Understanding online behavior: Exploring the probability of online personality trait using supervised machine-learning approach

Abstract

The notion of online anonymity is based on the assumption that on the Internet the means of identification are limited to network and system identifiers, which may not directly relate to the identity of the user. Personality traits as a form of identity have recently been explored. A myriad of relationships between the Internet and human personality traits have been examined based on correlation and regression of media usage specific to selected media platforms, such as social networking sites. In these studies, the link between humans and the Internet based on interests and disposition was studied. However, the paradigm of the existence of a platform-independent digital fingerprint of personality trait is yet to be explored. This paradigm considers the Internet an extension of human daily communication that is capable of exhibiting a digital behavioral signature. Therefore, in this study, using client-server interaction as the fundamental unit of online communication, the probability of a digital personality trait distinction was explored. A five-factor model of a personality trait measurement instrument and server-side network traffic data collected over 8 months from 43 respondents were analyzed using supervised machine-learning techniques. The results revealed a high probability that the signature of conscientiousness personality trait exists in online communication. This observation presents a novel platform for the exploration of online identity. Furthermore, it charts a new research focus on human digital signatures capable of characterizing online behavior.