https://github.com/semanticclimate/rag-llm-with-pdf-xml

https://github.com/semanticclimate/rag-llm-with-pdf-xml

Last synced: 9 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/semanticclimate/rag-llm-with-pdf-xml
• Owner: semanticClimate
• License: apache-2.0
• Created: 2025-07-29T08:41:20.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2025-08-01T09:40:54.000Z (11 months ago)
• Last Synced: 2025-09-04T18:59:38.677Z (10 months ago)
• Language: Jupyter Notebook
• Size: 85 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE
  • Citation: CITATION.cff

Awesome Lists containing this project

README

          
# RAG-LLM Pipeline for Extracting and Generating Insights from PDF/XML File



  



DOI Zenodo badge: 

[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.16675979.svg)](https://doi.org/10.5281/zenodo.16675979)

Citation:

Barbhuiya, S., Alwi, K. K., Kumari, R., S., A., Jawed, M., Simon, W., Yadav, G., & Murray-Rust, P. (2025). RAG-LLM Pipeline for Extracting and Generating Insights from PDF/XML File (0.2). Zenodo. https://doi.org/10.5281/zenodo.16675979

Description: 

This notebook demonstrates how to build a semantic question-answering system over scientific PDFs using Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs). It enables users to upload PDFs, extract content, embed it into a vector store, and query the document using natural language.

**Key Features**

- PDF Upload & Text Extraction: Extract raw text from research papers using PyMuPDF

- Text Chunking & Embeddings: Convert text into meaningful chunks and generate embeddings using models like sentence-transformers

- RAG Pipeline:

    - Store document chunks in a FAISS vector database

    - Retrieve top-matching chunks based on user queries

    - Generate context-aware answers with an LLM

    - Natural Language Q&A: Ask questions like “What is the main finding?” or “What methods were used?” and get accurate answers drawn directly from the paper

[Link to Notebook](https://colab.research.google.com/drive/17J9wEvkQvdaeOihN3N13u_ln5Oez8ssd?usp=sharing)

Reviewers & review process: \ 

---

Software citation information: [CITATION.cff](CITATION.cff)

License: Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ | License information: [LICENSE](LICENSE)