A Dynamic Normalized Weighting Based Context Reasoning in Home-Based Care

Some types of contextual information are more important than others for inferring a situation of the patient in home-based care. The weight of this information may be changed due to the variations of the sensed values over time. Some researches applied a static weighting based fusion process to their systems to model these variations. However, this fusion process sometimes reduces the reliability of contextual information, since it does not consider a dynamic change of the importance over time. We propose the dynamic weighting based evidential fusion process (DWEFP) as a context reasoning method. DWEFP applies a normalized weighting technique to dynamic evidential networks for improving the confidence level of contextual information. To show the improvement of the proposed method, we compare our DWEFP with previous static weighting based fusion methods such as Bayesian Networks (BNs), Dempster-Shafer Theory (DST), Static Evidential Fusion Process (SEFP), Dynamic Evidential Fusion Process (DEFP), and Dynamic Bayesian Networks (DBNs).