Ontology-based Diagnosis and Recommendation Model for Common Mental Disorders

Abstract

In the developing world, there is a high prevalence of common mental disorder such schizophrenia, generalized anxiety disorders, generalized depressive and other related disorders. We have created an ontology-based diagnosis and recommendation model to increase services and help the psychiatrist expert and people. To create domain-based knowledge, we applied Semantic Web Technology. Domain expert and other expert-written materials were used to acquire data for domain ontology-based knowledge building. The concepts, relationships, and qualities of this ontology model can be utilized in the approach to diagnosis and recommendation of common mental disorders. In order to determine the type of mental disorder based on the symptoms in the domain ontology, the SWRL rules are developed using legitimate relationships between ontology concepts. Our study was carried out using the Design Science research methodology. Using protégé editor, we modeled the acquired knowledge as an ontology-based architecture, and pellet reasoner was used to assess the consistency of the domain ontology we had constructed. We may search the domain ontology for treatments, drugs, advice, and other pertinent information. We have used the sample of 34 patients’ symptoms with common mental disorders diagnosed by domain experts from Jimma University Referral Hospital. The result shows 32 patients are correctly diagnosed by our developed model using SWRL rule (the correctness ratio is 94.1%). All the recommended information for each CMD was taken from domain experts and developed in our domain ontology model. Its performance was evaluated by non-psychiatrist and psychiatrist experts and the performance was 85.3%. The overall performance of our developed domain ontology model was average 89.7%.