There is a multitude of very complex and interconnected political, socio-economic and environmental issues facing our globalizing society. To address these, topic communities are essential of experts and stakeholders collaborating closely for a longer period of time. These topic communities often need to be created ad hoc and urgently, however, while demanding a unique mix of experience and expertise of their members. Thus, the formation of such communities is far from trivial. Existing scientific and political structures do not suffice to provide the right experts and stakeholders in time. I outline a socio-technical approach for topic community member selection, analyzing large corpora of blog posts to identify combinations of topics and bloggers relevant to the goals of the topic community. The technical basis for the approach is the tOKo tool for text analysis. The social aspect consists of a sequence of steps of human interpretation of the blog analysis results that tOKo produces. This socio-technical approach forms a ''pragmatic funnel'', producing a set of candidate topic community members likely to be relevant. I illustrate the approach with a realistic case.

Relevant links:
www.communitysense.nl
www.toko-sigmund.org

Watch the webcast replay >>