This research bridges the gap between general-purpose LLMs and specialized materials science knowledge extraction by fine-tuning language models to identify process-structure-properties relationships.

• Adapts large language models to overcome limitations in materials-specific information extraction
• Focuses on extracting process-structure-properties relationships crucial for materials discovery
• Addresses the challenge of limited expert annotations in materials science through specialized fine-tuning
• Creates new pathways for leveraging millions of academic papers for materials innovation

This advancement significantly impacts engineering by enabling automated extraction of valuable materials data from vast scientific literature, accelerating discovery and application of new materials in industrial contexts.

Structured Extraction of Process Structure Properties Relationships in Materials Science