Predicting the understandability of computational notebooks through code metrics analysis
Published in Empirical Software Engineering, 2025
This paper investigates how various code metrics can be utilized to evaluate and predict the understandability of computational notebooks. By analyzing these metrics, the study provides actionable insights for improving the readability, maintenance, and overall design of data science workflows.
An earlier version is available as a 2024 arXiv preprint. The preprint includes Masih Beigi Rizi as an additional coauthor; the author list above follows the published journal version.
Recommended citation: Ghahfarokhi, M. M., Asadi, A., Asgari, A., Mohammadi, B., & Heydarnoori, A. (2025). "Predicting the understandability of computational notebooks through code metrics analysis." Empirical Software Engineering, 30, 98.
Download Paper