Construction of user model for an assistance system with sensor network

In recently years, human´s life became convenient because various services or assistances supplied. However, because these services are developed for major users, they are provided for users in a uniform manner. The way of service or assistant is not always appropriate, because there always exists some people who are not major users. Thus, sometimes a gap, which exists between the service and user´s thought, will appear when a system provides service to users. Hence, the author thought that a system, which is called useful, has to include the ability of understanding user, and more over, being able to change the way of service or assistance like human. Thus, if user models are acquired by a system appropriately, the system will have the ability of serving or assisting a person who is not major user appropriately. And if the user use the system, he will allow use it without learning, and only to behave naturally. In this research, the authors use 3D accelerometer sensor to obtain data and data mining to construct user model. The user models are applied in an assistance system. Here we focus on assisting user in walking and driving a car. The experiment result shows assistance system in which the authors applied user model is effective.