Chronobiology applied to the development of human behavior computational models

Abstract

This paper proposes CHROMUBE. It is a methodology for the creation of computational models of human behavior. That kind of simulated model has a number of applications in the smart environments and ubiquitous computing domains. For example, an interesting application is using these models for testing services and applications in smart environments (e.g. a fall detection service for elders living independently in their own house, based on sensors deployed at home). The methodology uses techniques from Chronobiology (i.e. actograms, plexograms and cyrcadian rhythms detection) to analyze real data about the target subject's behavior. The analysis is intended to create a simulation model of the human starting from the result of such analysis. The effectiveness of the procedure is demonstrated with a running example: a real system that monitors elderly people at home. The application of CHROMUBE in such context had an influence on the pilot experience used to test the system. Thanks to CHROMUBE, its duration was reduced and the pilot more focused on the typical reliability and dependability aspects of a system that must work 24 × 7.