DICON: Interactive Visual Analysis of Multidimensional Clusters

Abstract

Fig. 1. DICON is a dynamic icon-based visualization technique that helps users understand, evaluate, and adjust complex multidimensional clusters. It provides visual cues describing the quality of a cluster as well as its multiple attributes, and can be embedded within many kinds of visualizations such as maps, scatter plots, and graphs.Abstract-Clustering as a fundamental data analysis technique has been widely used in many analytic applications. However, it is often difficult for users to understand and evaluate multidimensional clustering results, especially the quality of clusters and their semantics. For large and complex data, high-level statistical information about the clusters is often needed for users to evaluate cluster quality while a detailed display of multidimensional attributes of the data is necessary to understand the meaning of clusters. In this paper, we introduce DICON, an icon-based cluster visualization that embeds statistical information into a multi-attribute display to facilitate cluster interpretation, evaluation, and comparison. We design a treemap-like icon to represent a multidimensional cluster, and the quality of the cluster can be conveniently evaluated with the embedded statistical information. We further develop a novel layout algorithm which can generate similar icons for similar clusters, making comparisons of clusters easier. User interaction and clutter reduction are integrated into the system to help users more effectively analyze and refine clustering results for large datasets. We demonstrate the power of DICON through a user study and a case study in the healthcare domain. Our evaluation shows the benefits of the technique, especially in support of complex multidimensional cluster analysis.

Sci-Fin: Visual Mining Spatial and Temporal Behavior Features from Social Media

“…First, the fingerprint designs to capture the characteristics of check-in data from region, activity, and user aspect. The designs are intuitive, compact, and informative and leverage several well established techniques like WorldMapper and Voronoi Treemap; Second, the compact design can encode temporal changes especially periodic patterns with their circular layout and dynamic placement on maps to represent its spatial attributes; Third, the encoding components such as color and shape can be well scaled without any significant loss of information as in [42]. This allows the design to remain effective for both large and small sized icons; Finally, our fingerprint compresses check-in users’ spatial and temporal information with multiple attributes into relatively small icons which can easily be embedded within other visualizations as graphs or tables.…”

VINCENT: A visual analytics system for investigating the online vaccine debate

“…has incorporated similar data analysis techniques used in VINCENT. For example, researchers have investigated how incorporating multi-dimensional scaling of co-occurrence data (discussed in section 2.3) in VASes help users investigate entities and identify clusters in a variety of data sets [28,29]. As well, researchers have utilized emotion analysis (discussed in section 2.4) in VASes that help users investigate online text from both social media and the general web regarding a variety of topics [30][31][32].…”

SOMFlow: Guided Exploratory Cluster Analysis with Self-Organizing Maps and Analytic Provenance

“…Visual interactive clustering solutions exist for a variety of data types. E.g., the work of Andrienko et al [2] proposes methods to group movement trajectories, Ruppert et al [40] describe visual interactive workflows to cluster textual documents, Cao et al [16] focus on the interactive analysis of multidimensional clusters, and the approach by Nam et al [34] focuses on high-dimensional data. Some of them also let the user select a specific subset of interest where another subsequent computation of the clustering/classification is applied (e.g., the work by Choo et al [18]).…”