Title: Network dynamics in friend recommendation: a study of Indian engineering students
Authors: Pradip Swarnakar; Ajay Kumar; Himanshu Tyagi
Department of Humanities and Social Sciences, ABV-IIITM Gwalior, Gwalior-474015, MP, India
Department of Applied Sciences, ABV-IIITM Gwalior, Gwalior-474015, MP, India
Department of Information Technology, ABV-IIITM Gwalior, Gwalior-474015, MP, India
Abstract: Social networks are the frequently used service which over the past few years, have grown by leaps and bounds. In this study, a survey of engineering students has been conducted to find out the most relevant socio-demographic and webographic factors that are considered by users while sending/accepting friend requests. An extensive survey has been conducted and the responses were used to determine the influence of various factors in friend request attributes. Based on the collected responses, logistic regression and artificial neural network models have been developed for predicting the users' friend request attributes. A comparative performance analysis of these models to predict the friend request attributes has also been done. The results indicate that neural network model outperformed the logistic regression model when data are nonlinear. The study also shows that among all the factors, users' gender, photographs, hometown, age, and shared interests are the most significant factors.
Keywords: social networks; Facebook; user perception of friendship; logistic regression; artificial neural network; ANN.
Int. J. of Information Technology and Management, 2017 Vol.16, No.3, pp.287 - 300
Available online: 19 May 2017