MyMediaLite is a lightweight, open source and multi-purpose library of recommender system algorithms.
MyMediaLite addresses the two most common scenarios in collaborative filtering:
· rating prediction (e.g. on a scale of 1 to 5 stars), and
· item prediction from implicit feedback (e.g. from clicks or purchase actions).
Here are some key features of "MyMediaLite":
MyMediaLite gives you a choice of many recommendation methods:
· dozens of different recommenders
· methods can use collaborative and attribute/content data
MyMediaLite is ready to use:
· MyMediaLite includes evaluation routines for rating prediction and item prediction; it can measure MAE, RMSE, AUC, prec@N, MAP, and NDCG.
· It also comes with command line tools for both tasks that read a simple text-based input format.
· MyMediaLite is compact: The core library has a size of around 100KB.
· Portability: Written in C#, for the .NET platform; runs on every architecture supported by Mono: Mac OS X, Linux, Windows.
· Freedom: MyMediaLite is free software/open source software. It can be used, modified, and distributed under the terms of the GNU General Public License (GPL).
Additional features:
· Serialization: save and reload recommender models
· Real-time incremental updates for many recommenders
· Attribute-based diversification of recommendation lists