VisCube Dataset

Overview

The VisCube Dataset merges detailed user profiles with various data visualization types and corresponding visual analysis tasks to foster the development of intelligent, user-adaptive visualization systems. By documenting the interaction between user characteristics and visualization task performance, this dataset is uniquely positioned to drive advancements in personalized visualization tools. It is open-source and aims to support research and development in the field.

Dataset Description

The dataset is divided into three components:

1. User Test Data: Profiles covering long-term characteristics such as personality traits and cognitive abilities of users.
2. Data Visualizations: Data visualizations used to construct the visual analysis tasks.
3. Analysis Tasks Data: Interaction data of users with various visualizations as they perform specific analysis tasks, including measurements of short-term cognitive states.

User Characteristics

Captures 9 different user characteristics through a battery of well-established psychological and HCI-related tests:

• Locus of Control
• Need for Cognition
• Big-Five Personality Test
• 5-Dimensional Curiosity Scale
• Perceptual Speed
• Visual Scanning
• Visual Disembedding
• Spatial Memory

Visualizations

The dataset includes 25 different types of visualizations, each with a shorthand code for quick reference:

• BC: Simple Bar Chart
• SB: Stacked Bar Chart
• DB: Diverging Stacked Bar Chart (with Neutral Parts)
• HS: Histogram (from binned data)
• HM: 2D Histogram Heatmap
• CS: Colored Scatterplot
• BU: Bubble Plot (Natural Disasters)
• ML: Multi-Series Line Chart
• SG: Slope Graph
• CO: Comet Chart
• AR: Area Chart
• SA: Stacked Area Chart
• HT: Annual Weather Heatmap
• LS: Lasagna Plot (Dense Time Series Heatmap)
• MS: Mosaic Chart with Labels
• RD: Radial Plot
• WF: Waterfall Chart of Monthly Profit and Loss
• PL: Parallel Coordinate Plot
• SD: Scatterplot with Mean and SD Overlay
• BP: Box-Plot with Pre-Calculated Summaries
• DP: Ranged Dot Plot
• HR: Horizon Graph
• RC: Horizontally Repeated Charts
• CH: Choropleth of Unemployment Rate per County
• OD: One Dot per Zipcode in the US

Visual Analysis Tasks

Five categories of tasks, each with a shorthand code for referencing specific tasks related to each visualization type:

• ID: Identification Task
• C: Comparison Tasks
• T: Trend Analysis Tasks
• RC: Relationship and Correlation Tasks
• DE1: Data Exploration

Examples of task references:

• HT_ID: Identification task for the Annual Weather Heatmap.
• CS_C: Comparison task for the Colored Scatterplot.

Short Cognitive State Measurement

After each task, a shortened test inspired by the NASA-TLX is administered to measure cognitive load, capturing changes in real-time cognitive states throughout different tasks.

Applications

This dataset supports a variety of applications in fields such as data visualization, HCI, educational technology, and adaptive systems design. It is ideal for:

• Developing personalized visualization systems.
• Academic research into the effects of user characteristics on visualization effectiveness.
• Enhancing user experience in educational and professional settings through tailored visual presentations.

Getting Started

1. Clone the Repository
2. Explore the Datasets
   • The User Test data is located in the file named "Part1_User_Tests.csv".
   • The Analysis Tasks data is stored in the file "Part2_Analysis_Tasks.csv".

Authors

• Fernando Yanez
• Ella Hugie
• Carolina Nobre

Acknowledgments

• Participants and collaborators who contributed to the dataset collection and validation.
• Funding and research support from the University of Toronto.