LLM-Empowered Class Imbalanced Graph Prompt Learning for Online Drug Trafficking Detection

2025-02-28

AI Summary

Addressing the critical challenges of extreme class imbalance and scarce labeled samples in detecting illicit drug transactions on online platforms, this paper proposes LLM-HetGDT—a novel framework integrating Large Language Models (LLMs) with Heterogeneous Graph Neural Networks (HGNNs). To mitigate graph-level imbalance, it introduces the first LLM-driven synthetic minority-user-node generation method. It further designs a learnable soft-prompt mechanism tailored for heterogeneous graphs to enhance modeling of discriminative patterns specific to minority classes. Finally, it employs contrastive pre-training coupled with soft-prompt fine-tuning to improve generalization. Evaluated on our newly constructed Twitter-HetDrug dataset, LLM-HetGDT achieves a 12.6% absolute improvement in F1-score and a 19.3% gain in minority-class recall over state-of-the-art methods, demonstrating both effectiveness and practical applicability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

As the market for illicit drugs remains extremely profitable, major online platforms have become direct-to-consumer intermediaries for illicit drug trafficking participants. These online activities raise significant social concerns that require immediate actions. Existing approaches to combating this challenge are generally impractical, due to the imbalance of classes and scarcity of labeled samples in real-world applications. To this end, we propose a novel Large Language Model-empowered Heterogeneous Graph Prompt Learning framework for illicit Drug Trafficking detection, called LLM-HetGDT, that leverages LLM to facilitate heterogeneous graph neural networks (HGNNs) to effectively identify drug trafficking activities in the class-imbalanced scenarios. Specifically, we first pre-train HGNN over a contrastive pretext task to capture the inherent node and structure information over the unlabeled drug trafficking heterogeneous graph (HG). Afterward, we employ LLM to augment the HG by generating high-quality synthetic user nodes in minority classes. Then, we fine-tune the soft prompts on the augmented HG to capture the important information in the minority classes for the downstream drug trafficking detection task. To comprehensively study online illicit drug trafficking activities, we collect a new HG dataset over Twitter, called Twitter-HetDrug. Extensive experiments on this dataset demonstrate the effectiveness, efficiency, and applicability of LLM-HetGDT.

Problem

• Detects online drug trafficking using LLM-enhanced graph learning.
• Addresses class imbalance and scarce labeled data in detection.
• Improves detection with synthetic data and fine-tuned prompts.

Innovation

• Leverages LLM to enhance heterogeneous graph neural networks
• Generates synthetic user nodes for minority class augmentation
• Fine-tunes soft prompts on augmented graph for detection