A THESIS SUBMITTED TO THE SCHOOL OF GRADUATE STUDIES OF ARBA MINCH UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN COMPUTER SCIENC

Abstract

Abstract Stroke disease is a medical condition caused due to inadequate supply of blood to the brain cell that damages the cell and may result in death. It is the third cause of death and first leading to disability. In developing country like Ethiopia death of stroke patient increase from year to year due to scarcity of specialists and health facilities. In an effort to address such a problem, this study attempts to design and develop a prototype system by integrating data mining results with knowledge-based system that can facilitate diagnosis and treatment for a patient that provides advice and risk level for a patient. different researches, journal papers and guidelines that have been done on stroke, knowledge acquisition method, and integration of data mining results with KBS were reviewed. Besides this, related works were reviewed to identify the gap and formulate research questions of the study. Mixed research design was used, integrated knowledge acquisition techniques were used to acquire knowledge, Orange and WEKA were used as hybrid data mining tool to preprocess and analyzing datasets, classification algorithms were comparatively analyzed and finally JRIP were registered better accuracy 94.16% using 10 fold cross-validation, rule-based knowledge representation approach was used to represent knowledge in the knowledge base, SWIProlog was used to construct knowledge base, Java NetBeans was employed to design GUI for the KBS, JPL library was used as a middleware between knowledge base and designed GUI. finally, the performance of the system is evaluated by preparing twenty test cases and provided to domain experts. For user acceptance test users are evaluated the system through nine criterions prepared by the researcher and the system has scored 90% system performance and 89.9% user acceptance