Thursday, June 20, 2024, 10am

In this proposal, I investigate the collaboration challenges between software engineers and data scientists in building machine learning (ML) products, and propose interventions to facilitate their collaboration by bridging the identified knowledge boundaries.

Despite significant advancements in ML algorithms and model development, integrating ML models into operational products remains challenging, with collaboration issues frequently cited as one of the major challenges. I identify collaboration challenges, and triangulate them with existing domain knowledge through a qualitative interview study with industry practitioners and a comprehensive meta-summary study of academic literature. I demonstrate principles or ideas of how those collaboration problems can be solved, illustrated with three interventions: (a) a novel approach for supporting data scientists and software engineers in deriving actionable model requirements, which aims to bridge gaps during the requirements elicitation process, (b) an innovative method to engage practitioners in responsible AI practices, fostering a culture of ethical awareness and compliance, and (c) a policy for guiding the development of explainable AI, ensuring transparency and understandability of ML models within products. These interventions are designed to address the syntactic, semantic, and pragmatic knowledge boundaries that hinder effective teamwork in ML product development. Lastly, I compile a comprehensive dataset of ML products from GitHub to further support research and education in the domain. The methodological approach combines various research techniques tailored to address the specific research questions in each study.

By systematically identifying and addressing collaboration challenges among practitioners, this proposal aims to support the successful development and deployment of ML products in real-world settings.

Thesis Committee:
Christian Kästner (Chair)
Jim Herbsleb
Claire Le Goues
Ken Holstein
Samir Passi (Microsoft Research)

Additional Information

In Person and Zoom Participation. See announcement.

Event Type: Thesis Proposals
Room Number: In Person and Virtual - ET
Building: TCS Hall 360 and Zoom
Speaker's Name: NADIA NAHAR
Speaker Website: sites.google.com…
Speaker's Professional Title: Ph.D. Student, Ph.D. Program in Software Engineering, Software and Societal Systems Department, Carnegie Mellon University
Talk Title: Facilitating Collaboration in Building Machine Learning Products
Event Poster Title: Poster
Event Poster URL: www.cs.cmu.edu…
For More Information: aroudebu@andrew.cmu.edu | cherold@cs.cmu.edu
Affiliations: Software and Societal Systems Department (S3D)
Organization(s): School of Computer Science