Research Area 3: Ontologies & Knowledge Representation

This research area extends ontology engineering expertise developed in the geographic domain to other application areas. The work demonstrates how formal knowledge representation techniques can be applied to model complex domains, support reasoning, and enable intelligent systems.

Cross-Domain Expertise

While my primary expertise is in geographic ontologies, I collaborate with domain experts to apply ontology engineering methods to fields including mental health informatics, where formal knowledge representation can support diagnosis, classification, and analysis.

Research Themes

Biomedical Ontologies: Formal representation of medical and health-related knowledge
Ontology Reuse Patterns: Methods for integrating existing ontologies into new domains
Ontology Enrichment: Enhancing ontologies with knowledge extracted from text and data
Cross-Domain Applications: Applying ontology methods beyond the geographic domain

Biomedical Ontologies Mental Health Informatics OWL Ontology Reuse Knowledge Graphs Deep Learning NLP Text Classification

Research Projects

OCD Ontology & Online Detection

Areej Nasser Muhajab

PhD 2020–2024

An ontology-based framework for modelling Obsessive Compulsive Disorder, combining biomedical ontology engineering with NLP and deep learning for online detection from forum posts.

Interdisciplinary Collaboration

Research in this area involves close collaboration with domain experts. For the OCD ontology project, this included working with the School of Medicine at Cardiff University to ensure the ontology accurately represents clinical knowledge while being enriched with insights from patient-generated content online.

Methodological Contributions

Ontology Reuse Patterns: Strategies for full equivalence (import), partial equivalence (alignment), and nominal similarity (linking)
Hybrid Enrichment: Combining clinical resources (Bio_ClinicalBERT) with domain-specific corpora (forum posts)
Evaluation Methods: Expert surveys, SPARQL-based queries, and reasoning with use cases
Classification Integration: Linking ontological knowledge with machine learning classifiers

Resources

Ontologies developed in this research area are made publicly available to support further research:

OCD Ontology on BioPortal