Goal Processing and Semantic Matchmaking in Opportunistic Activity and Context Recognition Systems
Sprache des Vortragstitels:
Englisch
Original Tagungtitel:
The 9th International Conference on Autonomic and Autonomous Systems (ICAS2013)
Sprache des Tagungstitel:
Englisch
Original Kurzfassung:
The Opportunistic Sensing Paradigm shifts away the configuration of activity- and context recognition systems from design time of the system to a dynamic runtime configuration. Systems following this new paradigm have to autonomously configure themselves during runtime according to a stated recognition goal and the available sensing infrastructure. A crucial point therefore is to find a methodology to express the recognition goal itself and to map the requirements specified by the stated recognition goal to the capabilities of the available sensors in the ecosystem of the users. This paper presents an approach for a semantic engine capable of processing and matching a given recognition goal to the semantically described sensing infrastructure. This matching allows to autonomously spot and quantify the best set of sensors available at any point in time that can contribute to the stated recognition goal and therefore being configured to a sensing ensemble.