On the measurement of semantic reasoners in Ambient Assisted Living environments


The increasing lifetime and population of elderly people leads to a great amount of interest in Ambient Assisted Living (AAL) environments and applications. An AAL environment could be modeled by ontologies and using a semantic reasoner as a way to infer knowledge about the underlying context. An important question is to clarify the feasibility of employing a semantic reasoner in AAL applications. For this reason, in this work we present our results of a simulated AAL environment using a knowledge driven approach. Our aim is to determine and clarify which combination of semantic reasoners and frameworks is the most convenient for this task. In particular, we are interested in the combination which provides a better computational performance and reasonable memory requirements. The obtained results show that the fastest choice under the employed dataset is to use the OWL API accessing the Pellet reasoner. On other hand, the less used and committed memory consumption experiment is provided by a combination of HermiT and custom Java rules.