Story Clustering

Storyclash collects social media contributions from online media and companies. Headlines and interactions are stored. The WIN-LAB has developed a clustering procedure that groups the social media contributions into topics using unsupervised learning methods such as LDA. For each company, a topic profile will be automatically created that characterizes that company in terms of its social media publications.

Interactive data visualizations are used to interpret the clusters as well as the distances between the clusters.