MyMediaLite is a lightweight, multi-purpose library of recommender system algorithms.
It 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 positive-only feedback (e.g. from clicks or purchase actions).
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 [% INCLUDE link link='mae' %], [% INCLUDE link link='rmse' %], [% INCLUDE link link='cbd' %], [% INCLUDE link link='auc' %], prec@N, MAP, and [% INCLUDE link link='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 150KB.
Portability: Written in C#, for the .NET platform; runs on every architecture supported by [% INCLUDE link link='mono' %]: Linux, Windows, Mac OS X.
Freedom: MyMediaLite is [% INCLUDE link link='free_software' %]/[% INCLUDE link link='open_source' %]. It can be used, modified, and distributed under the terms of the [% INCLUDE link link='gpl' %].
Additional features:
- Serialization: save and reload recommender models
- Real-time incremental updates for many recommenders
- Attribute-based diversification of recommendation lists