Spam Review Detection with Graph Convolutional Networks

Li, Ao; Zhou, Qin; Liu, Runshi; Yang, Yiqun; Li, Dong

doi:10.1145/3357384.3357820

Cited by 192 publications

(95 citation statements)

References 25 publications

(22 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…(3) Existing method relying on reviews, items and users. GAS [15] first applies GNN to spam review detection in e-commerce scenarios. We include GAS as the baseline since it is the only work that tries to apply graph neural network to online review classification.…”

Section: Methodsmentioning

confidence: 99%

“…To detect spam reviews in a second-hand goods app, Li et al [15] further involved the information of the users and constructed a user-review-item graph. They formulated the problem as an edge classification task, and proposed GAS, which employed GNNs on the heterogeneous graph to further mine the high-order information in the graph, and generated a representation for each edge.…”

Section: Reviews + Items + Usersmentioning

confidence: 99%

“…(2) Fan et al [7] incorporated more information of the item and used deep neural networks to better predict the review helpfulness. (3) Li et al [15] further involved the information of the users and adopted a graph structure to organize the multiple data sources. Based on this graph representation, the authors modeled the target as an edgeclassification problem and proposed to use graph neural networks to train and classify the reviews.…”

Section: Introductionmentioning

confidence: 99%

“…Motivated by this, we propose CA-GNN to tackle both of the above two challenges, using graph neural network to identify helpful reviews in a multi-task manner. To combine multiple data sources, we follow [15] to organize them as a user-review-item graph. To handle the heterogeneity of criteria for classifying reviews, we employ GNN with one shared and many item-specific graph convolutions to learn the common features and each item's specific criterion for classifying reviews simultaneously.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Category-aware Graph Neural Networks for Improving E-commerce Review Helpfulness Prediction

Xiaoru

Wang

et al. 2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

Helpful reviews in e-commerce sites can help customers acquire detailed information about a certain item, thus affecting customers' buying decisions. Predicting review helpfulness automatically in Taobao is an essential but challenging task for two reasons: (1) whether a review is helpful not only relies on its text, but also is related with the corresponding item and the user who posts the review, (2) the criteria of classifying review helpfulness under different items are not the same. To handle these two challenges, we propose CA-GNN (Category-Aware Graph N eural N etworks), which uses graph neural networks (GNNs) to identify helpful reviews in a multi-task manner-we employ GNNs with one shared and many item-specific graph convolutions to learn the common features and each item's specific criterion for classifying reviews simultaneously. To reduce the number of parameters in CA-GNN and further boost its performance, we partition the items into several clusters according to their category information, such that items in one cluster share a common graph convolution.We conduct solid experiments on two public datasets and demonstrate that CA-GNN outperforms existing methods by up to 10.9% in AUC. We also deployed our system in Taobao with online A/B Test and verify that CA-GNN still outperforms the baseline system in most cases. CCS CONCEPTS • Information systems → Data mining; Electronic commerce; • Computing methodologies → Supervised learning; • Applied computing → Document management and text processing.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Reviews + Items + Usersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Category-aware Graph Neural Networks for Improving E-commerce Review Helpfulness Prediction

Xiaoru

Wang

et al. 2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

show abstract

“…Player2Vec leverages GCN & GAT to encode the intra-& inter-relation neighbor information. GAS [19] learns unique aggregators for different node types and updates the embeddings of each node types iteratively.…”

Section: Related Workmentioning

confidence: 99%

Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters

Dou

Liu

et al. 2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

262

160

View full text Add to dashboard Cite

Graph Neural Networks (GNNs) have been widely applied to fraud detection problems in recent years, revealing the suspiciousness of nodes by aggregating their neighborhood information via different relations. However, few prior works have noticed the camouflage behavior of fraudsters, which could hamper the performance of GNNbased fraud detectors during the aggregation process. In this paper, we introduce two types of camouflages based on recent empirical studies, i.e., the feature camouflage and the relation camouflage. Existing GNNs have not addressed these two camouflages, which results in their poor performance in fraud detection problems. Alternatively, we propose a new model named CAmouflage-REsistant GNN (CARE-GNN), to enhance the GNN aggregation process with three unique modules against camouflages. Concretely, we first devise a label-aware similarity measure to find informative neighboring nodes. Then, we leverage reinforcement learning (RL) to find the optimal amounts of neighbors to be selected. Finally, the selected neighbors across different relations are aggregated together. Comprehensive experiments on two real-world fraud datasets demonstrate the effectiveness of the RL algorithm. The proposed CARE-GNN also outperforms state-of-the-art GNNs and GNN-based fraud detectors. We integrate all GNN-based fraud detectors as an opensource toolbox 1. The CARE-GNN code and datasets are available at https://github.com/YingtongDou/CARE-GNN. CCS CONCEPTS • Security and privacy → Web application security; • Computing methodologies → Neural networks.

show abstract

A Graph Attentive Network Model for P2P Lending Fraud Detection

Wang

Liu

et al. 2020

Knowledge Science, Engineering and Management

View full text Add to dashboard Cite

Spam Review Detection with Graph Convolutional Networks

Cited by 192 publications

References 25 publications

Category-aware Graph Neural Networks for Improving E-commerce Review Helpfulness Prediction

Category-aware Graph Neural Networks for Improving E-commerce Review Helpfulness Prediction

Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters

A Graph Attentive Network Model for P2P Lending Fraud Detection

Contact Info

Product

Resources

About