Digital Library

cab1

 
Title:      EXAMINING FACTORS EFFECT TO DETERMINE ACCURACY OF PAGES RECOMMENDATION IN SOCIAL NETWORKS
Author(s):      Hamed Jafarpour, Ahmad A. Kardan
ISBN:      978-989-8533-39-5
Editors:      Ajith P. Abraham, Antonio Palma dos Reis and Jörg Roth
Year:      2015
Edition:      Single
Keywords:      Hybrid recommender system, recommender system, page recommendation, social network
Type:      Full Paper
First Page:      45
Last Page:      56
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      Social networks are popular platforms for interaction and collaboration among users. Users share their information and obtain others’ information shared in the networks. Generally, the information is published as pages. Hence, many pages are made in the networks. Therefore, in previous study, we proposed a new five dimensional hybrid recommender system for pages recommendation in social networks (Jafarpour & Kardan, 2014). On the other hand, whereas when people make decisions, they usually rely on recommendations; hence, accuracy of the recommendation plays significant role in users’ satisfaction. Consequently, in this paper, we examine and measure influence of factors to determine accuracy of pages recommendation in social networks. For this purpose, first, five factors which consist of social relation among users, user’s profile, page, event and group are considered. Second, according to the factors, sixteen categories are defined. Third, based on presented information by users, the users are placed in corresponding category and recommend pages by applying our proposed hybrid recommender system. Fourth, influence of each factors and their linear combination are computed in four sets of real dataset of Facebook to evaluate and validate the accuracy of recommendations and factors significance. The obtained results reveal that the accuracy of pages recommendation for the users who present their profiles is 89%. Moreover, the accuracy for the users who liked pages is 90%.
   

Social Media Links

Search

Login