Representing and Reasoning about Complex Human Activities – an Activity-Centric Argumentation-Based Approach
A groundbreaking research thesis aims to revolutionize computing machinery by enabling it to identify, represent, reason, and evaluate complex human activities. Unlike current methods that focus on basic actions, this groundbreaking approach captures the intricate elements of activities directed by an individual’s needs, goals, motives, preferences, and environment. The research introduces a novel formal activity-centric model that encompasses purposeful, social, mediated, hierarchically organized, and continuously evolving interactions between people and the world.
The core contribution of this research is the development of a common-sense reasoning method based on argumentation. It provides defeasible explanations of an individual’s activities, leveraging the activity-centric model of human activity. By utilizing a semantics-based inference system, the research introduces the notion of an argument that allows the construction of structured arguments and the derivation of consistent conclusions. These structured arguments offer bottom-up explanations of complex activities through the introduction of activity fragments. By analyzing these fragments, the research generates consistent interpretations of activity that align with the activity-centric model.
To assess activities similar to how clinicians perform deductive analysis, the research introduces two quantitative measurements to evaluate performance and capacity. These measurements are formalized and analyzed using different argumentation semantics, enabling the generation of useful information for various purposes. For example, they can be utilized to detect the risk of falling in older adults or provide specific information about activity performance and completion. Ultimately, this information can be leveraged to develop intelligent machinery capable of delivering tailored recommendations for individual needs.
The contributions of this research have been implemented in proof-of-concept systems designed to evaluate complex activities and improve individuals’ health in their daily lives. Empirical evaluations were conducted with potential users, showcasing the theories and methodologies’ potential in domains such as ambient assisted living, assistive technology, activity assessment, and self-management systems for health improvement.
With its focus on developing theories and formal methods for complex human activities, this research holds immense potential for enhancing various sectors related to human well-being and quality of life. By incorporating advanced reasoning and argumentation techniques, this approach opens up new possibilities for intelligent systems, providing personalized recommendations and improving individual health outcomes. As this research continues to progress, its impact could reverberate across the global stage, contributing to a future where technology optimizes human experiences and promotes overall wellness.
