Abstract: As described in this paper, we propose a supervisory system that
considers actual situations and social aspects of users in a ubiquitous
computing environment. To realize gentle and safe supervision while providing
efficient supervisory services, the system must recognize the situations of a
watched person, such as the person's physical condition. To achieve this, we
have proposed a ubiquitous supervisory system "uEyes", which introduces
Social Context Awareness: a distinguishing feature for supervision. Using this
feature, the system can combine environmental information acquired from sensors
in the real world and common-sense knowledge related to human activities in
daily life. As described in this paper, we specifically examine design of
Social Context Awareness using ontology technologies. Based on this advanced
feature, a live video streaming system is configured autonomously depending on
the users' circumstances in runtime. We implemented a uEyes prototype for
supervising elderly people and performed some experiments based on several
scenarios. Based on those experimental results, we confirmed that the social
contexts are handled effectively to support the supervision.
Keywords: Supervisory system, Social Context Awareness, ontology, temporal concept
