Related Work — ChE Curriculum KG / OBE Analytics

Comparison table

Where the gap sits

None of the above combine all four of:

1. Formal OBE/accreditation structure (CLO → PLO, Bloom's, Washington Accord / PEC / ABET)
2. Semantic, topic-level extraction validated against human raters (κ)
3. A specific engineering discipline (not generic, not CS-only)
4. Programme-level analytic findings (overlap/redundancy, Bloom's progression) as the output — not personalization/recommendation, not matrix arithmetic

Sharpest unclaimed angles:

• Extraction input ablation: measuring signal loss when using CLOs-only vs. CLOs + full course outlines (Jaccard similarity, topic count deltas) — nobody has studied this explicitly.
• Human validation of automated topic extraction: existing curriculum-KG papers validate graph structure (modularity, centrality), not agreement with human raters; existing CLO-PLO papers don't extract topics at all. The κ=0.75 benchmark against teacher voting is unclaimed.
• Redundancy/overlap as a derived finding: existing OBE tools only answer "is coverage numerically balanced." Nobody visualizes or quantifies where content actually repeats across courses, or how cognitive demand (Bloom's) shifts semester-to-semester as an emergent, visualized result rather than a manually-tagged cell.

To-Do: Ontology / Domain-Model Papers to Read Properly

Read these specifically for their class/module structure — useful for grounding or contrasting against the Knowledge Area / Knowledge Unit hierarchy in the ChE Ontology view.

• OntoCAPE (Marquardt, Morbach, Wiesner et al.) — OntoCAPE: A Re-Usable Ontology for Chemical Process Engineering (Springer book) and the shorter ScienceDirect overview papers. Focus on Ch. 7–9 (material/chemistry concepts, process system concepts, mathematical models) and the modular/layered structure discussion (Ch. 10 design principles: coherence, conciseness, intelligibility, adaptability, minimal ontological commitment, efficiency). This is the most likely direct citation for "prior formal ChE domain ontology."

  • https://link.springer.com/book/10.1007/978-3-642-04655-1
  • https://www.sciencedirect.com/science/article/abs/pii/S0952197606001199

• OntoKin — combustion chemistry / kinetic reaction mechanisms ontology. Check scope note: reaction mechanism, phase, chemical reaction, rate coefficient, chemical species modules.

  • https://pubs.acs.org/doi/10.1021/acsomega.0c02055

• OntoRXN (Garay-Ruiz & Bo, 2022) — reaction network KG, built on OntoCAPE + OntoCompChem, undirected graph treatment of reaction networks, SPARQL query examples.

  • https://pmc.ncbi.nlm.nih.gov/articles/PMC9153116/

• KG framework for digital twins of chemical processes (Nature Chemical Engineering, 2026) — most recent OntoCAPE-adjacent work, LLM agents assisting KG construction. Worth reading for how they frame LLM+ontology integration — directly relevant to your Ollama layer + normalization dictionary approach.

  • https://www.nature.com/articles/s44286-026-00392-1

• "An Ontology for Representing Curriculum and Learning Material" (arXiv 2506.05751) — uses ACM/IEEE-CS Knowledge Area → Knowledge Unit → Learning Outcome as its backbone. Read closely: this is the closest existing precedent for your Knowledge Area/Unit terminology decision — check whether to cite it directly as justification for the naming choice.

  • https://arxiv.org/pdf/2506.05751

• LLM-Assisted Knowledge Graph Completion for Curriculum and Domain Modelling (arXiv 2501.12300) — Topic/Sub-Topic ontology classes, LLM-assisted fine-grained topic extraction with human collaboration, evaluated via graph modularity + degree centrality. Compare their evaluation methodology against your κ approach — good contrast material for the "validation gap" argument.

  • https://arxiv.org/html/2501.12300v1

• PALM: Panoramic Learning Map (arXiv 2507.18393) — cites Harden 2001 AMEE Guide and Wager 1978 on curriculum mapping functional relationships; check its reference list for older curriculum-mapping foundational sources worth pulling into your own related-work section.

  • https://arxiv.org/pdf/2507.18393

• CyBOKClaw (arXiv 2605.24663) — human-in-the-loop mapping of syllabus/lecture topics to a formal Body of Knowledge (CyBOK). Structurally close to what a "ChE Body of Knowledge" alignment might look like if you extend beyond your own extracted taxonomy toward an external standard.

  • https://arxiv.org/pdf/2605.24663

• Chemical species ontology for data integration (Pascazio, Rihm, Naseri, Mosbach, Akroyd, Kraft, 2023, J. Chem. Inf. Model.) — referenced inside the tmQM-RDF paper; check if it's a useful smaller-scope citation alongside OntoCAPE for chemical species representation specifically.

• ACM/IEEE-CS/AAAI Computer Science Curricula (the source of the Knowledge Area / Knowledge Unit / Learning Outcome framework) — not primarily an "ontology" paper but the origin of the terminology you're adopting. Worth reading the actual curricula guidelines document (not just secondary citations of it) to cite it directly rather than second-hand.

Notes while reading (fill in during office review)

• OntoCAPE scope vs. your scope — where does subject-matter ontology end and curriculum-topic ontology begin? Is there a clean line, or should ChE Ontology view eventually link to OntoCAPE classes rather than reinvent them?
• Does arXiv 2506.05751's KA/KU/LO ontology have a public schema/OWL file worth structurally comparing against data.json?
• Check if any of these papers report inter-rater agreement numbers at all (for comparison against κ=0.75 benchmark) — note here if found: ___