ChronAtlas: A Visualization for Dynamic Topic Exploration

Documents in rich text corpora such as digital libraries and social media often contain complex information. These data resources are huge in amount, dynamic in nature and contain multifaceted information. This poster presents ChronAtlas, a visual analytic technique for visually exploring topics in multifaceted dynamic data. ChronAtlas simultaneously visualizes the topic distribution of the underlying entities from one facet together with keyword distributions that convey the semantic definition of each cluster along a secondary facet. ChronAtlas combines several visual techniques including 1) topic contour clusters and interactive multifaceted keyword topic rings, 2) a global layout optimization algorithm that aligns each topic cluster with its corresponding keywords, and 3) an optimal temporal network summarization and clustering algorithm that renders evolution trends of clusters.