Agent-Based Model of Virtual Community Cohesion (S/O: 13443)

• Main Authors: Yusop, Nor Iadah, Mat Aji, Zahurin, Ab. Aziz, Azizi, Md Dahalin, Zulkhairi
• Format: Monograph
• Language: English
• Published: UUM 2021
• Subjects: HT Communities. Classes. Races
• Online Access: https://repo.uum.edu.my/id/eprint/31660/1/13443.pdf; https://repo.uum.edu.my/id/eprint/31660/

This report provides an analysis and evaluation of a study conducted to predict the state of community cohesion through simulation. Community cohesion is a broad concept that commonly connotes better relationships among people of different backgrounds. Cohesion is a state that can only be witnessed as a result of long-term interactions amongst the participating members of a community, commonly in the physical community. In the era in which social media is becoming a dominant tool for social interactions, it is crucial to understand if virtual communities could be used to attain cohesion in the future. This study was conducted to predict the dynamicity of cohesion among members of the virtual community through simulation. The virtual community cohesion model was adopted as the theoretical basis for this exercise. Originally, the model lists the factors and components that are pertinent to virtual community cohesion. In achieving the aim of this study, an agent-based model (ABM) was applied to help understand the interactions between people (agents) who, in turn, influence one another. ABM bridges the descriptive approach used in the social sciences and the formal approach used in the hard sciences by focusing on the processes or behaviours that build the social reality. The relationships among agents were represented as a conceptual model for virtual community cohesion. Simple individual rules or actions are applied as changes in these individual acts can affect the collective group in any given population, and are expected to reveal the coherent behaviour as well as the emergent group behaviour. Those relationships were then modelled using a set of mathematical equations. The formal specifications (differential equations) equations form the basis of algorithm development, which preceded the development of a virtual community cohesion prediction simulator. The simulator was used as an experiment to explore exciting patterns and traces that explain the agent model's behaviour related to virtual community cohesion as asserted in related literature. The simulations results reveal several interesting traces (patterns) based on four selected conditions (three baselines and one mixed variation case). These traces show the necessary what-if conditions as the resulting inputoutput collaboration patterns in the entire virtual community cohesion emerge. Also, it is shown that agents' behaviour based on individual attributes can reduce the variability of cohesion levels. This result can be explained by the fact that virtual community cohesion is based on the perceived experiences, trust, participation and interactions by all community members. Moreover, this model has been evaluated using mathematical analysis (whereby theoretical analysis and value substitution analysis are performed) and automated logical verification (internal validation based on selected previous empirical study. In conclusion this study has achieved the objectives that were set. The simulation traces and patterns provide indications that the internal validation has been successful. In the future the external validation process to evaluate the model could be done. As such, a study could be conducted to confirm if the model holds true based on empirical evidences