User care preference-based service discovery in ubiquitous environment

Abstract

Recently, research efforts on service discovery have worked on abstract representation on functional effects. Abstract representation inherently adheres to a high level of semantics to maximize the availability of service. However, services discovered by effects are differently accepted to a user according to user``s condition like medical symptom. For example, when message needs to be delivered to a user, system discovers available services such as PDA and TV which support ``display effect`` for message delivery. However, TV which has higher sound effect as well as display effect might be better service to the user who experiences a poor eye-sight because preferred service is determined according to user``s condition. Thus, to consider user``s condition, we introduce a user care preference which is classified to multiple effects with different weight. It is used to measure how much each effect of user is sensitive to each of multiple effects which service supports. In addition, we propose a systematic process of design to support semantic service discovery with the attempts of user request classified by five senses. Symptom analysis is executed in context manger to generate user care preference and the relative degree with proximate services is measured in a service manager. Finally, when some event happens, system can discover the best service based on user``s condition. Our approach helps service discovery using the context incorporating user condition to represent user``s intention and sensitivity richer.