On the Role of Social Identity and Cohesion in Characterizing Online Social Communities

Hemant Purohit, Yiye Ruan, David Fuhry, Srinivasan Parthasarathy, Amit P. ShetharXiv 2012 · arXiv

Abstract

This paper investigates the drivers of online community sustainability by comparing two social psychology theories: social cohesion (structural ties) and social identity (shared characteristics). By analyzing Twitter data from major events such as Hurricane Sandy and Occupy Wall Street, the authors demonstrate that shared identities and structural "tightness" are key indicators of a group's ability to maintain a focused, sustainable discussion. The research introduces "Topic Divergence" as a novel metric for quantifying community focus over time.

1. Introduction

In the rapidly evolving landscape of social media, communities form and dissolve with high frequency. Determining what makes a group "sustainable" is a critical challenge. Traditional metrics like group size or growth rate are often "leaky"; a group can expand in numbers while losing its core members or its thematic focus. This study bridges the gap between social psychology and computer science to identify the socio-psychological features that predict community longevity and focus.

2. Methodology

The researchers proposed a multi-faceted approach to modeling group behavior through three primary lenses:

  • Group Identification: Utilizing multi-level graph clustering based on actual interactions (retweets, replies, and mentions) rather than simple hashtag tracking to identify organic social groups.
  • Quantifying Social Identity: Measuring member similarity via Entropy across three dimensions: Regional (geographic location), Expertise (professional background extracted from bios), and AID Identity (Activity, Influence, and Diffusion metrics).
  • Quantifying Social Cohesion: Analyzing the "tightness" of the follower graph using density, transitivity, and average shortest path length, comparing both reciprocal and undirected ties.
  • Topic Divergence: A novel metric using Dynamic Topic Models (DTM) to calculate how much an individual member’s discussion strays from the group’s central conversation using Kullback-Leibler (KL) divergence.

3. Datasets and Event Classification

The study utilized five large-scale Twitter datasets representing diverse real-world events. These events were categorized based on their duration (Lasting vs. Transient) and their structural nature (Loose vs. Compact).

Classification of studied events based on duration and structural tightness.

As shown in the classification table, events like Hurricane Irene and Sandy are transient and loose, whereas Occupy Wall Street represents a lasting but loose community. Anti-SOPA stands out as a compact, transient event. The scale of these datasets, ranging from 21K to 1.8M users, provides a robust foundation for analyzing different community dynamics.

Summary statistics of the five Twitter datasets used in the study, including duration and volume.

4. Key Insights

  • Sustainability as Focus: A group is considered sustainable if its members remain "on topic." High topic divergence indicates a loss of community voice.
  • Identity in Crisis: In transient events like natural disasters, shared identity (specifically physical location) is a stronger predictor of sustainability than structural ties. These ad-hoc groups rely on shared geographic fate rather than established follower networks.
  • The Reciprocity Myth: Data suggests that mutual following (reciprocal ties) is not necessarily a better indicator of sustainability than one-way following in digital environments.
  • Size vs. Cohesion: Traditional growth rates showed almost no correlation with identity or cohesion features, confirming that "getting bigger" and "staying together" are distinct social phenomena.

5. Results

The empirical analysis yielded several significant correlations:

  • Identity Correlation: Regional identity entropy showed a moderate to high positive correlation with topic divergence (0.25 to 0.58), meaning geographically diverse groups were more likely to have unfocused discussions.
  • Structural Cohesion: High group density and low average shortest path lengths were strong indicators of lower topic divergence. For Hurricane Irene, the correlation between path length and topic divergence was 0.57.
  • Event Characteristics: Protests like Occupy Wall Street exhibited higher identity entropy than disasters, likely because they revolve around diverse opinion leaders rather than a single shared geographic event.

6. Conclusion

This research highlights that both structural cohesion and shared social identity are vital for maintaining focused online communities. While the study is currently limited to Twitter and relatively simple identity models, it provides a framework for future cross-domain validation in professional networks and other social platforms. Future work will explore more nuanced identity features and the utility of curated data sources like Twitter Lists.

Summary updated

Abstract

Two prevailing theories for explaining social group or community structure are cohesion and identity. The social cohesion approach posits that social groups arise out of an aggregation of individuals that have mutual interpersonal attraction as they share common characteristics. These characteristics can range from common interests to kinship ties and from social values to ethnic backgrounds. In contrast, the social identity approach posits that an individual is likely to join a group based on an intrinsic self-evaluation at a cognitive or perceptual level. In other words group members typically share an awareness of a common category membership. In this work we seek to understand the role of these two contrasting theories in explaining the behavior and stability of social communities in Twitter. A specific focal point of our work is to understand the role of these theories in disparate contexts ranging from disaster response to socio-political activism. We extract social identity and social cohesion features-of-interest for large scale datasets of five real-world events and examine the effectiveness of such features in capturing behavioral characteristics and the stability of groups. We also propose a novel measure of social group sustainability based on the divergence in group discussion. Our main findings are: 1) Sharing of social identities (especially physical location) among group members has a positive impact on group sustainability, 2) Structural cohesion (represented by high group density and low average shortest path length) is a strong indicator of group sustainability, and 3) Event characteristics play a role in shaping group sustainability, as social groups in transient events behave differently from groups in events that last longer.
BibTeX
@misc{purohit2012role,
  author = {Hemant Purohit and Yiye Ruan and David Fuhry and Srinivasan Parthasarathy and Amit P. Sheth},
  title = {{On the Role of Social Identity and Cohesion in Characterizing Online Social Communities}},
  year = {2012},
  eprint = {1212.0141},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1212.0141},
}
Figure