Introduction
As Data Science (DS) continues to solidify its place in both academia and industry, the definition of its "computing" component has remained somewhat fluid. While the interdisciplinary nature of the field—blending statistics, mathematics, and computer science—is widely accepted, the specific technical competencies required for a computer science-based DS degree have lacked a standardized framework. To address this, the ACM Data Science Task Force has been developing a formal set of curricular recommendations aimed at defining the computing-specific knowledge required for undergraduate practitioners.
Defining the Computing Core
The primary challenge addressed by the Task Force is the lack of granular guidance for departments building DS programs. Many existing guidelines offer high-level overviews but fail to specify the rigorous computing foundations necessary to distinguish a professional data scientist from a casual user of data libraries. By establishing a "Knowledge Area" (KA) framework, the Task Force seeks to move beyond "coding for data" toward a curriculum that emphasizes algorithmic design, big data systems, and sophisticated AI competencies.
Methodological Approach
The development of these recommendations relied on a three-pronged methodology:
- Comparative Analysis: The Task Force reviewed international benchmarks, including the European EDISON project and reports from the National Academies, to identify existing gaps in DS education.
- Community Surveys: In 2018, the group conducted surveys across both academic and industrial sectors. These surveys were designed to distinguish between "required" and "elective" computing concepts based on real-world demand.
- Competency-Based Design: Knowledge was organized into nine distinct areas. For each area, the Task Force defined specific, measurable competencies—skills that students must demonstrate to prove mastery.
The Nine Knowledge Areas of Data Science
The Task Force identified nine core Knowledge Areas (KAs) that constitute the computing foundation of a Data Science curriculum:
- Programming, Algorithmic Design, and Development: The fundamental ability to create efficient code.
- Data Acquirement and Governance: The logistics of sourcing and managing data.
- Data Management, Storage, and Retrieval: Database systems and data architecture.
- Data Privacy, Security, and Integrity: Ensuring the safety and reliability of data.
- Machine Learning: The theoretical and practical application of predictive models.
- Data Mining: Discovering patterns in large datasets.
- Systems for Big Data: High-performance computing (HPC) and large-scale data processing.
- Professionalism: Ethics, collaboration, and professional decision-making.
- Analysis and Presentation: Visualization and Human-Computer Interaction (HCI).
The Role of Machine Learning and Artificial Intelligence
A central pillar of the proposed curriculum is the treatment of Machine Learning (ML) as a multi-faceted discipline. Rather than treating ML as a single skill, the Task Force breaks it down into 12 distinct competencies. These range from the ability to derive algorithms from first principles to understanding the "curse of dimensionality" and evaluating model interpretability.
Furthermore, the Task Force argues that AI in Data Science extends far beyond ML. Subfields such as Natural Language Processing (NLP), ontologies, and Human-Computer Interaction are becoming increasingly vital. The curriculum also emphasizes "Data Science in Context," noting that technical skills must be applied to external domains—such as biology, economics, or social sciences—to be truly effective.
Ethics as a Technical Foundation
One of the most significant shifts in this framework is the treatment of ethics. Rather than being relegated to a "soft skill" or an elective, professionalism and ethics—including the mitigation of algorithmic bias—are integrated as core technical competencies. This reflects the growing industry and academic consensus that ethical data handling is inseparable from technical proficiency.
Conclusion and Future Directions
The current framework represents a "work in progress" intended to spark dialogue within the AI and computing communities. While the focus is strictly on the computing contributions to Data Science, the authors acknowledge that a comprehensive curriculum must also integrate deep dives into statistics and mathematics. Future iterations of the report will likely refine the roles of emerging subfields, such as the semantic web and speech processing, to ensure the curriculum remains relevant in a rapidly evolving technological landscape.
Summary updated