Intelligent Decision Support System for Dementia Care Through Smart Home
K. Gayathri, K. S. Easwarakumar · Procedia Computer Science · 2016
As per the statistics of WHO (World Health Organization), major percentages of aged society across the globe are affected with memory related disease - dementia. The percentage of dement people would be doubled in future and hence assistive health care systems have become predominant. Smart home, an ubiquitous environment offers ambient assisted living to its occupant through activity recognition and decision making process. This research work proposes an assistive dementia care environment through smart home that aid mentally disabled people with many different types of assistance during emergency. The proposed system models “Intelligent Decision Support System” that identifies deviation of the occupant from their regular activities of daily routines and decides on appropriate alerts to handle these situations. Significant criterion to model dementia care is to handle incomplete event sequences (produced due to memory loss) and to model occupant specific knowledge (provided by the care taker / doctors). Markov Logic Network (MLN), an approach of statistical relational learning models uncertain data and domain knowledge within a single framework. Thus, the proposed approach of decision support system for dementia care effectively utilizes MLN for its modeling. The experimental study made with smart home dataset showcased the competence of MLN approach of decision making has higher F-measure than existing approaches.