OutlineSpark: Igniting AI-powered Presentation Slides Creation from Computational Notebooks through Outlines

Fengjie Wang,Yanna Lin,Leni Yang,Haotian Li,Mingyang Gu,Min Zhu,Huamin Qu
2024-03-14
Abstract:Computational notebooks are widely utilized for exploration and analysis. However, creating slides to communicate analysis results from these notebooks is quite tedious and time-consuming. Researchers have proposed automatic systems for generating slides from notebooks, which, however, often do not consider the process of users conceiving and organizing their messages from massive code cells. Those systems ask users to go directly into the slide creation process, which causes potentially ill-structured slides and burdens in further refinement. Inspired by the common and widely recommended slide creation practice: drafting outlines first and then adding concrete content, we introduce OutlineSpark, an AI-powered slide creation tool that generates slides from a slide outline written by the user. The tool automatically retrieves relevant notebook cells based on the outlines and converts them into slide content. We evaluated OutlineSpark with 12 users. Both the quantitative and qualitative feedback from the participants verify its effectiveness and usability.
Human-Computer Interaction
What problem does this paper attempt to address?
### The Problem the Paper Aims to Solve This paper aims to address the tedious and time-consuming process of creating presentation slides from computational notebooks (such as Jupyter Notebooks). Specifically: - **Current Issues**: Computational notebooks are commonly used for data exploration and analysis, but they are often lengthy and poorly formatted. Users need to present key analysis details and results from these notebooks to team members or stakeholders, which requires them to use external tools (such as Google Slides or Microsoft PowerPoint) to create slides. This process involves conceptualizing the slide structure and content, locating and extracting relevant cells, and adjusting the slide structure, all of which are very time-consuming. - **Existing System Shortcomings**: Existing systems like NB2Slides and Slide4N attempt to automate slide generation, but the former is only suitable for notebooks that build machine learning models and requires high-quality documentation support; the latter allows users to select cells of interest to generate slides but still requires users to have a clear understanding of the overall slide structure. - **Solution**: This paper proposes OutlineSpark, a tool that automatically generates slides based on user-written outlines. OutlineSpark improves the slide creation process in the following ways: - Supports users in writing outlines to conceptualize the slide structure. - Automatically retrieves notebook cells related to outline items and converts them into slide content. - Provides an interactive interface that allows users to easily modify and optimize the generated slides. In this way, OutlineSpark simplifies the entire process of creating presentation slides from computational notebooks, enabling users to focus more on the content and logical structure of their presentations.