Abstract
<h2>Abstract</h2> The Dual Training Error based Correction approach (DTEC) is applied following the execution of a recommender system to improve its recommendation accuracy. These corrections are performed in user and item view point, so that finally a dual system efficiently combines both of them. DTEC is applicable to any model-based recommender system with positive training error. An open-source Matlab implementation with no external dependencies of this algorithm is available.
Highlights
Recommender Systems (RS) try to predict the preferences of users for specific items, based on an analysis of previous consumer preferences [1]
Given a set U of users, a set I of items and a set R of ratings of users for items, the goal of a recommender system is to predict the rating for a user–item pair which is not in R
We provide an open-source MATLAB implementation of a parameter free system, that can be executed after any model-based recommender system in order to improve its recommendation accuracy
Summary
Recommender Systems (RS) try to predict the preferences of users for specific items, based on an analysis of previous consumer preferences [1]. We provide an open-source MATLAB implementation of a parameter free system, that can be executed after any model-based recommender system in order to improve its recommendation accuracy. This system is called Dual Training Error based Correction approach. The goal of the DTEC model is to improve the accuracy of a RS by adjusting its output based on a weighted average of the errors in the training set. In [6], the proposed method is successfully applied to SCoR [1] and to other state-of-the-art recommender systems, demonstrating the efficiency and high performance of DTEC since it potentially increases the accuracy of the recommendations. The wide applicability of the proposed DTEC method on any model-based recommender system makes this software quite useful
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
