CypherBench: Towards Precise Retrieval over Full-scale Modern Knowledge Graphs in the LLM Era

2024-12-24

AI Summary

Large-scale modern knowledge graphs (e.g., Wikidata) pose significant accuracy and efficiency bottlenecks for LLM-based structured retrieval due to opaque identifiers, heterogeneous relations, and massive scale. Method: We propose a semantics-preserving RDF-to-property-graph transformation paradigm, enabling efficient Cypher-native graph querying. Contribution/Results: We introduce CypherBench—the first Cypher-centric benchmark covering 11 domains, 7.8 million entities, and over 10,000 questions—and design a composite evaluation framework grounded in semantic equivalence and execution accuracy. Experiments demonstrate substantial improvements in recall and Cypher generation fidelity for complex graph queries, establishing a scalable, production-ready knowledge retrieval infrastructure for GraphRAG and related systems.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Retrieval from graph data is crucial for augmenting large language models (LLM) with both open-domain knowledge and private enterprise data, and it is also a key component in the recent GraphRAG system (edge et al., 2024). Despite decades of research on knowledge graphs and knowledge base question answering, leading LLM frameworks (e.g. Langchain and LlamaIndex) have only minimal support for retrieval from modern encyclopedic knowledge graphs like Wikidata. In this paper, we analyze the root cause and suggest that modern RDF knowledge graphs (e.g. Wikidata, Freebase) are less efficient for LLMs due to overly large schemas that far exceed the typical LLM context window, use of resource identifiers, overlapping relation types and lack of normalization. As a solution, we propose property graph views on top of the underlying RDF graph that can be efficiently queried by LLMs using Cypher. We instantiated this idea on Wikidata and introduced CypherBench, the first benchmark with 11 large-scale, multi-domain property graphs with 7.8 million entities and over 10,000 questions. To achieve this, we tackled several key challenges, including developing an RDF-to-property graph conversion engine, creating a systematic pipeline for text-to-Cypher task generation, and designing new evaluation metrics.

Problem

• Large Language Models
• Knowledge Graph Retrieval
• Information Accessibility

Innovation

• CypherBench
• Attribute Graphs
• Cypher Query Language