LEC-KG: An LLM-Embedding Collaborative Framework for Domain-Specific Knowledge Graph Construction -- A Case Study on SDGs

TL;DR

LEC-KG is a collaborative framework that combines LLMs and knowledge graph embeddings to effectively construct domain-specific knowledge graphs from unstructured text, especially handling long-tail relations and unseen entities.

Contribution

It introduces a bidirectional, iterative approach integrating LLMs with KGE for improved knowledge graph construction from unstructured data.

Findings

Significant improvement over LLM baselines on SDG reports

Effective handling of low-frequency relations

Reliable transformation of unstructured text into validated triples

Abstract

Constructing domain-specific knowledge graphs from unstructured text remains challenging due to heterogeneous entity mentions, long-tail relation distributions, and the absence of standardized schemas. We present LEC-KG, a bidirectional collaborative framework that integrates the semantic understanding of Large Language Models (LLMs) with the structural reasoning of Knowledge Graph Embeddings (KGE). Our approach features three key components: (1) hierarchical coarse-to-fine relation extraction that mitigates long-tail bias, (2) evidence-guided Chain-of-Thought feedback that grounds structural suggestions in source text, and (3) semantic initialization that enables structural validation for unseen entities. The two modules enhance each other iteratively-KGE provides structure-aware feedback to refine LLM extractions, while validated triples progressively improve KGE representations. We evaluate LEC-KG on Chinese Sustainable Development Goal (SDG) reports, demonstrating substantial improvements over LLM baselines, particularly on low-frequency relations. Through iterative refinement, our framework reliably transforms unstructured policy text into validated knowledge graph triples.