Abstract
In many real-world settings, it is essential for a group of interacting individuals to reach shared understanding and consensus on a given issue. Consensus can strengthen groups and their impact on society. The popularity of online social and collaboration networks has influenced the way individuals interact with each other. Many collaboration sites (i.e., StackExchange, Reddit or Wikipedia) enable users to exchange opinions, discuss certain topics and solve problems while interacting with other online users. This thesis aims at uncovering the dynamics of consensus building among users collaborating online. Consensus dynamics is closely related to the process of opinion dynamics. Opinion dynamics has been studied from the perspective of social sciences, physics, mathematics, complex system studies and network science. However, such studies often remain confined to these disciplines. Therefore, this thesis applies an interdisciplinary approach. It builds hypotheses based on social science theories, simulates
opinion dynamics utilizing agent-based models from statistical physics and applies social network analysis on empirical datasets extracted from the Web. Methodologically, this thesis contributes a novel framework to study the role and interplay of some of the main factors in consensus building (i.e., users social status, network structure, users similarity and content creation). The presented method can be applied to run extensive simulations of opinion dynamics in arbitrary collaboration networks. The empirical findings of this thesis help draw recommendations on how to integrate the influence of user characteristics (e.g., social status) in opinion dynamics to optimize consensus building. Additionally, this thesis experimentally demonstrates how content dynamics drives the process of agreement and disagreement between users collaborating online. These results add to our understanding of the challenges of designing and implementing services that promote consensus building.
opinion dynamics utilizing agent-based models from statistical physics and applies social network analysis on empirical datasets extracted from the Web. Methodologically, this thesis contributes a novel framework to study the role and interplay of some of the main factors in consensus building (i.e., users social status, network structure, users similarity and content creation). The presented method can be applied to run extensive simulations of opinion dynamics in arbitrary collaboration networks. The empirical findings of this thesis help draw recommendations on how to integrate the influence of user characteristics (e.g., social status) in opinion dynamics to optimize consensus building. Additionally, this thesis experimentally demonstrates how content dynamics drives the process of agreement and disagreement between users collaborating online. These results add to our understanding of the challenges of designing and implementing services that promote consensus building.
| Original language | English |
|---|---|
| Qualification | Doctor of Technology |
| Awarding Institution |
|
| Supervisors/Advisors |
|
| Publication status | Published - Jan 2019 |
Keywords
- Consensus Building
- Opinion Dynamics
- Content Dynamics
- Agent-Based Models
- Online Collaboration Networks
- Social Influence
- Social Science
- Statistical Physics
- Mathematical Models
- Complexity Studies
- Network Science
- Social Network Analysis
- Computational Social Science
- Q&A Sites
- StackExchange
- Wikipedia
- Co-Authorship Networks