NotePlayer: Engaging Jupyter Notebooks for Dynamic Presentation of Analytical Processes

TL;DR

NotePlayer tackles the challenge of communicating analytical workflows in Jupyter notebooks by linking notebook cells to video segments and automating narration with language models. Guided by a formative study and content analysis of 38 tutorial videos, the authors design an architecture with interactive modules and a computational engine that uses a design script and LLM assistance to generate dynamic, scene-based presentations. The system is evaluated via a usage scenario and a user study (n=12), which indicate improved expressiveness, usability, and learnability, while also highlighting areas for refinement such as more flexible editing and richer animations. Overall, NotePlayer offers a practical, semi-automatic pathway to produce expressive notebook tutorials that combine code, narration, and visuals, with potential to extend to other domains and deeper integration with existing authoring tools.

Abstract

Diverse presentation formats play a pivotal role in effectively conveying code and analytical processes during data analysis. One increasingly popular format is tutorial videos, particularly those based on Jupyter notebooks, which offer an intuitive interpretation of code and vivid explanations of analytical procedures. However, creating such videos requires a diverse skill set and significant manual effort, posing a barrier for many analysts. To bridge this gap, we introduce an innovative tool called NotePlayer, which connects notebook cells to video segments and incorporates a computational engine with language models to streamline video creation and editing. Our aim is to make the process more accessible and efficient for analysts. To inform the design of NotePlayer, we conducted a formative study and performed content analysis on a corpus of 38 Jupyter tutorial videos. This helped us identify key patterns and challenges encountered in existing tutorial videos, guiding the development of NotePlayer. Through a combination of a usage scenario and a user study, we validated the effectiveness of NotePlayer. The results show that the tool streamlines the video creation and facilitates the communication process for data analysts.

Figures (9)

• Figure 1: Boxplot of durations for different video segments. Mean durations are as follows: Background at 104.7s, Code Interpretation at 126.7s (the longest), Result Description at 94.9s, Insight at 99.5s, Conclusion at 73.5s (the shortest), Transition at 105.6s, Direction at 118.3s, and Question at 77.4s.
• Figure 2: The overview of NotePlayer presents two primary components: interactive modules and computational engines. The interactive modules provide intuitive controls and real-time feedback, while the computational engines, serving as the backbone of NotePlayer, process data and generate outputs for the interactive modules with assistance from language models.
• Figure 3: The Interactive Modules of NotePlayer comprise (A) the Logic Flow Representation, (B) the Organization Panel, (C) the Emphasis Record, and (D) Scene Playback. The Organization Panel features crucial elements: (B1) code explanation and (B2) narration section, along with (B3) detailed notebook information, all essential components of the content creation process.
• Figure 4: Steps for conducting emphasis: (1) select code snippet, (2) click the button, and (3) add annotation in the pop-up window.
• Figure 5: The narration presentation includes three key types: 1) transition, 2) direction, and 3) question. Users can click the Q button to transform the sentence from declarative to an interactive question-and-answer format.