From Visual Exploration to Storytelling and Back Again.

Area of Science:

• Data Visualization
• Scientific Communication
• Human-Computer Interaction

Background:

• Documenting visual data exploration is crucial for insight reproducibility.
• Current methods like images and videos lack provenance and interactivity.
• Existing approaches fail to capture the full context and history of discovery.

Purpose of the Study:

• To present CLUE (Capture, Label, Understand, Explain), a model integrating data exploration with discovery presentation.
• To enable users to document and share their visual data exploration process.
• To facilitate the review, reproduction, and extension of data-driven insights.

Main Methods:

• CLUE captures provenance data during visual data exploration.
• Users can extract key exploration steps, add annotations, and create "Vistories".
• Vistories are visual stories based on the exploration history, allowing for review and extension.

Main Results:

• A prototype implementation of the CLUE model is provided.
• The model demonstrates applicability in public health data exploration and molecular biology.
• Vistories can be shared for viewing, retracing, and extending analyses.

Conclusions:

• CLUE offers a robust solution for documenting and communicating visual data discoveries.
• The Vistories approach enhances the reproducibility and collaborative potential of data exploration.
• Integrating CLUE into visualization tools can significantly improve scientific communication and insight validation.