Algorithms for user centred, problem driven automated coaching: Functional test specification and demo scenario
Year: 2013
Type: report
Abstract: In todays society, obtaining and maintaining a healthy lifestyle has become difficult for large groups of people. The easy availabilityof unhealthy food and the increasing number of (stressful) office jobs for example provide obstacles for sticking to a healthy diet, obtaining sufficient physical activity throughout the day and findingtime to relax. It is possible to obtain support to overcome theseobstacles from human coaches, or automated coaching systems. The latter are taking over the market for behavior modification support, due to the fact that they require less investment in terms both of money and time. Furthermore, such systems can provide more precise andbetter timed feedback on behavior, by using sensing and feedback devices (e.g. small accelerometers and a mobile phone) that users carrywith them througout the day. Such automated coaching systems currently follow one of the following approaches: one size fits all: all users follow the same trajectory throughout the program personalized coaching: at some points in the trajectory, the next coachingaction/trajectory is based on certain personal charac-teristics Human coaches still have an advantage over these automated coach-ing systems; they can acknowledge a clients personal issues. In PR-TN2012/00300 (M. Hendriks, M.A. Weffers, M. Spit and A.T. van Halteren, Algorithms for user centred, problem driven automated coaching), an automated coaching system is described which tailors the coachingtrajectory based on the problems that an individual user is encountering while trying to learn the new behaviour, using our coach-ing system. We call these problems/issues user dilemmas. Detection of ausers personal dilemma is done automatically. These algorithms are generically applicable, to any coaching system, or any service where user dilemmas can be identified. A generic component is planned to be implemented in the Digital Innovations platform, which deploysthese algorithms. This document describes A set of functional tests for this component, and a concept demonstrator of applicationof this component to a small use case.
Authors: Hendriks, M., Spit, M., Weffers-Albu, M.A.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W591701889