Structural Creation of a Social Network: An Analysis of Longitudinal Development of Online Communities
Summary
Over the past ten years, online communities have attracted significant attention from a wide range of organizations in various markets. Social interactions within these virtual communities have logged an enormous amount of digital traces, and opened up new opportunities to researchers in the field of social network studies. However, current research in this field seems mainly focused on the individual participant, while very little has been researched from a structural, time-based [longitudinal] network perspective. One of the crucial issues that comes with it, is that there is insufficient knowledge regarding the aspect of evolution in online communities. Such issues get aggravated when the data extracted from online communities are unstructured (i.e. without explicit relational references) and are cumulated over a long period of time.
With this in mind, in this study we carried out a series of experiments and analysis on a public online forum [online community] with a hierarchical network structure. The objective is to identify scientific methods from existing literature for creating network structures based on unstructured data, as well as scientific approaches for longitudinal SNA. Finally, we try to determine the validity and applicability of existing literature on social network analysis in online communities from a practical point of view, in order to contribute to a scientific approach in investigating longitudinal development of online communities, from a network’s perspective.
As a result, we identified certain limitations in the existing literature and proposed our own methods and guidelines for social network creation and longitudinal SNA. The results of our work revealed the critical issues related to working with unstructured and longitudinal communication data retrieved from an online forum.. The analytical results of edge-ratio analysis suggest that the development of an online community can be monitored periodically based on fractions of the whole network structure (i.e. network snapshots). Indications are that the development of an online community is two-folded, i.e. participation of “old” users as well as fresh in-streams of new users is roughly equally important. Last but not least, an attempt to visualize an online community’s development in a longitudinal manner with the Gephi software package has provided discernible insights, as compared to other methods.