Master theses

Ontology change is unavoidable. Domain concepts evolve, regulations change, and modeling decisions are refined over time. Yet while software engineering has mature practices for managing change — semantic versioning, CI/CD pipelines, release artifacts, and automated database migrations — ontology engineering often lacks comparable operational tooling. Many ontology updates are still handled manually, with limited impact analysis and little automation for downstream consumers.

This gap has real consequences. Knowledge Graphs, data integration pipelines, validation rules, and APIs often depend on ontology terms and constraints. When an ontology changes, dependent systems can silently drift out of sync, causing broken queries, invalid data, and costly maintenance. Without structured release and migration mechanisms, teams either postpone ontology evolution or absorb high migration overhead each time a new version is published.

Recent work such as this article highlights the need for systematic ontology evolution strategies, but significant opportunities remain in turning those ideas into practical engineering workflows. The missing piece is an ontology catalogue approach that treats ontologies as living software artifacts: versioned, diffable, releasable, and accompanied by machine-actionable migration guidance.

In this thesis, you will design and prototype a framework for operational ontology evolution that supports change dissemination, release publication, and automated migration assistance for dependent data systems. Your objective is to reduce the cost and risk of ontology updates while preserving interoperability across versions.

Concretely, you will analyse ontology update strategies (e.g., backward-compatible vs breaking changes), define a change taxonomy, and model how different change types affect consumers such as datasets, SHACL shapes, mappings, and query templates. You will then design an ontology catalogue system that publishes versioned releases, exposes structured change logs, and generates migration suggestions or scripts that downstream systems can use.

You will evaluate your approach on three dimensions: change traceability (can updates be understood and communicated clearly), migration support quality (are generated migration artifacts useful and correct), and operational impact (does the framework reduce update friction for data teams). The expected outcome is a practical blueprint for ontology lifecycle management that brings software-grade DevOps principles to Semantic Web engineering and helps organisations evolve ontologies safely at scale.

The expected result is a prototype ontology-catalogue workflow with automated migration assistance for selected change classes, not a full enterprise governance platform.