LINK-KG: LLM-Driven Coreference-Resolved Knowledge Graphs for Human Smuggling Networks

2025-10-30

AI Summary

Human trafficking network analysis is hindered by lengthy legal documents, ambiguous coreference, and lack of structural annotation, resulting in fragmented knowledge graph construction and inconsistent entity linking. To address these challenges, this paper proposes a large language model–driven, three-stage coreference resolution framework. It introduces a novel type-aware Prompt Cache mechanism to ensure cross-textual consistency in entity referencing, and integrates modular knowledge extraction to substantially suppress node redundancy and noise. Experimental results demonstrate that, compared to baseline methods, the proposed approach reduces node duplication by 45.21% and noise nodes by 32.22%. The resulting knowledge graph exhibits significant improvements in coherence, completeness, and interpretability for criminal network analysis.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Knowledge Representation Languages
• Machine Learning: Graph-based Machine Learning

Application Category

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

Abstract

Human smuggling networks are complex and constantly evolving, making them difficult to analyze comprehensively. Legal case documents offer rich factual and procedural insights into these networks but are often long, unstructured, and filled with ambiguous or shifting references, posing significant challenges for automated knowledge graph (KG) construction. Existing methods either overlook coreference resolution or fail to scale beyond short text spans, leading to fragmented graphs and inconsistent entity linking. We propose LINK-KG, a modular framework that integrates a three-stage, LLM-guided coreference resolution pipeline with downstream KG extraction. At the core of our approach is a type-specific Prompt Cache, which consistently tracks and resolves references across document chunks, enabling clean and disambiguated narratives for structured knowledge graph construction from both short and long legal texts. LINK-KG reduces average node duplication by 45.21% and noisy nodes by 32.22% compared to baseline methods, resulting in cleaner and more coherent graph structures. These improvements establish LINK-KG as a strong foundation for analyzing complex criminal networks.

Problem

• Resolving coreference ambiguities in legal case documents for smuggling networks
• Constructing consistent knowledge graphs from long unstructured legal texts
• Reducing entity duplication and noise in automated knowledge graph extraction

Innovation

• LLM-guided coreference resolution pipeline for knowledge graphs
• Type-specific Prompt Cache tracks references across documents
• Modular framework reduces node duplication and noise