Libraries used in MyMediaLite
The MyMedia Project
- Dynamic Personalization of Multimedia (MyMedia)
- The MyMedia Framework is a framework for Multimedia recommendations (non-commercial use only).
- The MyMedia project was part of the User Centric Media Cluster.
Free/Open Source Recommender System Software
Frameworks:
- Apache Mahout (Java)
- GraphLab collaborative filtering library (C++)
- LensKit (Java)
- Waffles (C++)
- easyrec
- RecLab – test recommender system models on synthetic and live data
- Crab (Python)
- recommenderlab (R)
- Duine (inactive)
- MultiLens (inactive)
- Taste.NET (inactive)
Implementations of single methods:
- Likelike – LSH (locality sensitive hashing) on Hadoop
- CofiRank – Maximum Margin Matrix Factorization for Collaborative Ranking
- SVDFeature – Feature-based matrix factorization (for rating prediction)
- Jellyfish – Large-scale parallel matrix factorization (for rating prediction)
- wooflix – SVD++ implementation in Python
- PyRSVD – Regularized SVD solver in C/Python
- OpenSlopeOne – SlopeOne with PHP and MySQL
- Hypergraph Playlist – Playlist generation from hypergraphs (Python)
Specialized systems with recommendation functionality:
- AppRecommender – software packages
- Collection Workflow Integration System – knowledge management
Free/Open Source Machine Learning Software
- mloss.org: directory of free/open source machine learning software
- LIBSVM
- LIBLINEAR
- Vowpal Wabbit
- GraphLab
- Weka
- R
Recommender Systems in General
- RecSysWiki: wiki about recommender systems
- ACM Conference on Recommender Systems: series of conferences on recommender systems
Development Tools
- Mono (free and portable .NET implementation)
- MonoDevelop (IDE)
- NUnit unit testing framework
- Template Toolkit (used to create this website)
- Gitorious (version control hosting)
- github (version control hosting)
- Strawberry Perl (Perl for Windows boxes)