BotHash: Efficient and Training-Free Bot Detection Through Approximate Nearest Neighbor

2025-06-25

AI Summary

To address the growing sophistication of large language model (LLM)-powered social bots and the poor generalizability and delayed response of conventional detection methods—largely stemming from their reliance on labeled data and supervised training—this paper proposes a training-free, lightweight, and efficient social bot detection framework. Our method maps user behavioral sequences into compact hash-based representations and identifies anomalous behavior clusters via approximate nearest neighbor (ANN) search. Notably, this is the first work to systematically integrate unsupervised ANN techniques into social bot detection, enabling zero-training overhead, minimal label dependency, and early-stage detection capability. Extensive evaluation across multiple real-world datasets demonstrates strong robustness against LLM-generated content and consistently superior performance: our approach achieves significantly higher accuracy and F1-score compared to state-of-the-art baselines.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Online Social Networks (OSNs) are a cornerstone in modern society, serving as platforms for diverse content consumption by millions of users each day. However, the challenge of ensuring the accuracy of information shared on these platforms remains significant, especially with the widespread dissemination of disinformation. Social bots -- automated accounts designed to mimic human behavior, frequently spreading misinformation -- represent one of the critical problems of OSNs. The advent of Large Language Models (LLMs) has further complicated bot behaviors, making detection increasingly difficult. This paper presents BotHash, an innovative, training-free approach to social bot detection. BotHash leverages a simplified user representation that enables approximate nearest-neighbor search to detect bots, avoiding the complexities of Deep-Learning model training and large dataset creation. We demonstrate that BotHash effectively differentiates between human and bot accounts, even when state-of-the-art LLMs are employed to generate posts' content. BotHash offers several advantages over existing methods, including its independence from a training phase, robust performance with minimal ground-truth data, and early detection capabilities, showing promising results across various datasets.

Problem

• Detects social bots mimicking human behavior without training
• Addresses misinformation spread by automated accounts in OSNs
• Overcomes challenges posed by LLMs in bot detection

Innovation

• Training-free bot detection using nearest-neighbor search
• Simplified user representation for efficient bot identification
• Effective against LLM-generated bot content