Improvement of behavior detection by dynamic threshold
Hiroyuki Yamahara, Fumiko Harada, Hideyuki Takada, Hiromitsu Shimakawa · 2007
We are developing a context-aware application for use in homes, which provides services according to user behavior proactively, by detecting high-level user behavior such as leaving the home. For the detection, a behavioral pattern is created by extracting frequent characteristics from the user's behavior logs acquired from sensors online, using an extraction threshold based on the criterion of frequency. A conventional model determines a fixed threshold value common to all users. However, proper values vary with user. This paper proposes a detection method using a model which dynamically determines the threshold value for individual behavioral pattern.