NodeXL, Small Data and Information Architecture

NodeXL is feted as a tool for dealing with large and complex social networks, but it also has uses in much smaller data sets.

My PhD research considers participatory media networks that circulate information locally and allow people to contribute to and shape their local communities.

I’m essentially taking the global two-way communication networks and tools underpinned by the Internet and considering how they operate on a local scale, down to the local government level. My case studies are three local government areas in the Australian state of New South Wales. Two of them are semi-rural areas with only about 30,000–40,000 residents and the third is a coastal city with about 250,000 residents.

The basic necessity in considering these kinds of networks is establishing the details of the architectures they operate within both for the purposes of description and to guide further methodological development. For local governments with many disparate business units, information networks like social network sites and websites are operated almost independently by each unit. The outcome is that often there is not very well-defined information flow or even understanding of what sites exist, who runs them, and how they’re used.

After trialling a few other design tools, I decided to map the information architectures of each selected local government in NodeXL. I prepared details of entity-entity relationships and entity-account relationships organised by media type. Importing these into NodeXL and manually adjusting the networks to show each media type and each council as major nodes allowed simple visualisation which nonetheless provides useful overview of the information architecture. In addition to the combined graph above, I prepared individual graphs for each of the three local governments.

I’m supplementing this basic information with more detailed graphs (draft example below), also prepared in NodeXL, that visualise broader networks of connections and socialisation between the councils, their constituents and other groups on Twitter and Facebook.

Using NodeXL to map very small datasets like council media networks provides a convenient way to provide visually appealing overviews of information architectures. This methodology could prove useful for many small to medium organisations with multiple subunits to account for the information networks maintained by or on behalf of their organisation.