Sequential pattern profiling based bio-detection for smart health service

Abstract

Due to the development of IT convergence technologies, increased attention has focused on smart health service platforms to detect emergency situations related to chronic disease, telemedicine, silvercare, and wellness. Moreover, there is a high demand for technologies that can properly judge a situation and provide suitable countermeasures or health information if an emergency situation occurs. In this paper, we propose the sequential pattern analysis based bio-detection for smart health services. A smart health service platform is able to save bio-images and their locations detected in a smart health surveillance area where CCD cameras are installed. When a person's figure is saved, the route tracing detects any movement and then traces its location. In addition, the platform analyzes the perceived bio-images and sequential patterns in order to determine whether or not the emergency situation is normal. Using AprioirAll algorithm-based sequential pattern profile analysis, bio-detection can detect a user who is undergoing an emergency based on abnormal patterns. It performs this task by managing information obtained from data and trace analyses, and it starts bio-detection only when there are patterns not conforming to sequential patterns. In other words, bio-detection detects the maximum sequence that can satisfy the minimum support in a given transaction. Sequential pattern profile analysis based on life-logs can analyze normal and abnormal profiles to provide health guidelines.