Ulster University Logo

ADAPTIVE EMPATHETIC INTERFACES FOR AGEING PEOPLE

Sayers, Heather, Mulvenna, Maurice, McDaid, Liam and Hunter, A (2007) ADAPTIVE EMPATHETIC INTERFACES FOR AGEING PEOPLE. In: Tromsø Telemedicine and eHealth Conference (TTeC) 2007 : Elderly - Who Cares?, Tromsø. Norwegian Centre for E-health Research. 1 pp. [Conference contribution]

[img] Text - Accepted Version
Indefinitely restricted to Repository staff only.

62kB

Abstract

The aim of this research is to investigate how adaptive, empathetic computer interfaces can help maintain the quality of the lives of ageing people who wish to remain in their homes for as long as possible and lead active and fulfilling lives in the heart of their communities. Ageing people constitute an increasing part of modern populations and they are as diverse, if not more so, as their younger counterparts. They experience many physiological and psychological changes as they age, including declines in physical mobility, attention span, hearing, vision, memory and cognitive capabilities. Research also shows that they experience significant usability problems with computers. Computer technology, however, offers the potential for ageing users to monitor their own health and at the same time enjoy fuller participation in society. Activities of Daily Living (ADLs) are functions that people perform in everyday life such as eating, washing and shopping, and ADL performance metrics are often used to provide objective measurements that help assess the level of independence of ageing people. We introduce the concept of e-ADLs to monitor users’ computer interaction activities. By mapping changes in this data to associated changes in physical and cognitive abilities, it may be possible for ageing users to monitor and assess their own abilities. Intelligent techniques can be used to address the usability problems experienced by establishing and dynamically implementing changes in interface representation to meet the needs of each individual user as they arise and change over time. By supporting ageing users in carrying out daily browsing tasks, for example, they can be helped to maintain vital social contacts with carers, friends and family, remain informed and carry out everyday tasks such as shopping or paying bills from the comfort of their own homes. An intelligent, empathetic interface could dynamically adapt to assist ageing users, and evolve to help mitigate the effects of decline associated with the normal ageing process, while users could, at the same time, utilise measurements derived from the recognised e-ADLs to monitor changes in individual physical and cognitive abilities.This research presents results from our initial work on Internet browsing which clearly demonstrate that it is possible to monitor and log detailed interaction activity and to classify interaction patterns using intelligent techniques. An experiment was carried out to establish some typical browsing event data. The participants in the experiment consisted of groups of students in laboratory classes who were asked to complete two sets of browsing tasks in 20 minute sessions (A and B). Both sets of tasks required the users were designed to require more interaction (A) or less interaction (B). The users’ interactions were logged for each session in the form of an XML document. These documents were then parsed and loaded into a relational database and data on the number of mouse clicks for each user in each session was extracted. An Artificial Neural Network (ANN) was then trained to classify users into categories A or B. The aim was to train the network to recognise when a user was carrying out tasks in Session A or in Session B by analysing the number of mouse clicks. This network yielded an accuracy of 98%, and when tested with unseen data the classification accuracy was 91%. In this instance one event -mouse click- was classified by the network but this can be extended to include keyboard inputs, time spent on a page, etc. Future work also involves incorporating intelligent functionality to recognise when changes in individual users’ interaction patterns have occurred and mapping these to changes in physical and cognitive capabilities. This would offer ageing users home-based health and well-being assessment capabilities while making the experience more enjoyable through the ease-of-use catered for by the underlying intelligent system.The ultimate research aim is to develop a networked application which would be both useful in healthcare support and also usable and enjoyable. An application based upon the use of novel game playing and role playing themes is envisaged which would allow ageing people to assess their own general well-being through the use of objective assessment techniques and scales commonly used by healthcare professionals, and which would also provide an enjoyable context to the assessment and care processes. The benefits of using home-based self- service are objective measurement, more regular measurement and the provision of longitudinal data for trend analysis. The primary benefit is of course that these measurements are available to the healthcare professionals responsible for the care and well-being of ageing people. Novel game and role playing techniques are planned, for example with an avatar representing the primary method of communication with the ageing person. Medical and social knowledge can also be incorporated into the application to provide an empathetic interface based upon advances in affective and emotion-orientated computing research.

Item Type:Conference contribution (Paper)
Keywords:Empathetic Interfaces
Faculties and Schools:Faculty of Computing & Engineering
Faculty of Computing & Engineering > School of Computing and Mathematics
Faculty of Computing & Engineering > School of Computing and Intelligent Systems
Research Institutes and Groups:Computer Science Research Institute > Intelligent Systems Research Centre
Computer Science Research Institute
Computer Science Research Institute > Artificial Intelligence and Applications
ID Code:37066
Deposited By: Professor Maurice Mulvenna
Deposited On:27 Feb 2017 12:29
Last Modified:17 Oct 2017 16:28

Repository Staff Only: item control page