Keynote talk #2: Health informatics: A step forward to design an intelligent system for monitoring personal healthcare

Abstract

In recent years, the growing demand of personal healthcare has attracted much attention from both researchers and practitioners. The motivation of designing an efficient intelligent system that keeps track physical conditions, makes diagnosis based on the symptoms, and recommends appropriate treatments is the long-term objective in Health Informatics. In this talk, we will present our recent results of creating such the system using knowledge model and the Internet-of-Thing (IoT) technology. The system firstly requests the IoT device to send the physical conditions of a patient using specialized sensors namely LM35 (temperature measurement), Pluse Sensor (heartbeat) and ESP8266 (network connection). The collected personal symptoms are then integrated to a server for diagnosis of possible diseases such as viral fever, hypothermia, tachycardia and heart failure. This is done by a new technique called the intuitionistic fuzzy recommender system (IFRS), which in essence is a recommender system deployed in the intuitionistic fuzzy set for diagnosing of diseases under uncertain environments. We will present the theoretical basis of IFRS including: i) the formulation of single-criterion and multi-criteria IFRS accompanied with some essential properties; ii) a hybrid model between picture fuzzy clustering and IFRS called HIFCF; iii) intuitionistic fuzzy vector (IFV) with intuitionistic vector similarity measure (IVSM); and iv) linguistic similarity measure. Using the model, diseases are ranked according to the current symptoms and stored in the server. Information of diseases and appropriate treatment therapies is sent back to the patient as well as stored in a web portal for personal monitoring.