A machine-learning based tool for extracting and analyzing scientific paper references. It parses the references in a bibliography (.bib) file and allows searching of referenced sentences contained in the introduction of any of the articles in the bibliography.
How it works:
- parses bibliography for article DOIs
- looks up the PMC identifiers for each of the DOIs
- parses the PMC html for Introduction/Main/Background section
- matches sentence and reference for each statement in the article's introdution section
- sentence fragments that are incomplete clauses, are reworded using Ollama to generate standalone sentences that best represent what is being stated in the article sentence.
- SentenceTransformers are used to compute the semantic embedding of each sentence which is compared to the user's search term by cosine similarity between the embeddings
FactYou uses a few machine learning libraries most of which can be installed with pip from the requirements.txt. The exception to this is Ollama which must be installed by the user before FactYou can be run. Installation insctructions for Ollama on desktop can be found here.
# Clone the repository
git clone https://github.com/seanlaidlaw/FactYou.git
cd FactYou
# Install dependencies
pip install -e .To launch the application run the module with python:
python -m factyu.mainThis uses a persistent database stored in your user data directory in which it stores the extracted information from the bibliography files (.bib) passed to it.
The application will listen on 127.0.0.1:5000 by default. If port is already in use, a different port can be manually set from the command line argument:
python -m factyu.main --host 0.0.0.0 --port 80The application stores data in a SQLite database. The default location is:
- Linux:
~/.local/share/FactYou/references.db - macOS:
~/Library/Application Support/FactYou/references.db - Windows:
C:\Users\<Username>\AppData\Local\FactYouApp\FactYou\references.db