Using cognitive models for behavioral assistance of humans

Abstract Cognitive impairments are a rising challenge in society. Everybody has experienced events of forgetting what one was going to do when entering a certain room, or where one has put the keys, the purse or the smartphone. Such problems mostly occur under stress or fatigue. And they may increase with age up to diseases like senile dementia or Alzheimer which affect mental tasks like memory or reasoning, and thus often lead to the need of comprehensive assistance. Due to the societal change, the number of people suffering from such impairments is continuously growing. Cognitive modeling may provide human-centered solutions for this challenge since a cognitive model of a person's behavior, regarding activities of daily living, can serve as a knowledge base for support actions. This paper presents a Domain Specific Modeling Language for Ambient Assistance: The Human Cognitive Modeling Language (HCM-L). It was developed to preserve the episodic memory of a person in the form of conceptual behavior models including relevant context. The work is part of the Human Behavior Monitoring and Support (HBMS) project, a research project in the field of Ambient Assisted Living funded by the Klaus Tschira Stiftung gGmbH. HBMS aims at monitoring a person's behavior using activity recognition techniques, generating models from their output, and providing focused and timely support by the use of intelligent reasoning mechanisms.