Briefly - Legal Precedent Retrieval System

This is a legal case search engine that allows you to find relevant case precedents using keywords (BM25).

How to Run the App

1. Install Dependencies

Make sure you have Python 3.9+ installed.

pip install -r requirements.txt

2. Prepare Data & Build Indices

You need to run these scripts once to download data and build the search indices.

# 1. Download and process the legal dataset
python briefly/precedent_data_extraction.py

# 2. Build the BM25 search index
python briefly/bm25_pipeline.py

# 3. Build Dense/Hybrid search embeddings (takes a few minutes)
python briefly/build_dense_embeddings.py

# 4. Build the Citation Graph
python briefly/build_citation_graph.py

3. Run the Search App

Start the web interface:

streamlit run app.py

The app will open in your browser at http://localhost:8501.

Evaluation

To check the performance (MAP/Recall) of the system:

python evaluate_retrieval.py