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:

• API service: http://localhost:8000
• attu GUI for Milvus: http://localhost:3000
• Milvus vector database:
• Endpoint: http://localhost:19530
• Diagnostic Web-UI: http://localhost:9091/webui/
• MinIO (internal service)
• etcd (internal service)

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