Knowledge Graph Question Answering for Materials Science (KGQA4MAT): Developing Natural Language Interface for Metal-Organic Frameworks Knowledge Graph (MOF-KG) Using LLM

TL;DR

This paper introduces a new benchmark dataset and a systematic approach using ChatGPT to facilitate natural language question answering over a materials science knowledge graph focused on metal-organic frameworks, aiming to improve accessibility for domain experts.

Contribution

It provides a comprehensive benchmark dataset for KGQA in materials science and demonstrates a novel method using LLMs to translate natural language questions into formal KG queries.

Findings

ChatGPT effectively translates complex natural language questions into KG queries.

The benchmark includes 161 complex questions with multiple variations.

The approach shows promise for improving user-friendly access to domain-specific knowledge graphs.

Abstract

We present a comprehensive benchmark dataset for Knowledge Graph Question Answering in Materials Science (KGQA4MAT), with a focus on metal-organic frameworks (MOFs). A knowledge graph for metal-organic frameworks (MOF-KG) has been constructed by integrating structured databases and knowledge extracted from the literature. To enhance MOF-KG accessibility for domain experts, we aim to develop a natural language interface for querying the knowledge graph. We have developed a benchmark comprised of 161 complex questions involving comparison, aggregation, and complicated graph structures. Each question is rephrased in three additional variations, resulting in 644 questions and 161 KG queries. To evaluate the benchmark, we have developed a systematic approach for utilizing the LLM, ChatGPT, to translate natural language questions into formal KG queries. We also apply the approach to the well-known QALD-9 dataset, demonstrating ChatGPT's potential in addressing KGQA issues for different platforms and query languages. The benchmark and the proposed approach aim to stimulate further research and development of user-friendly and efficient interfaces for querying domain-specific materials science knowledge graphs, thereby accelerating the discovery of novel materials.