Uncertainty management of situations in a housing use context

Cedric Deffo Sikounmo (1) , Stéphane Perrin (1) , Eric Benoit (1)

1 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance

This paper proposes a new contribution to manage uncertainty in the context of a situation recognition. It gives an application to the measurement of housing use situations. It extends a previous communication that proposes an ontological based model of situations. In this model, a situation is set of entities linked by relationships themselves determined using sensors. However, such model doesn’t allow to manage uncertainty. In order to take into account an uncertainty related to the identification of some entities, we propose to model the set of possible eligible entities, called the universe of discourse, by an anonymous instance in the ontology. This anonymous instance is linked with a “same as” relationships with all possible eligible entities. In addition, we propose to complete this model by the association of a mass with each “same as” relationship. This paper presents a probabilistic semantic to these masses. A practical example on a housing use situations measurement illustrates the presented approach.

Mots clés: ontology; uncertainty; home; situation detection

Domaines: Modélisation et simulation; Intelligence artificielle [cs.AI]

Cedric Deffo Sikounmo, Stéphane Perrin, Eric Benoit. Uncertainty management of situations in a housing use context. 2017 IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA) (CIVEMSA 2017), Jun 2017, ANNECY, France. pp.129-134. ⟨hal-01558391⟩