Problem
Research questions and friction points this paper is trying to address.
Knowledge Graph
Machine Learning Interatomic Potential
Materials Science
Automatic Construction
Automatically Building and Updating a Knowledge Graph of MLIP Models
🤖 AI Summary
This study addresses the fragmentation of domain knowledge and the difficulty of automatically constructing structured knowledge graphs in the field of machine learning interatomic potentials (MLIPs). To overcome these challenges, this work proposes a closed-loop framework driven by large language models (LLMs) that integrates multi-step information extraction with SHACL shape constraint validation. The proposed method leverages LLMs to automatically parse scientific literature and employs SHACL validation to enable incremental, iterative updates of the knowledge graph while ensuring model-level representational consistency. Applying this approach, we successfully constructed an MLIP knowledge graph encompassing models from the Matbench Discovery leaderboard. The resulting graph supports complex semantic queries and significantly enhances both knowledge management efficiency and the level of automation within the MLIP research community.
📝 Abstract
Complementing the many efforts in providing semantic representations of concepts, notions, and entities in materials science, we report and illustrate a process by which we can automatically build a knowledge graph of the fast evolving field of machine learning applied to the prediction of material properties, focusing on MLIP (Machine Learning Interatomic Potential). This LLM-based process relies on multiple steps, from information extraction in documents and articles to a validation loop using SHACL constraints to detect and correct errors. It is carried out on a model-by-model basis, focusing on the consistency of representation, therefore enabling an iterative construction where the addition of new models is facilitated. We illustrate the process by showing a few interesting aspects that can be queried from a knowledge graph built from the models listed in the Matbench Discovery leaderboard.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Knowledge Graph
Large Language Model
Machine Learning Interatomic Potential
SHACL Constraints
Information Extraction
🔎 Similar Papers
No similar papers found.
