Home // SOFTENG 2026, The Twelfth International Conference on Advances and Trends in Software Engineering // View article

Knowledge Reuse in ML Prototyping: Insights from an Interview Study

Authors:
Selin Coban
Patrick Chrestin
Horst Lichter

Keywords: Knowledge Reuse; Prototyping; Machine Learning

Abstract:
Prototyping is a core activity in Machine Learning (ML) solution development, yet research on reuse in this context has largely focused on source code, overlooking more comprehensive forms of knowledge that ML developers rely on. This paper reports on a qualitative interview study with 18 ML developers from academia and industry that investigates how knowledge is sought, evaluated, reused, and retained during ML prototyping. Our thematic analysis reveals that ML developers reuse not only code but also declarative knowledge (e.g., terminology and foundational concepts) and procedural knowledge (e.g., design patterns and instructions). Search strategies are opportunistic and context-driven. Reuse decisions are based on quality, intuition, and direct experimentation. Notably, retention practices are often unsystematic, relying on memory or unstructured notes, which hinders the effective transfer of knowledge across projects. With this paper, we contribute a more structured perspective on knowledge reuse in ML prototyping. We further highlight the need for dedicated tools to support knowledge management in ML solution prototyping and suggest directions for systematically retaining this knowledge to increase its reuse.

Pages: 1 to 8

Copyright: Copyright (c) IARIA, 2026

Publication date: May 24, 2026

Published in: conference

ISSN: 2519-8394

ISBN: 978-1-68558-398-9

Location: Venice, Italy

Dates: from May 24, 2026 to May 28, 2026