ChronoDeck: A Visual Analytics Approach for Hierarchical Time Series Analysis

Area of Science:

• Data Visualization
• Human-Computer Interaction
• Time Series Analysis

Background:

• Hierarchical time series data analysis is crucial for insights in retail, finance, and energy.
• Current interactive tools lack comprehensive analysis across aggregation levels and complex tasks.
• Existing methods are limited to basic operations like summarization and comparison.

Purpose of the Study:

• To develop an advanced visual analytics approach for hierarchical time series data.
• To address the limitations of existing interactive exploratory analysis tools.
• To support complex analytical tasks beyond simple summarization and comparison.

Main Methods:

• Developed a generalized taxonomy for hierarchical time series analysis tasks.
• Created ChronoDeck, an interactive system with multi-column visualization.
• Integrated coordinated dimensionality reduction and small multiples visualizations.
• Incorporated interactions like highlight, align, filter, and select.

Main Results:

• ChronoDeck facilitates visualization, comparison, and transformation of hierarchical time series.
• The system aids in identifying entities of interest within the data.
• Case studies on real-world datasets and expert interviews validated its effectiveness.

Conclusions:

• ChronoDeck offers a novel and effective solution for analyzing hierarchical time series data.
• The system enhances user capabilities in distilling insights from complex datasets.
• The developed taxonomy and system advance the field of visual analytics for time series.