RoSGAS: Adaptive Social Bot Detection with Reinforced Self-Supervised GNN Architecture Search

TL;DR

RoSGAS introduces a reinforcement learning-based framework that adaptively searches for optimal GNN architectures for social bot detection, leveraging self-supervised learning to improve embedding discrimination and model performance.

Contribution

The paper presents RoSGAS, a novel adaptive GNN architecture search method that combines reinforcement learning and self-supervised learning for improved social bot detection.

Findings

Outperforms state-of-the-art methods in accuracy on Twitter datasets.

Achieves better training efficiency and stability.

Demonstrates improved generalization to unseen samples.

Abstract

Social bots are referred to as the automated accounts on social networks that make attempts to behave like human. While Graph Neural Networks (GNNs) has been massively applied to the field of social bot detection, a huge amount of domain expertise and prior knowledge is heavily engaged in the state-of-the art approaches to design a dedicated neural network architecture for a specific classification task. Involving oversized nodes and network layers in the model design, however, usually causes the over-smoothing problem and the lack of embedding discrimination. In this paper, we propose RoSGAS, a novel Reinforced and Self-supervised GNN Architecture Search framework to adaptively pinpoint the most suitable multi-hop neighborhood and the number of layers in the GNN architecture. More specifically, we consider the social bot detection problem as a user-centric subgraph embedding and classification task. We exploit heterogeneous information network to present the user connectivity by leveraging account metadata, relationships, behavioral features and content features. RoSGAS uses a multi-agent deep reinforcement learning (RL) mechanism for navigating the search of optimal neighborhood and network layers to learn individually the subgraph embedding for each target user. A nearest neighbor mechanism is developed for accelerating the RL training process, and RoSGAS can learn more discriminative subgraph embedding with the aid of self-supervised learning. Experiments on 5 Twitter datasets show that RoSGAS outperforms the state-of-the-art approaches in terms of accuracy, training efficiency and stability, and has better generalization when handling unseen samples.