RecSysMF

Matrix Factorization for Recommender System with Parallel Stochastic Gradient Descent (with locks).

If you use RecSysMF please cite the following paper:

• F. Petroni and L. Querzoni: "GASGD: stochastic gradient descent for distributed asynchronous matrix completion via graph partitioning." In: Proceedings of the 8th ACM Conference on Recommender systems (RecSys), 2014.

###Usage:

To run the project from the command line, go to the dist folder and type the following:

java -Xmx§GB -jar RecSysMF.jar inputtrainfile [options]

Parameters:

• §: number of GB for the java virtual machine
• inputfile: the name of the file that stores the <user,item,rating> triples for train.

Options:

• -separator string -> specifies the separator between user, item and rating in the input file . Default '\t'.
• -test string -> the name of the file that stores the <user,item,rating> triples for test.
• -output string -> specifies the name of the file where the output (the loss) will be stored.
• -iterations integer -> specifies how many iterations to be performed by the sgd algorithm. Default 30.
• -lambda double -> specifies the regularization parameter for the sgd algorithm. Default 0.05.
• -learning_rate double -> specifies the learning rate for the sgd algorithm. Default 0.01.
• -rank integer -> specifies the number of latent features for the low rank approximation. Default 50.

Example

download the MovieLens 100k dataset from http://grouplens.org/datasets/movielens/ and put the folder ml-100k in the root of the project.

java -jar dist/RecSysMF.jar ml-100k/u1.base -test ml-100k/u1.test -output ml-100k/output_loss.dat -iterations 100 -lambda 0.01 -learning_rate 0.05 -rank 100