Release Notes: Important changes: fold-in support for UserKNN for item recommendation; less verbose evaluation output; and many bugfixes.
Release Notes: The release added NaiveBayes for attribute-based rating prediction, MostPopularByItemAttributes, and improvements and bugfixes in WeightedBPRMF, SigmoidCombinedAsymmetricFactorModel, and ItemKNN. A bug in the Map@k routine was fixed. Support for Million Song Dataset Challenge data was improved.
Release Notes: This release adds several new rating predictors, all of them variants of asymmetric factor models (AFMs). The new item recommender MostPopularByItemAttributes generalizes an idea presented by the organizers of the Million Song Database Challenge. There are now 27 different rating predictors and 18 different item recommenders in MyMediaLite.
Release Notes: This release adds the new rating predictor SigmoidItemAsymmetricFactorModel, improvements and bugfixes for SocialMF, speed-ups for ItemKNN rating predictors, faster writing of item recommendations, and fast (de)serialization of data sets.
Release Notes: Float (32-bit) is now used instead of double (64-bit) to store ratings and model parameters. The incremental update API now accepts several feedback events at once. A new SVD++ rating predictor was added. LogisticRegressionMatrixFactorization and MultiCoreMatrixFactorization were merged into BiasedMatrixFactorization. There were many small enhancements and fixes, and polishing.
Release Notes: Similarity computations are now faster and consume less memory. This release adds the new rating prediction evaluation criterion CBD (capped binomial deviance), new recommenders (MultiCoreBPRMF and LogisticRegressionMatrixFactorization), and bugfixes and other improvements for the recommenders BPRMF, MultiCoreMatrixFactorization, TimeAwareBaseline, UserItemBaseline, and ItemKNNCosine.
Release Notes: Can now be built without an IDE. Command line tools: scripts for easy deployment on Unix; executable names changed to lower case; an option to ignore the first line of a file. New examples in F# and simplified examples in C#, Python, and Ruby. Evaluation methods are much easier to call. There is a new baseline rating predictor: co-clustering. There are bugfixes and other improvements for BPRMF, MultiCoreMatrixFactorization, TimeAwareBaseline, and KNN.
Release Notes: A crash in the item recommendation tool has been fixed.
Release Notes: Rating data with timestamps is now supported. There are two time-aware baselineline recommenders, TimeAwareBaseline and TimeAwareBaselineWithFrequencies (from the Netflix Grand Prize solution), and chronological splits. Some matrix functions are now implemented as extension methods, which makes them easier to use in many cases. The script download_movielens.sh has been updated because the download URLs have changed. Lots of improvements have been made under the hood.
Release Notes: MyMediaLite now needs Mono 2.8 or later (2.10.x is recommended, or .NET 4.0). Multi-core support has been added for several components: item recommendation evaluation, cross-validation for rating prediction and item recommendation, and distributed SGD training for rating prediction matrix factorization. Long int user and item IDs are now supported. Different modes for selecting candidate items now also work for cross-validation. Mean reciprocal rank (MRR) has been added as a new evaluation measure. Numerous improvements have been made to the documentation, API, commandline tools, and helper scripts.