Researchllmgraph neural networkscitation analysisknowledge distillation
LAGMiD Detects Miscitation in Citation Networks
8.1
Huidong Wu et al. (Mar 10, 2026) introduce LAGMiD, an LLM-augmented graph learning framework for detecting miscitation in scholarly citation networks. LAGMiD uses evidence-chain chain-of-thought multi-hop citation tracing, distills intermediate LLM reasoning into GNN embeddings, and routes hard cases to the LLM; experiments on three real-world benchmarks report state-of-the-art detection with substantially reduced inference cost.
Key Points
- 1Introduces LAGMiD, combining LLM chain-of-thought reasoning with graph neural networks for miscitation detection
- 2Applies evidence-chain multi-hop citation tracing to assess semantic fidelity across citation graph contexts, improving detection accuracy
- 3Distills LLM intermediate reasoning into GNN embeddings and routes complex cases to LLMs, enabling scalable lower-cost deployment
Scoring Rationale
Strong novel method combining LLM reasoning and GNN distillation, but remains an arXiv preprint without peer review.
Sources
Public references used for this report.
Practice with real Logistics & Shipping data
90 SQL & Python problems · 15 industry datasets
Used by DS/ML engineers at top companies
High-Value Overnight OrdersEasyDelivered International ShipmentsMediumOn-Time Delivery Rate by CarrierHard
250 free problems · No credit card
See all Logistics & Shipping problems