Automated, LLM Enabled Extraction of Synthesis Details for Reticular Materials from Scientific Literature
Abstract
Automated knowledge extraction from scientific literature can potentially accelerate materials discovery. We have investigated an approach for extracting synthesis protocols for reticular materials from scientific literature using large language models (LLMs). To that end, we introduce a Knowledge Extraction Pipeline (KEP) that automatizes LLM-assisted paragraph classification and information extraction. By applying prompt engineering with in-context learning (ICL) to a set of open- source LLMs, we demonstrate that LLMs can retrieve chemical information from PDF documents, without the need for fine-tuning or training and at a reduced risk of hallucination. By comparing the performance of five open-source families of LLMs in both paragraph classification and information extraction tasks, we observe excellent model performance even if only few example paragraphs are included in the ICL prompts. The results show the potential of the KEP approach for reducing human annotations and data curation efforts in automated scientific knowledge extraction.
Cite
Text
da Silva et al. "Automated, LLM Enabled Extraction of Synthesis Details for Reticular Materials from Scientific Literature." NeurIPS 2024 Workshops: AI4Mat, 2024.Markdown
[da Silva et al. "Automated, LLM Enabled Extraction of Synthesis Details for Reticular Materials from Scientific Literature." NeurIPS 2024 Workshops: AI4Mat, 2024.](https://mlanthology.org/neuripsw/2024/dasilva2024neuripsw-automated/)BibTeX
@inproceedings{dasilva2024neuripsw-automated,
title = {{Automated, LLM Enabled Extraction of Synthesis Details for Reticular Materials from Scientific Literature}},
author = {da Silva, Viviane Torres and Rademaker, Alexandre and Lionti, Krystelle and Giro, Ronaldo and Lima, Geisa and Fiorini, Sandro Rama and Archanjo, Marcelo and Carvalho, Breno W S R and Ferreira, Rodrigo Neumann Barros and Souza, Anaximandro and de Souza, João Pedro Gandarela and de Valnisio, Gabriela and Paz, Carmen and Cerqueira, Renato and Steiner, Mathias B},
booktitle = {NeurIPS 2024 Workshops: AI4Mat},
year = {2024},
url = {https://mlanthology.org/neuripsw/2024/dasilva2024neuripsw-automated/}
}