A Tale of Two Cohesion: A Review of Group Cohesion in Social Psychology and Social Computing

Yining Zhao, Sourav S Bhowmick, Nastassja L. Fischer, Annabel Chen Shen-Hsing

Abstract:

Community detection and community search are two prominent social computing problems aimed at identifying cohesive groups, with wide-ranging social applications. In most studies, the precise quantification of cohesion is fundamental to the effective identification of communities. However, although the concept of cohesion originates in social psychology, its structural dimensions are the only aspects that have been formally captured through cohesiveness metrics. As a result, existing algorithms may be inadequate for addressing the nuanced requirements of community-driven social applications.

In this paper, we present a narrative review of cohesion definitions and measurement approaches across both social psychology and social computing, guided by our proposed unified framework. By examining the connections and disparities between these disciplines, we draw on insights from social psychology to inform the design of cohesiveness metrics that are psychologically meaningful. This review lays the groundwork for bridging these traditionally disparate fields, fostering interdisciplinary collaboration, and advancing the development of cohesiveness metrics capable of identifying communities that embody psychology-informed cohesion within online social networks.