About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
AI/Data Informatics: Computational Model Development, Verification, Validation, and Uncertainty Quantification
|
Presentation Title |
Extracting Complex Concentrated Alloys Properties from Scientific Literature with LLMs |
Author(s) |
Joshua Berry, Alan Thomas, Xianyuan Liu, Haiping Lu, Nicola Morley, Katerina A. Christofidou |
On-Site Speaker (Planned) |
Joshua Berry |
Abstract Scope |
Materials informatics tools necessitate the availability and robustness of large databases that in the complex concentrated alloy (CCA) space, are unavailable. However, interrogation of the literature using natural language processing (NLP) and large language models (LLMs) offers a unique opportunity to create, collate and curate databases. Despite the wealth of CCA papers, the diverse contexts and complex data dissemination within these papers pose significant challenges to the application of NLP and LLMs. To understand the challenges and opportunities of such techniques for developing relevant databases for alloy design, this work presents the repeated construction using LLM of two manually collated published databases consisting of compositional information and material properties of CCAs. This work enables the construction of comprehensive datasets essential for advancing research in alloy development and materials discovery, facilitating the exploration of novel compositions.
This work was supported by Oerlikon AM Europe GmbH, EPSRC-UK [EP/S022635/1] and SFI [18/EPSRC-CDT/3584]. |
Proceedings Inclusion? |
Planned: |
Keywords |
Machine Learning, Computational Materials Science & Engineering, High-Entropy Alloys |