NReco Recommender is based on complete .NET port of Apache Mahout "Taste" - a flexible, fast collaborative filtering engine for Java. Recommendation mining takes users' behavior and from that tries to find items users might like. It can bring real-life intelligence to the almost any application!

Includes:
  • User-based and Item-based recommendation algorithms
  • A lot of real-life similarity functions
  • Evaluation framework for choosing best algorithm / parameters
  • SVD-based recommender

typical usage scenario

  1. Decide what you want to suggest (this may be anything: books, movies, users, profiles, tasks, articles)
  2. Determine how to infer preferences dataset from existing DB or users behaviour
  3. Evaluate and choose appropriate recommendation algorithm (NReco.Recommender can help)
  4. Integrate recommendations into your application (NReco.Recommender includes examples)

Links

Last edited Oct 28, 2016 at 12:44 PM by fedorchenko, version 8