Subgroup Fairness in Graph-based Spam Detection

Fake reviews are prevalent on review websites such as Amazon and Yelp. GNN is the state-of-the-art method that can detect suspicious reviewers by exploiting the topologies of the graph connecting reviewers, reviews, and target products. However, the discrepancy in the detection accuracy over different groups of reviewers causes discriminative treatment of different reviewers of the websites, leading to less engagement and trustworthiness of such websites. The complex dependencies over the review graph introduce difficulties in teasing out subgroups of reviewers that are hidden within larger groups and are treated unfairly. There is no previous study that defines and discovers the subtle subgroups to improve equitable treatment of reviewers. This paper addresses the challenges of defining, discovering, and utilizing subgroup memberships for fair spam detection. We first define a subgroup membership that can lead to discrepant accuracy in the subgroups. Since the subgroup membership is usually not observable while also important to guide the GNN detector to balance the treatment, we design a model that jointly infers the hidden subgroup memberships and exploits the membership for calibrating the target GNN's detection accuracy across subgroups. Comprehensive results on two large Yelp review datasets demonstrate that the proposed model can be trained to treat the subgroups more fairly.

Interpretable and Effective Reinforcement Learning for Attacking against Graph-based Rumor Detection

Social networks are polluted by rumors, which can be detected by machine learning models. However, the models are fragile and understanding the vulnerabilities is critical to rumor detection. Certain vulnerabilities are due to dependencies on the graphs and suspiciousness ranking and are difficult for end-to-end methods to learn from limited noisy data. With a black-box detector, we design features capturing the dependencies to allow a reinforcement learning to learn an effective and interpretable attack policy based on the detector output. To speed up learning, we devise: (i) a credit assignment method that decomposes delayed rewards to individual attacking steps proportional to their effects; (ii) a time-dependent control variate to reduce variance due to large graphs and many attacking steps. On two social rumor datasets, we demonstrate: (i) the effectiveness of the attacks compared to rule-based attacks and end-to-end approaches; (ii) the usefulness of the proposed credit assignment strategy and control variate; (iii) interpretability of the policy when generating strong attacks.