An Approach for Incorporating End User Knowledge into Applications of Information Extraction Classification and Fusion in Institutional Eldercare

Abstract

Sensor-based smart spaces have been developed to help caring of the elderly. Selection of sensors, layout of smart space, design of fusion algorithms, and knowledge of domain experts are critical to the development of a useful ICT system for caring the elderly in the institutional facilities. We apply multi-modal sensor fusion technique to improve the confidence level of our system, and leverage the knowledge of domain experts, which is stored in ontology database, to detect the sequence of events that can lead to dangerous situations.