Hierarchical representation of tagged data for visualization, aggregation and navigation

Digital social networks generate massive amounts of data that are sometimes tagged by their users. In this work we have developed an analytic pipeline to extract and visualize a hierarchical representation of the data for navigation and aggregation purposes.For this thesis, we gathered a collection of Q&A forums containing millions of tagged ques- tions. Our objective is to create a navigable structure from the data that provides a con- tinuum of views from the big picture, down to the details.We also devised a new algorithm to prune the data while preserving the important parts and relationships among data for very large datasets. We also showed a recursive aggregation approach to generate labels and timelines for intermediate nodes in the hierarchy.