This repository contains all outcomes created in the 2025 Eleuther AI Summer of Open AI Research program for project #5, Scientific Literature Knowledge Extraction Tool.

• UV (Python package installer and resolver)
• pre-commit (Git hooks manager)
• docker (to start local FalkorDB instance)

1. Create and activate a virtual environment using UV:

   uv venv
   source .venv/bin/activate  # On Unix/macOS
   # or
   .\.venv\Scripts\activate  # On Windows

2. Install development dependencies:

   uv sync

3. Install pre-commit hooks:

   uv run pre-commit install

Pre-commit hooks are configured to run on each commit to ensure code quality.

To manually run all pre-commit checks:

uv run pre-commit run --all-files

You can put extracted files into FalkorDB to visualize. Start an instance with

docker run -p 6379:6379 -p 3000:3000 -it --rm falkordb/falkordb:latest

then fill it out with

uv run src/db.py and merge with uv run src/merge.py

and you can see the result at http://localhost:3000/graph

• Automated Discovery of Emerging Research Fronts and "Invisible Colleges": identify clusters of papers that are conceptually related, even if they don't cite each other directly
• Knowledge Gap Detection and Research Opportunity Mapping: identify weakly connected or isolated concepts, then prioritize them as potential research frontiers
• Consensus / Controversy Atlas (claim-level meta-analysis): aggregate by concept and time to generate consensus scores and timelines

Also other projects that should already be possible with a citation graph:

• Cross-Disciplinary Concept Bridging Tool: take the most influential concepts in one subgraph and combine them with concepts from another subgraph (transformers in protein folding)
• Concept Evolution and Trend Forecasting