Title :
Neural networks in human-computer interaction: a view of user modelling
Author :
Finlay, Janet ; Beale, Russell
Author_Institution :
Dept. of Comput. Sci., York Univ., UK
Abstract :
Neural networks have proved valuable in solving pattern recognition and classification problems. Primarily they have been successfully applied to the fields of image and speech recognition, and much research is continuing in these areas. However in the field of human-computer interaction, pattern recognition is also a common requirement. The authors identify two related applications of neural techniques to recognizing patterns in user behaviour. The first is dynamic user modelling, which may be used in the development of adaptive interfaces, or intelligent help and tutoring systems, to provide the system with a model of the particular user. The second can be viewed as a form of static user modelling: analysing behavioural traces or protocol analysis
Keywords :
computerised pattern recognition; human factors; interactive systems; user interfaces; adaptive interfaces; behavioural traces; dynamic user modelling; human-computer interaction; intelligent help; neural networks; neural techniques; pattern recognition; protocol analysis; static user modelling; tutoring systems; user behaviour;
Conference_Titel :
Neural Nets in Human-Computer Interaction, IEE Colloquium on
Conference_Location :
London
