The Story in the Notebook - Exploratory Data Science Using a Literate Programming Tool

Explore the challenges and behaviors of data scientists using literate programming tools in this 21-minute conference talk from the ACM CHI Conference on Human Factors in Computing Systems. Dive into research findings from interviews with 21 data scientists and a survey of 45 professionals, uncovering insights on coding practices, version control, and narrative creation in notebook environments. Learn about the difficulties in maintaining detailed experimentation histories and how data scientists curate their notebooks into narratives. Gain valuable design guidance for future literate programming tools, including suggestions for history search functionality based on contextual details. Understand the importance of balancing readability with comprehensive version tracking in data science workflows.

Intro
What is Data Science
Programming for Data Science
Challenges of Data Science
Literate Programming
Behavioral Evidence
Research Questions
Keeping a Record
Methodology
Results
Narrative Matters
Scratch Pads
Annotations
Expand Reduce
Reduce
Participants
Main takeaway
Experiment history design
Key takeaways
Questions