Bringing complex event processing into multitree modelling of sensors

Abstract

The recent advances in the Internet of Things allow deploying a large variety of applications for smart cities, home automation or the industry of the future. These applications generate a large amount of data that can be challenging to manage; identifying and parsing this data become a prominent problem. In order to address this issue, we propose a multitree model for the sensor representation which matches the need for heterogeneous applications and user support. From there, we define a complex event processor based on a new language and grammar, in order to filter and identify user specific events. We show that we considerably reduce the size of queries by focusing on end-users knowledges as semantics for data streams.