A fuzzy logic based context-aware reminder for elders with mild dementia
Résumé
Purpose : Dementia is a progressive, disabling, chronic disease affecting 5% of all persons above 65 years old and over 40% of people above 90. In Europe, about 1.9 million people experience mild dementia. The term dementia refers to a combination of symptoms involving impairments of memory, thought, speech, perception and reasoning. Early impairments in performing complex tasks lead to an inability to perform even the most basic functional activities such as washing and eating. The objective of this paper is to apply pervasive technologies to support elders with mild dementia to live independently. Specifically, the paper intends to develop novel solutions for providing appropriate alerting, interventions or assistance to the elders in the right time. Method : The common way to design reminding systems is to pre-define necessary ADLs into a plan and then prompt the elders according to the plan. The plan expresses the temporal constraints of various daily activities. If the subjects forget to execute those activities in the plan, the reminding system will prompt them accordingly. However, the dynamic nature of people's daily activities poses many challenges in building the reminding systems. Problems arise when the planned activities are interrupted by "disruptive" activities, as those disruptive activities that cannot be predicted in advance, thus the system has to evaluate, between the planned and disruptive activity, which has higher priority in real-time environment and then decide to either deliver, or delay, or even cancel the reminder for the planned activities. In our previous work, we proposed a general framework for designing context-aware reminders and analyzed the possible conflict among different activities. Results & Discussion : In this paper, by quantifying the interruption degree of the "disruptive" activity and the urgency of the planned activity via fuzzy logic, we formulate the context-aware reminder problem as a fuzzy decision making one. Then, through fuzzy logic reasoning, the system generates an appropriate strategy to resolve the activity conflict and infer a proper prompting time.