Large language models (LLMs) have the potential to enhance evidence synthesis efficiency and accuracy. This study assessed LLM-only and LLM-assisted methods in data extraction and risk of bias ...

Extracting data from scientific literature and constructing machine learning-ready (ML-ready) datasets is crucial for data-driven materials discovery. However, material properties under various ...