Life Supporting Technologies (LifeSTech), Superior Technical School of Telecommunication Engineers, Universidad Politécnica de Madrid (UPM), Ciudad Universitaria, 28040 Madrid, Spain.
Sensors (Basel). 2021 Dec 27;22(1):166. doi: 10.3390/s22010166.
Utilizing context-aware tools in smart homes (SH) helps to incorporate higher quality interaction paradigms between the house and specific groups of users such as people with Alzheimer's disease (AD). One method of delivering these interaction paradigms acceptably and efficiently is through context processing the behavior of the residents within the SH. Predicting human behavior and uncertain events is crucial in the prevention of upcoming missteps and confusion when people with AD perform their daily activities. Modelling human behavior and mental states using cognitive architectures produces computational models capable of replicating real use case scenarios. In this way, SHs can reinforce the execution of daily activities effectively once they acquire adequate awareness about the missteps, interruptions, memory problems, and unpredictable events that can arise during the daily life of a person living with cognitive deterioration. This paper presents a conceptual computational framework for the modelling of daily living activities of people with AD and their progression through different stages of AD. Simulations and initial results demonstrate that it is feasible to effectively estimate and predict common errors and behaviors in the execution of daily activities under specific assessment tests.
在智能家居 (SH) 中使用上下文感知工具有助于在房屋和特定用户群体(如阿尔茨海默病 (AD) 患者)之间融入更高质量的交互模式。通过上下文处理 SH 内居民的行为来以可接受且高效的方式提供这些交互模式的一种方法。预测人类行为和不确定事件对于预防 AD 患者进行日常活动时出现的未来失误和困惑至关重要。使用认知架构对人类行为和心理状态进行建模,可以产生能够复制实际用例场景的计算模型。通过这种方式,一旦 SH 对认知能力下降患者日常生活中可能出现的失误、中断、记忆问题和不可预测事件有足够的了解,它们就可以有效地加强日常活动的执行。本文提出了一种用于建模 AD 患者日常生活活动及其在 AD 不同阶段进展的概念计算框架。模拟和初步结果表明,在特定评估测试下,有效地估计和预测日常活动执行过程中的常见错误和行为是可行的。