Just another WordPress site - Ruhr-Universität Bochum

Reproducibility of computational results remains a challenge in materials science, as simulation workflows are often reported in unstructured text. While literature is valuable for validation and reuse, the lack of machine-readable workflow descriptions prevents large-scale curation and systematic comparison. Existing text-mining approaches typically extract entities or pairwise relationships but do not capture computational workflows. An ontology-driven, large language model (LLM)-assisted framework is introduced for the semi-automated extraction and structuring of computational workflows from the literature. The approach focuses on density functional theory-based stacking fault energy calculations in hexagonal close-packed magnesium and its alloys. It uses a filtering strategy together with prompt-engineered LLM extraction, resulting in a structured dataset with 711 data points. Extracted information is unified into a canonical schema and aligned with established materials ontologies, enabling the construction of a knowledge graph using atomRDF. While full computational reproducibility remains limited by missing or implicit metadata such as relaxation constrains, the primary purpose of the framework is to enable semantic organization, comparison of reported results, and reuse of computational workflows. By representing extracted information in an ontology-aligned, semantically interoperable form, the framework improves transparency and reusability of computational materials data, while establishing a foundation toward future automated reproduction.