# An Ontology-Based Expert System Approach for Hearing Aid Fitting in a Chaotic Environment

## TL;DR

This paper introduces DHAFES, an offline expert system for hearing aid fitting that uses an ontology to provide personalized recommendations in low-resource settings.

## Contribution

The novel contribution is DHAFES, an ontology-based expert system for hearing aid fitting that operates offline and supports multilingual self-assessment.

## Key findings

- DHAFES supports 33 core complaint classes and ensures transparent and traceable recommendations.
- The system improves accessibility in resource-limited environments by operating offline and remotely.
- Its ontology-based design allows adaptation to diverse clinical contexts and future AI integration.

## Abstract

Background/Objectives: Hearing aid fitting is critical for hearing loss rehabilitation but involves complex, interdependent parameters, while AI-based technologies offer promise, their reliance on large datasets and cloud infrastructure limits their use in low-resource settings. In such cases, expert knowledge, manufacturer guidelines, and research findings become the primary sources of information. This study introduces DHAFES (Dynamic Hearing Aid Fitting Expert System), a personalized, ontology-based system for hearing aid fitting. Methods: A dataset of common patient complaints was analyzed to identify typical auditory issues. A multilingual self-assessment questionnaire was developed to efficiently collect user-reported complaints. With expert input, complaints were categorized and mapped to corresponding hearing aid solutions. An ontology, the Hearing Aid Fitting Ontology (HAFO), was developed using OWL 2. DHAFES, a decision support system, was then implemented to process inputs and generate fitting recommendations. Results: DHAFES supports 33 core complaint classes and ensures transparency and traceability. It operates offline and remotely, improving accessibility in resource-limited environments. Conclusions: DHAFES is a scalable, explainable, and clinically relevant solution for hearing aid fitting. Its ontology-based design enables adaptation to diverse clinical contexts and provides a foundation for future AI integration.