SuperNOVA: Design Strategies and Opportunities for Interactive Visualization in Computational Notebooks (2305.03039v2)

Abstract: Computational notebooks, such as Jupyter Notebook, have become data scientists' de facto programming environments. Many visualization researchers and practitioners have developed interactive visualization tools that support notebooks, yet little is known about the appropriate design of these tools. To address this critical research gap, we investigate the design strategies in this space by analyzing 163 notebook visualization tools. Our analysis encompasses 64 systems from academic papers and 105 systems sourced from a pool of 55k notebooks containing interactive visualizations that we obtain via scraping 8.6 million notebooks on GitHub. Through this study, we identify key design implications and trade-offs, such as leveraging multimodal data in notebooks as well as balancing the degree of visualization-notebook integration. Furthermore, we provide empirical evidence that tools compatible with more notebook platforms have a greater impact. Finally, we develop SuperNOVA, an open-source interactive browser to help researchers explore existing notebook visualization tools. SuperNOVA is publicly accessible at: https://poloclub.github.io/supernova/.

• The paper systematically analyzed 161 visualization tools in computational notebooks to derive effective design strategies.
• It introduces a framework that categorizes tools based on motivation, target users, and design dimensions for seamless integration.
• The study provides actionable insights on integration levels and user feedback to enhance interactive visualization in data workflows.

SuperNOVA: Design Strategies and Opportunities for Interactive Visualization in Computational Notebooks

The paper "SuperNOVA: Design Strategies and Opportunities for Interactive Visualization in Computational Notebooks" by Zijie J. Wang et al. investigates the landscape of interactive visualization tools integrated within computational notebooks such as Jupyter and Colab. It addresses an existing research gap by systematically analyzing 161 visualization and analytics tools (VA tools) for notebooks, thus shedding light on effective design strategies for developers and researchers.

Systematic Analysis of Notebook VA Tools

The authors conducted a comprehensive review of notebook VA tools, examining 64 systems introduced in academic papers, and 103 systems identified from a large dataset of 8.6 million GitHub-hosted notebooks. User feedback from 15 studies and 379 GitHub issues were also analyzed. The paper provides detailed insights into the motivation behind creating these tools, their targeted users, and various design considerations.

Design Framework and SuperNOVA Development

The paper develops an organizational framework categorizing notebook VA tools based on:

• Motivation: Seamless integration with user workflows, access to computational artifacts, and the portability of notebooks.
• Target Users: Data scientists, scientific researchers, and educational audiences.
• Design Dimensions: The degree of VA-notebook integration, data source/type, display styles, and modularity.

This framework clarifies the landscape, helping researchers design adaptable tools for various user needs. Furthermore, the authors introduced SuperNOVA, an open-source interactive tool enabling exploration and discovery of existing notebook VA tools to inspire new designs and aid in research.

Insights into VA-Notebook Integration and Data Communication

The paper identifies three levels of integration based on data communication:

1. No Direct Communication: Tools that don't need data input directly from notebooks.
2. One-way Communication: Tools receiving data (e.g., dataframes) from notebooks to create visualizations.
3. Bidirectional Communication: Tools that send data back to notebooks, enabling deeper interaction by adjusting notebook states or generating code.

Advantages and limitations of varying integration levels are discussed, and implementation methods are outlined, helping developers decide on appropriate strategies.

User Experience and Feedback Analysis

User satisfaction with seamless VA-tool integration into notebooks was generally positive, particularly among data scientists working in exploratory roles. However, automatic content updates in notebooks received mixed reactions. Some users welcomed code-generation features for productivity, whereas others preferred minimal intervention.

The paper documents compatibility challenges across multiple platforms such as Colab and Jupyter, emphasizing the importance of considering platform interoperability during development.

Implications and Future Directions

This systematic exploration suggests that notebook VA tools can significantly transform data analysis and dissemination if they leverage notebook environments' native strengths. For developers and researchers, this means utilizing notebook artifacts and carefully considering integration levels. Furthermore, SuperNOVA could act as a catalyst for generating innovative tools that accommodate the evolving needs of data-centric workflows.

In conclusion, the paper provides a robust foundation for future research and development efforts aimed at enhancing the usability and functionality of visualization tools in computational notebooks. The findings offer a significant contribution to the field of visual analytics, guiding the development of new systems that are well-tuned to the context of computational notebooks.