PyramidViz: Visual Analytics and Big Data Visualization of Frequent Patterns

Frequent pattern mining aims discover implicit, previously unknown and potentially useful knowledge in the form of frequently co-occurring items, events, or objects. These discovered frequent patterns helps reveal interesting relationships such as consumer shopper behaviour. Existing mining algorithms mostly return a long textual list of frequent patterns to users. Such a long list may not be comprehensible by many users. As a picture is worth a thousand words, visual representation of frequent patterns is more comprehensible. Consequently, several visualizers have been proposed. While they are popular and benefit from a few advantages, they also suffer from some disadvantages. In this paper, we present a visual analytic solution, called PyramidViz, for visualizing and analyzing frequent patterns. PyramidViz shows frequent patterns in an informative and intuitive fashion so that users can easily get an insight about frequency of frequent patterns and relationships (e.g., prefix-extension relationships) among related frequent patterns. Evaluation results show the effectiveness and practicality of PyramidViz in visual analytics and big data visualization of frequent patterns for various real-life applications.