from previous research, which has mostly used behavioral indicators and structural indicators, we propose a new perspective on spammer detection. In our approach, we portray reviewers as a comment-based reviewer network through a new collusion similarity measure, divide reviewers into different communities using an effective community detection method and separate spammer communities from normal reviewer communities through network structure. We find that spammer communities have different network structural features from normal reviewer communities, a high clustering coefficient and high self-similarity. In our experiments, we show that our method achieves a detection accuracy of 94.59% - substantially higher than the current state-of-the-art methods which achieve an 80.00% accuracy.
