Discover Gorse: A Go-Powered Open-Source Recommendation Engine

Gorse is an open‑source recommendation system written in Go, designed to be a general‑purpose engine that can be quickly integrated into online services.

Key features include:

Automation: backend model search automatically selects the best recommendation model and strategy.

Distributed recommendation: single‑node training, distributed prediction, and horizontal scaling during the recommendation phase.

RESTful API: provides CRUD operations and recommendation requests via a RESTful interface.

Dashboard: data import/export, monitoring, and cluster status checking.

Gorse stores data in MySQL or MongoDB, caches intermediate data in Redis, and consists of a master node, multiple worker nodes, and server nodes. The master handles model training, non‑personalized item recommendation, configuration, and user management. Server nodes expose the RESTful API and deliver real‑time recommendations, while workers generate offline recommendations for each user.

The main workflow is:

User feedback is collected into the data store.

Archived feedback is extracted to train ranking and CTR models.

Offline recommendations are generated for all items and cached.

Online recommendations are served in real time from the cached offline results.

As of now, Gorse has received 3.8K stars and 274 forks on GitHub.

GitHub repository: https://github.com/zhenghaoz/gorse