Large Scale Behavioral Analytics via Topical Interaction
This addresses the challenge of efficiently summarizing and interacting with large-scale behavioral data for analysts, though it appears incremental as it builds on existing dimension reduction methods.
The paper tackles the problem of visualizing massive behavioral data by introducing the split-diffuse algorithm, which distributes data points uniformly across visualization space to create topic grids, resulting in more perceivable topical analysis and interaction.
We propose the split-diffuse (SD) algorithm that takes the output of an existing dimension reduction algorithm, and distributes the data points uniformly across the visualization space. The result, called the topic grids, is a set of grids on various topics which are generated from the free-form text content of any domain of interest. The topic grids efficiently utilizes the visualization space to provide visual summaries for massive data. Topical analysis, comparison and interaction can be performed on the topic grids in a more perceivable way.