🤖 AI Summary
Large language models (LLMs) suffer from frequent hallucinations, poor traceability, and outdated knowledge in knowledge-intensive domains such as healthcare. Method: This paper proposes Tripartite-GraphRAG—a novel retrieval-augmented generation framework built upon the first tripartite knowledge graph jointly modeling domain entities, ontology concepts, and raw text snippets. It employs concept anchoring for fidelity-preserving pre-analysis and formalizes prompt generation as an unsupervised node classification task—bypassing bottlenecks in traditional entity resolution and deduplication. The method integrates lexicon-guided graph construction, Markov random field–driven prompt generation, and statistics-based concept-level embedding similarity scoring. Results: On medical history analysis tasks, Tripartite-GraphRAG significantly improves prompt information density, coverage, and structural coherence; reduces prompt length and inference overhead; and enhances the consistency and reliability of LLM outputs.
📝 Abstract
Large Language Models (LLMs) have shown remarkable capabilities across various domains, yet they struggle with knowledge-intensive tasks in areas that demand factual accuracy, e.g. industrial automation and healthcare. Key limitations include their tendency to hallucinate, lack of source traceability (provenance), and challenges in timely knowledge updates. Combining language models with knowledge graphs (GraphRAG) offers promising avenues for overcoming these deficits. However, a major challenge lies in creating such a knowledge graph in the first place. Here, we propose a novel approach that combines LLMs with a tripartite knowledge graph representation, which is constructed by connecting complex, domain-specific objects via a curated ontology of corresponding, domain-specific concepts to relevant sections within chunks of text through a concept-anchored pre-analysis of source documents starting from an initial lexical graph. As a consequence, our Tripartite-GraphRAG approach implements: i) a concept-specific, information-preserving pre-compression of textual chunks; ii) allows for the formation of a concept-specific relevance estimation of embedding similarities grounded in statistics; and iii) avoids common challenges w.r.t. continuous extendability, such as the need for entity resolution and deduplication. By applying a transformation to the knowledge graph, we formulate LLM prompt creation as an unsupervised node classification problem, drawing on ideas from Markov Random Fields. We evaluate our approach on a healthcare use case, involving multi-faceted analyses of patient anamneses given a set of medical concepts as well as clinical literature. Experiments indicate that it can optimize information density, coverage, and arrangement of LLM prompts while reducing their lengths, which may lead to reduced costs and more consistent and reliable LLM outputs.