Visualizing Collective Discursive User Interactions in Online Life Science Communities

TL;DR

This paper introduces BioViz, a visualization tool for analyzing collective user interactions in online life science communities, revealing patterns of collective behavior and organization.

Contribution

The paper presents a novel matrix-based visualization approach for understanding collective interactions in diverse online science forums.

Findings

Patterns of collective behavior identified

Visualization reveals organizational structures

Potential applications in collective intelligence study

Abstract

This paper highlights the rationale for the development of BioViz, a tool to help visualize the existence of collective user interactions in online life science communities. The first community studied has approximately 22,750 unique users and the second has 35,000. Making sense of the number of interactions between actors in these networks in order to discern patterns of collective organization and intelligent behavior is challenging. One of the complications is that forums - our object of interest - can vary in their purpose and remit (e.g. the role of gender in the life sciences to forums of praxis such as one exploring the cell line culturing) and this shapes the structure of the forum organization itself. Our approach took a random sample of 53 forums which were manually analyzed by our research team and interactions between actors were recorded as arcs between nodes. The paper focuses on a discussion of the utility of our approach, but presents some brief results to highlight the forms of knowledge that can be gained in identifying collective group formations. Specifically, we found that by using a matrix-based visualization approach, we were able to see patterns of collective behavior which we believe is valuable both to the study of collective intelligence and the design of virtual organizations.