Computational Curriculum Studies
Navigating Pathways for Learning in the Information Age
Compared to the cultural evolution of scientific understanding, the design of general education curricula has remained relatively stagnant since the origins and spread of public schooling in the last two centuries. The rise of AI technologies presents new challenges and opportunities, not simply to how teachers develop lessons, but to the deeper underlying structures of knowledge which form the basis of the learning pathways that the majority of the world’s population are required or encouraged to engage.Â
As part of our AI innovation strategy, OpenEvo aims to support the emergence of an ethical and critical approach to the nascent field of Computational Curriculum Studies, the interdisciplinary study of the use of information technologies to describe, analyze, represent, and improve the design of learning pathways in formal education contexts.
Ethical Principles for Computational Curriculum Studies (CCS)
- CCS does not assume that computational approaches are necessarily ‘better’ than ‘traditional’ approaches to curriculum design, or even that a ‘traditional’ curriculum is in any way ‘optimal’ for human development.
- CCS is a methods development project, but methods must serve questions, not the other way around.
- CCS must focus on explainable, participatory, and open science approaches to the computationalization of curriculum design processes.
CCS has a dual emphasis on theory and practice.
CCS Theory aims to...
- Understand the shift from static, human-readable curricula to dynamic, machine-readable and human-accessible learning systems.
- Investigate how cultural evolution shapes curriculum design, balancing standardized structures with flexible, high-dimensional learning pathways.
- Understand the curriculum overload paradox by leveraging computational models to optimize cognitive load over time.
- Integrate explainable AI, participatory design, and human-in-the-loop approaches to ensure transparency and equity in educational innovation.
CCS Practice aims to...
- Develop innovative prototypes, including scientific knowledge graphs and intelligent Curriculum Agents.
- Analyze K-12 education curriculum frameworks towards richer, multi-dimensional spaces.
- Expand the role of the Nature of Science (NOS) as both a core content area and a contextual framework for inquiry-based learning.
- Construct an open ecosystem of knowledge graphs, curriculum agents, and other tools to enable adaptive, data-driven, and collaborative curriculum design.
The Curriculum Overload Paradox
Humans keep accumulating more and more knowledge. It might, therefore, make sense that the school curriculum keeps growing as well. Unfortunately, for many teachers and students, this growth in content has led to curriculum overload.Â
The rise of AI seemingly adds yet another topic to the already overloaded stack of topics that all students must study. Yet, a different perspective – both conceptually and technologically – may suggest a different interpretation.
Technology overview
We are focusing our approach on a few spaces of emerging technological potential.
Ontologies are like structured maps of ideas—they define the key concepts in a topic (like evolution, behavior, and sustainability) and how those concepts relate to one another. Knowledge graphs use these maps to connect information in smart, flexible ways, showing not just isolated facts but the rich web of relationships between them. At OpenEvo, we are working to use ontologies and knowledge graphs to model how knowledge is organized and connected—so learning can follow many different paths, adapt to different learners, and provide a conceptual richness and coherrence not possible from static curriculum representations.
Curriculum Agents are AI-powered systems that support collaboration with educators, students, and experts to analyze and refine school curricula. Leveraging knowledge graphs, large language models, and participatory AI, they make curriculum structures more dynamic, adaptable, and transparent. Rather than replacing human decision-making, they support flexible learning pathways, manage cognitive load, and enhance curriculum evolution. By integrating explainable AI and human-in-the-loop approaches, Curriculum Agents help bridge the gap between rigid, standardized curricula and real-world knowledge complexity.
New techniques in CCS, such as the use of Knowledge Graphs (KGs) to represent learning pathways, combined with the strategic use of LLMs (e.g. ChatGPT), may enable new directions in curriculum design in which, paradoxically, an increase in the complexity of curriculum (as representations of knowledge structures) may be capable of leading to a decrease in the complexity experienced by teacher and students – due to the potential to more efficiently – and more flexibly – manage cognitive load through personalized learning supports.Â
Interdisciplinary educational design research will have to clarify the potential value proposition.
Our vision for CCS at OpenEvo
To explore the opportunities and challenges described above, we are outlining a vision for technological development.Â
“The EvoMentor AI agent is powered by the OpenEvo Knowledge Graph, empowering students, teachers, and researchers with new tools for developing scientifically reflected explanations of the biological and cultural diversity of our world. From specific explanations for individuals, to the scope and sequencing of entire school curricula, EvoMentor provides expert-curated explainable AI analyses and prototyping of innovations with a precision and depth not possible by humans alone. “ – Our proposed vision
The diagram depicts the architecture of a system where the EvoMentor AI agent leverages the OpenEvo Knowledge Graph (OE-KG) to enhance science education. The OE-KG, containing rich data on trait variation, explanatory structures, conceptual structures, and learning trajectories, acts as the core knowledge base. EvoMentor utilizes this information to provide educational users with services such as scientific explanation mentoring and personalized learning pathways, as well as to assist with lesson, unit, and curriculum design. Expert users play a crucial role in contributing to and refining the OE-KG. To ensure broader applicability and interoperability, the OE-KG integrates with external ontologies. The interactions within this system provide valuable data that informs OpenEvo improvement research, driving iterative development of both the knowledge graph and the AI agent.
References
The scope of CCS is potentially vast, and this initial concept site can not yet map the full space. Below we offer a few helpful references to orient interested researchers to relevant contexts for future development.
Abu-Rasheed, H., Abdulsalam, M. H., Weber, C., & Fathi, M. (2024). Supporting student decisions on learning recommendations: An llm-based chatbot with knowledge graph contextualization for conversational explainability and mentoring. arXiv preprint arXiv:2401.08517.https://arxiv.org/pdf/2401.08517Â
Abu-Salih, B., & Alotaibi, S. (2024). A systematic literature review of knowledge graph construction and application in education. Heliyon, 10(3).
Smaldino, P. E., Russell, A., Zefferman, M., Donath, J., Foster, J., Guilbeault, D., … & Miton, H. (2024). Information Architectures: A Framework for Understanding Socio-Technical Systems (No. c7vrw). Center for Open Science. https://culturologies.co/files/IASociotechnical.pdfÂ
