Determining Big Data Dynamics

We continue at the CCC Symposium with the great Alex Halavais, who is interested in the first place in the hidden patterns in data, and the learning - the evolution of ideas - which might result from them. But how do we detect such learning, such change? One indicator could be the popularity of content or users - success may be measured in the amount of attention received, for example. But all of this also happens at multiple scales, in multiple contexts - search engines cannot simply produce one result, for example, but must produce the right results for a wide variety of different search contexts.

Of particular importance in this are the dynamics of activity, of course - how does learning take place; how do participants change, adapt (improve?) their activities in order to be more successful; how do influential users rise to their positions of influence? What are the strategies users employ, in what circumstances, and what effects do they have? What are the patterns that lead to specific roles in a community?

But of course it's not only the lead, central users who are worthy of research - those users who just 'hang out' rather than participating in depth are of interest as well, but are often considered to be outside the intended scope of research projects.

At any rate, what's especially interesting here are changes over time: the dynamics of networks. There's a feedback loop which is important here: reporting back to the subjects of the network mapping process – making them aware of the fact that they're being mapped – may affect their behaviours, for example, as they attempt to change how they appear in the research results.

Quantitative and qualitative research both involve interpretation - though this tends to be more evident in qualitative work, where interpretation is generally accepted as necessary, while the results of quantitative work are often presented as objective and self-evidently true; if so, this too is a problem to be addressed.