Patrick and his students did a pretty good job to visualize blogposts. They wrote a script that parsed my RDF file, then extracted the most important topics following Wise and colleagues' method. At the end of the road, Patrick used R to compute two visualizations. The first one is a representation of the sematic distance between blogposts (thanks to Multidimensional Scaling). The second one depicts a results of a Clara cluster analysis, in which 10 clusters have been built. At the center of each cluster is a prototypical article. Another visualisation, using “magnet” layout to browse topics and posts is under construction. Even tough these visualizations are nice and interesting, they've having hard times making sense of them! Especially the second one, trying to infer meaning from cluster is difficult (getting back to the prototypical posts at the center of the medoids is a solution to get some insights about it).