I.J. Information Technology and Computer Science, 2017, 10, 13-28 Published Online October 2017 in MECS (http://www.mecs-press.org/) DOI: 10.5815/ijitcs.2017.10.02 Copyright © 2017 MECS I.J. Information Technology and Computer Science, 2017, 10, 13-28 Fuzzy Based Multi-Fever Symptom Classifier Diagnosis Model Ighoyota Ben Ajenaghughrure and Dr. P. Sujatha Vels University, Department of computer science, Chennai, 600117, India Ighoyotaben@yahoo.com, suja.research@gmail.com Dr. Maureen I. Akazue Delta State University, Department of Computer Science, Abraka, 330106, Nigeria Akazuem@gmail.com Received: 08 November 2016; Accepted: 31 August 2017; Published: 08 October 2017 Abstract—Fever has different causes and types, but with similar symptoms. Therefore, making fever diagnosis with human physiological symptoms more complicated. This research project delves into the design of a web based expert multi-fever diagnosis system using a novel fuzzy symptom classifier with human self-observed physiological symptoms. Considering malaria, Lassa, dengue, typhoid and yellow fever. The fuzzy-symptom classifier has two stages. Fist stage is fever type confirmation using common fever symptoms, leading to five major fuzzy rules and the second phase is determining the level of infection (severe or mild) of the confirmed type of fever using unique fever symptoms. Furthermore, Case studies during the system implementation yielded data collected from 50 patients of having different types of fever. The analysis clearly shows the effectiveness and accuracy in the system performance through false result elimination. In addition, acceptability of the system was investigated through structured questionnaire administered to same 50 patients. This result clearly indicates that the system is well accepted, by users and considered fairly easy to use, time and cost saving. Index Terms—Fuzzy classifier, fever diagnosis, multi fever, expert fever diagnosis. I. INTRODUCTION Fever is a change in the human body temperature, both minimum and maximum [4],[5]. During this condition, a sufferer generally experiences cold and muscle seizure [3], resulting to alteration of body temperature regulation system producing more heat in a bid to sustain normal body temperature, which when restored, leads to excessive sweating[25]. Although there is discrepancy in the normal human temperature [1],[2], but fever occurs in body temperature between the range 41 to 42 °C (105.8 to 107.6 °F)[5]. The causes of fever varies from infectious to non- infectious diseases, but the case of body temperature changes also referred to as hypothermia is completely different[7]. Since it is not a result of either causative above. Treatment of both fever and hyperthermia to reduce or subside its presence is not necessary[6],[24], but a direct treatment of its associated symptoms such as muscle pains, headache etc is considered more useful[8], using common drug such as paracetamol to more intensive care methodology, depending on the severity of the sufferers health status[8],[9]. Being a common symptom of most health problem, fever is accountable for approximately 30% of children healthcare-centers visit [6] and dominates up to 75% of critically ill adults [10]. The common types of fever widely reported with scientifically available medication includes malaria, dengue, typhoid, Lassa, and yellow fever. These has great similarities in symptom’s irrespective of carriers, infection type(bacteria, virus etc) and treatment. This relationship between various types of fever symptoms, makes diagnosis of fever with human physiological symptoms is difficult, for example symptoms of Lassa is very hard to differentiate from those of malaria, yellow, dengue and typhoid fever [11],[12]. In addition to the high cost associated with acquiring wet lab fever diagnosis service, lack of medical expert availability and accessibility, has prompted the development of computer aided expert fever diagnosis systems for diagnosing single or multiple types of fever. Unfortunately, the existing computer based expert fever diagnosis systems do not take symptoms relationship into consideration, which has significant impact on the accuracy of the fever diagnosis results they produce. To overcome this challenge, a fuzzy-based multi-fever symptom classifier for fever diagnosis implemented as a web system for accessibility is proposed and designed in this research. The fuzzy-based multi-fever symptom classifier put into consideration all the various types of fever related symptoms, and accurately determine the type fever and level of infection (mild/acute) a patient is suffering from, based on user input of physiological symptoms. The implemented fuzzy-based multi-fever symptoms classifier is known as e-fever portal. Fuzzy technique derived from artificial intelligence was specifically chosen, because of its longstanding successful application