A Matrix Factorization Algorithm for Efficient Recommendations in Social Rating Networks Using Constrained Optimization

Nicholas Ampazis,Theodoros Emmanouilidis,Flora Sakketou

doi:10.3390/make1030053

Nicholas Ampazis, Theodoros Emmanouilidis + Show 1 more

Open Access

https://doi.org/10.3390/make1030053

Copy DOI

Abstract

In recent years the emergence of social media has become more prominent than ever. Social networking has become the de facto tool used by people all around the world for information discovery. Consequently, the importance of recommendations in a social network setting has urgently emerged, but unfortunately, many methods that have been proposed in order to provide recommendations in social networks cannot produce scalable solutions, and in many cases are complex and difficult to replicate unless the source code of their implementation has been made publicly available. However, as the user base of social networks continues to grow, the demand for developing more efficient social network-based recommendation approaches will continue to grow as well. In this paper, following proven optimization techniques from the domain of machine learning with constrained optimization, and modifying them accordingly in order to take into account the social network information, we propose a matrix factorization algorithm that improves on previously proposed related approaches in terms of convergence speed, recommendation accuracy and performance on cold start users. The proposed algorithm can be implemented easily, and thus used more frequently in social recommendation setups. Our claims are validated by experiments on two real life data sets, the public domain Epinions.com dataset and a much larger dataset crawled from Flixster.com.

Highlights

Matrix factorization in collaborative filtering recommender systems is usually performed by unconstrained gradient descent for learning the feature components of the user and item factor matrices [1]
In this paper, following proven optimization techniques from the domain of machine learning with constrained optimization, and modifying them in order to take into account the social network information, we propose a matrix factorization algorithm that improves on previously proposed related approaches in terms of convergence speed, recommendation accuracy and performance on cold start users
We proposed an efficient constrained matrix factorization algorithm called SocialFALCON, for providing recommendations in social rating networks

Summary

Introduction

Matrix factorization in collaborative filtering recommender systems is usually performed by unconstrained gradient descent for learning the feature components of the user and item factor matrices [1] This is essentially a “black box” approach, where apart from the minimization of an objective function (usually the Root Mean Squared Error (RMSE) over the known ratings), generally no other information or knowledge is taken into account during the factorization process. We exploit this information within the FALCON framework and propose a matrix factorization algorithm for recommendation in social rating networks, called SocialFALCON.

Matrix Factorization for Recommender Systems

SocialMF

Overview of the Falcon Framework

The SocialFALCON Algorithm

Desirable Properties of SocialFALCON

Complexity Analysis of SocialFALCON

Datasets

Experimental Results

Conclusions