Measuring Completeness of Ontology-to-Ontology Mappings: A Case Study on Transforming IMF to the Industrial Data Ontology

Topic Description 
This thesis investigates the problem of assessing and ensuring completeness of ontology-to-ontology mappings. The case study focuses on transformations from the Information Modelling Framework (IMF), represented as an OWL ontology, to the Industrial Data Ontology (IDO). The current transformation is implemented using SPARQL CONSTRUCT rules, but alternative approaches, such as OTTR templates, could also serve as mapping mechanisms. 

A key research problem is to measure the completeness of the mapping, i.e., when it covers all possible or legal inputs expressible in the source metamodel, and produces outputs that are both semantically valid and faithful to the target ontology. The thesis will develop formal criteria and practical evaluation methods for measuring coverage and correctness of ontology mappings. 

By analyzing IMF–IDO mappings, the work could also explore the trade-offs between rule-based (SPARQL) and template-based (OTTR) approaches, evaluate their completeness, and propose techniques for detecting gaps, redundancies, and inconsistencies. 

Research Questions 

1. What does mapping completeness mean in the context of ontology-to-ontology transformations? 

1. Coverage of source ontology constructs? 

2. Conformance with target ontology constraints? 

3. Preservation of semantics (no loss of information)? 

2. How can completeness of a mapping be measured systematically? 

1. Structural coverage (all classes, properties, constraints handled) 

2. Semantic coverage (all legal combinations of constructs handled) 

3. Instance-level testing (can all valid source instances be transformed?) 

3. How do SPARQL CONSTRUCT rules compare with OTTR templates as mapping formalisms in terms of: 

1. Readability and maintainability 

2. Guaranteeing completeness 

3. Extensibility for future updates 

4. What methods (test case generation, reasoning, coverage analysis) can be used to detect missing or redundant mappings? 

5. How can the findings generalize beyond IMF–IDO to other ontology-to-ontology transformation scenarios? 

Methodology 

1. Background & Literature Review – Study ontology mapping frameworks, completeness criteria, coverage metrics, and validation methods. 

2. Mapping Implementation – Develop mappings from IMF to IDO using both SPARQL CONSTRUCT rules and OTTR templates. 

3. Formalization of Completeness – Define criteria for mapping coverage (syntactic, semantic, instance-level). 

4. Evaluation Techniques: 

1. Structural coverage analysis (ensuring all IMF classes/properties are mapped). 

2. Reasoning-based validation (checking if mapped data satisfies IDO constraints). 

3. Test data generation (systematic creation of legal IMF instances to test mapping exhaustiveness). 

5. Comparison of Mapping Formalisms – Analyze advantages/disadvantages of SPARQL vs OTTR for maintaining completeness. 

6. Case Study Execution – Apply methods to selected parts of the IMF and IDO ontologies. 

7. Synthesis – Provide practical guidelines and metrics for measuring and improving mapping completeness. 

Expected Contributions (suggestions) 

• A formal definition of completeness for ontology-to-ontology mappings. 

• Methods for measuring and validating completeness of mappings (structural + semantic). 

• A comparative analysis of SPARQL CONSTRUCT and OTTR template approaches for ontology mappings. 

• A case study on IMF-to-IDO transformation, including gap analysis and completeness evaluation. 

• Generalizable guidelines for ontology engineers working with industrial data interoperability.