Chahal, P., Agarwal, U., Umra, M., Kumari, R., Barbhuiya, S., Naqvi, A. M., Simon, W., Yadav, G.& Murray-Rust, P. (2026). A Scientific-Literature Question-Answering Pipeline Using Retrieval-Augmented Generation (RAG) (Version 0.1) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.21803196
This notebook provides an end-to-end Retrieval-Augmented Generation (RAG) workflow for exploring and querying scientific literature. Starting with a user-defined research topic, it automatically retrieves relevant open-access publications, extracts and cleans their full text, and builds a searchable knowledge base that can be queried using a Large Language Model (LLM).
Unlike general-purpose chatbots, this RAG pipeline generates responses that are grounded in the retrieved scientific literature. Every answer is supported by citations, including the publication title and DOI (where available), enabling users to verify the source of each claim rather than relying solely on the model's pre-trained knowledge.
Developed as a teaching resource for the FSCI 2026 (FORCE11 Scholarly Communication Institute) programme, this notebook introduces researchers, librarians, and information professionals to modern LLM and RAG techniques through practical, hands-on exercises. It is designed to help users build literature-aware AI applications using their own research domains.
Software citation information: CITATION.cff
License: Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ | License information: LICENSE