Towards activity recognition using probabilistic description logics

A major challenge of pervasive context-aware comput- ing and intelligent environments resides in the acquisi- tion and modelling of rich and heterogeneous context data. Decisive aspects of this information are the ongo- ing human activities at different degrees of granularity. We conjecture that ontology-based activity models are key to support interoperable multilevel activity repre- sentation and recognition. In this paper, we report on an initial investigation about the application of proba- bilistic description logics (DLs) to a framework for the recognition of multilevel activities in intelligent envi- ronments. In particular, being based on Log-linear DLs, our approach leverages the potential of highly expres- sive description logics with probabilistic reasoning in one unified framework. While we believe that this ap- proach is very promising, our preliminary investigation suggests that challenging research issues remain open, including extensive support for temporal reasoning, and optimizations to reduce the computational cost.