Building Domain-Specific Machine Learning Workflows - A Conceptual Framework for the State-of-the-Practice

Aug 4, 2026

Domain experts increasingly apply machine learning to address domain-specific problems, yet translating expertise into executable computational workflows remains challenging. This article introduces a conceptual framework capturing the route of transformations a domain expert may take, and identifies six key challenges faced when moving from problem definition to implementable workflow systems. It grounds the framework in the current state-of-the-practice by examining textual and graphical workflow systems, classifying tool support and domain-specificity, and highlighting gaps in automation and transformation coverage, while proposing research directions to bridge software engineering and scientific domains.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages50

Reading Time2 h 6 min

AccessFree to download

Introduction

Conceptual framework and route of transformations

Six key challenges in domain expert workflows

State-of-the-practice workflow systems and support

Research directions and automation of tools

Titre: Building Domain-Specific Machine Learning Workflows: A Title: Conceptual Framework for the State-of-the-Practice Auteurs: Bentley Oakes, Michalis Famelis, & Houari Sahraoui Authors: Date: 2024 Type: Article de revue / Article Oakes, B. , Famelis, M. , & Sahraoui, H. (2024) . Building Domain-Specific Machine Référence: Learning Workflows: A Conceptual Framework for the State-of-the-Practice. ACM Citation: Transactions on Software Engineering and Methodology, 33(4), 1-50. [https://doi.org/10.1145/3638243](https://doi.org/10.1145/3638243) | Document en libre accès dans PolyPublie
Open Access document in PolyPublie
| | | | | --- | --- | --- | --- | | URL de PolyPublie:
PolyPublie URL: | [https://publications.polymtl.ca/57010/](https://publications.polymtl.ca/57010/) | | | | Version: | Version finale avant publication / Accepted version Révisé par les pairs / Refereed | | | | Conditions d’utilisation:
Tous droits réservés / All rights reserved
Terms of Use:
| | | | | Document publié chez l’éditeur officiel
Document issued by the official publisher
| | --- | | Titre de la revue: ACM Transactions on Software Engineering and Methodology (vol. 33, Journal Title: no. 4)
Maison d’édition:
Publisher: Association for Computing Machinery URL officiel:
Official URL: [https://doi.org/10.1145/3638243](https://doi.org/10.1145/3638243)
Mention légale: © 2024 Copyright held by the owner/author(s) . Publication rights licensed to ACM.
Legal notice:
| Ce fichier a été téléchargé à partir de PolyPublie, le dépôt institutionnel de Polytechnique Montréal This file has been downloaded from PolyPublie, the institutional repository of Polytechnique Montréal [https://publications.polymtl.ca](https://publications.polymtl.ca) Building Domain-Specific Machine Learning Workflows: A Conceptual Framework for the State-of-the-Practice BENTLEY JAMES OAKES, Département de génie informatique et génie logiciel, Polytechnique Montréeal, Canada MICHALIS FAMELIS, Département d’informatique et de recherche opérationnelle, Université de Montréal, Canada HOUARI SAHRAOUI, Département d’informatique et de recherche opérationnelle, Université de Montréal, Canada Domain experts are increasingly employing machine learning to solve their domain-specific problems. This article presents to software engineering researchers the six key challenges that a domain expert faces in addressing their problem with a computational workflow, and the underlying executable implementation. These challenges arise out of our conceptual framework which presents the “route” of transformations that a domain expert may choose to take while developing their solution. To ground our conceptual framework in the state-of-the-practice, this article discusses a selection of available textual and graphical workflow systems and their support for the transformations described in our framework. Example studies from the literature in various domains are also examined to highlight the tools used by the domain experts as well as a classification of the domain-specificity and machine learning usage of their problem, workflow, and implementation. The state-of-the-practice informs our discussion of the six key challenges, where we identify which challenges and transformations are not sufficiently addressed by available tools. We also suggest possible research directions for software engineering researchers to increase the automation of these tools and disseminate best-practice techniques between software engineering and various scientific domains. CCS Concepts: • Computing methodologies → Machine learning; Knowledge representation and reasoning. Additional Key Words and Phrases: computational workflow, workflow composition, domain experts, machine learning, machine learning pipelines, software engineering framework ACM Reference Format: Bentley James Oakes, Michal

FAQs

What does the article propose for building domain-specific machine learning workflows?

It proposes a conceptual framework that models the route of transformations a domain expert can take when developing a computational workflow and its executable implementation.

What are the six key challenges discussed in the paper?

The paper identifies six key challenges a domain expert faces in turning their problem into a computational workflow, and it examines which challenges and transformations existing tools do not sufficiently address.

How does the paper evaluate the state-of-the-practice?

It reviews and analyzes available textual and graphical workflow systems, studies examples from the literature across domains, and uses these to assess tool support and categorize domain-specificity and machine learning usage.

Aug 4, 2026

Upload by Emma Wilson