User Behavior Profiling in Social Media Applications
BOJA CATALIN, ZAMFIROIU ALIN, Mădălina Zurini, Bogdan Iancu · ECONOMIC COMPUTATION AND ECONOMIC CYBERNETICS STUDIES AND RESEARCH · 2019
Given the increase of sensitive data stored within mobile devices and social media applications, the need for creating a user behaviour profile is in high demand for preventing security breaches, impersonation or unauthorized access to resources.The behaviour of a user is defined by the aggregation of different patterns that are obtained while constantly using a software service on a mobile device or a computer.The scope of this study is to identify a core group of characteristics that can be further used in profiling a user based on his behaviour.For that, a survey consisting of 20 questions was conducted having a set of 356 responders.The obtained data were pre-processed and used as input for supervised and unsupervised classification techniques.Two applications, one web-based and one mobile, were implemented in order to expand the prior selection of characteristics and to verify the obtained results.They were used to measure interaction events and ways of using social media on both versions of the applications.Taking into account these different behaviour based characteristics, we defined different groups of users that are uniquely identifiable by limited sets of characteristics.