A New Contextual Influencer User Measure to Improve the Accuracy of Recommender System

A New Contextual Influencer User Measure to Improve the Accuracy of Recommender System

Maryam Jallouli (Miracl Laboratory, Technopole of Sfax, Sfax University, Tunisia), Sonia Lajmi (Miracl Laboratory, Technopole of Sfax, Sfax University, Tunisia & Al Baha University, Al Baha, Saudi Arabia) and Ikram Amous (Miracl Laboratory, Technopole of Sfax, Sfax University, Tunisia)
DOI: 10.4018/IJSITA.2018100103

Abstract

In the last decade, social-based recommender systems have become the best way to resolve a user's cold start problem. In fact, it enriches the user's model by adding additional information provided from his social network. Most of those approaches are based on a collaborative filtering and compute similarities between the users. The authors' preliminary objective in this work is to propose an innovative context aware metric between users (called contextual influencer user). These new similarities are called C-COS, C-PCC and C-MSD, where C refers to the category. The contextual influencer user model is integrated into a social based recommendation system. The category of the items is considered as the most pertinent context element. The authors' proposal is implemented and tested within the food dataset. The experimentation proved that the contextual influencer user measure achieves 0.873, 0.874, and 0.882 in terms of Mean Absolute Error (MAE) corresponding to C-cos, C-pcc and C-msd, respectively. The experimental results showed that their model outperforms several existing methods.
Article Preview
Top

In the literature, and according to (Gediminas, Adomavicius, & Tuzhilin, 2005), RS can be classified into three classes: content-based (Sánchez, 2013), collaborative filtering (Gediminas, Adomavicius, & Tuzhilin, 2005), (Francesco Ricci & Kantor, 2011) and hybrid approaches (Burke, 2002). In collaborative filtering, Matrix Factorization (MF) (Salakhutdinov & Mnih, 2008) is considered as a good method of predicting the missing ratings. According to this method, the data are organized in a “User x Item” matrix, as illustrated by Figure 1.

Figure 1.

A structure of a rating matrix

IJSITA.2018100103.f01

In Figure 1, the rows represent the users “IJSITA.2018100103.m01, ... IJSITA.2018100103.m02” and the columns constitute the items “IJSITA.2018100103.m03, ... IJSITA.2018100103.m04”. This matrix is called rating matrix, denoted by IJSITA.2018100103.m05, which m and n refers to the number of users and items respectively. Each row of this matrix corresponds to a specified user u, and each column corresponds to a specified item j while the intersection of a row and a column corresponds to a rating value IJSITA.2018100103.m06 of a user’s rating u given to an item j. As we can notice, the matrix R is sparse (more than 99% of the entries are missing), and the aim of a recommender system is to predict the missing entries. Once a recommendation approach is applied and a missing rating of a user u about an item j is calculated, we are talking about a predicted rating IJSITA.2018100103.m07.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 11: 4 Issues (2020): Forthcoming, Available for Pre-Order
Volume 10: 4 Issues (2019): 3 Released, 1 Forthcoming
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing