Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
October 20, 2025Open Access

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

View Full Paper
Ask AI
Bookmark
Share

Authors

TMTianyi MaYQYefei QianZWZehong Wang

Discussion

Loading...

Member takes

Overview

Novel framework improves drug trafficking detection in online platforms, highlighting the challenges of class imbalance and synthetic data generation.

Key Points

  • The proposed method boosts drug trafficking detection effectiveness in scenarios with class imbalance.
  • Extensive experiments on the Twitter-HetDrug dataset illustrate that LLM-HetGDT can successfully identify illicit drug trafficking activities.
  • The framework incorporates heterogeneous graph neural networks, capitalizing on synthetic nodes to aid detection in minority classes.
  • The approach includes a contrastive pretext task to leverage unmatched graph information prior to the actual detection task.

Cite This Study

Ma et al. (2025) studied this question.

synapsesocial.com/papers/68f5fcd68d54a28a75cf2014https://doi.org/10.48550/arxiv.2503.01900
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1LLM-Enhanced User-Item Interactions: Leveraging Edge Information for Optimized Recommendations2024 · 5 citations
  2. 2Large Language Models for Link Stealing Attacks Against Graph Neural Networks2024 · 1 citations
  3. 3LOGIN: A Large Language Model Consulted Graph Neural Network Training Framework2024 · 1 citations
  4. 4LLMRec: Large Language Models with Graph Augmentation for Recommendation2024 · 95 citations
  5. 5Deep Learning Solution for Monitoring Social Media Drug Sales2026