Skip to content

Getting Started

BEAR is built for easy institutional deployment. Use this guide to quickly set up a proof-of-concept for semantic search and expert discovery at your university.

Prerequisites

  • Git
  • Docker and Docker Compose

Installation

1. Clone the Repository

git clone https://github.com/uw-madison-dsi/bear.git
cd bear

2. Install Dependencies

BEAR uses uv for dependency management:

curl -LsSf https://astral.sh/uv/install.sh | sh
uv sync

3. Configuration

Copy the example environment file and configure your institutional settings:

cp example.env .env

Edit the .env file with your specific configuration. Key settings include:

  • Institution identifier (OpenAlex format)
  • Embedding model preferences
  • Custom data source configurations
  • API keys for external services

See the Config Reference for detailed configuration options.

4. Start Backend

docker compose up -d

This will start:

5. Crawl Academic Data

Crawl data from OpenAlex for your institution:

uv run bear/crawler.py <your-institution-name>

For example, for University of Wisconsin-Madison:

uv run bear/crawler.py uw-madison

6. Ingest Data

Process and vectorize the crawled data:

# Test run first
uv run bear/ingest.py --test

# Full ingest
uv run bear/ingest.py

The API will be available at http://localhost:8000.

Testing the Installation

Test your installation with a sample API call:

curl "http://localhost:8000/search_author?query=data%20science"

Next Steps

  • Explore the API Usage for hands-on examples