Detection of User Mode Shift in Home

A ubiquitous environment enable us to enjoy various services “anytime” ”anywhere”. However, “everyone” is not realized. We research an intelligent space “everyone” can enjoy services. This paper proposes a method to detect user behavior to provide services according to user context in home. We focus on scenes user’s mode significantly changes, such as going out and going to bed. People often have characteristic behavior in these scenes. Our method extracts this characteristic as a behavioral pattern and detects user behavior in these scenes by matching current user behavior online with it. The method characterizes each scene with kind of objects a user touched and the order of them. The method realizes early start of providing services by creating a behavioral pattern from user behavior logs in short duration. The experiment proves the high potency of our method and discusses its weakness at the same time.

• Type: Conference
• Year: 2007
• Where: UCS
• Authors: Hiroyuki Yamahara, Hideyuki Takada, Hiromitsu Shimakawa
• Keywords: Applied Computing, Scenes User, UCS 2007, User Behavior, User Behavior Online