City Research Online

Abstract

This research forms part of a larger project, the eventual aim of which is the design and implementation of an intelligent interface for document retrieval systems. A number of functions which must be performed by the human intermediary in order to successfully interact with the user have been identified. The research presented here is concerned with one function in particular: the user modelling function, which aims to describe and model various aspects of the user’s background, personal characteristics, goals and knowledge. An assumption underlying this research is that an intelligent interface should simulate the functional behaviour of a competent human intermediary. Therefore the ways in which human intermediaries carry out user modelling and employ these models, have been investigated.

The primary method was to make audio recordings of seven human user/human intermediary interviews in online search service settings, and to subject the transcripts to detailed functional discourse analysis. This analysis produced a specification for the User Model, and identified its components and the knowledge resources that are needed by the intermediary, whether human or automatic, to carry out the function of user modelling. This analysis was supplemented by the examination of a number of users’ problem statements, together with their accompanying recordings, which had been collected for another project, and by interviews with three intermediaries. The discourse analysis revealed that the User Model interacts with the other interface functions, and this interaction was also investigated.

The results showed that the User Model comprises a number of subfunctions, requires extensive knowledge resources, and interacts with the other functions, in particular providing information necessary for the other functions’ own processing. A formalism for representing the User Model in a computer system is suggested, and an attempt is made to validate the User Model by applying it to a new dialogue. The results of the validation suggested that the User Model is independent of the data on which it is based, and that the formalism can adequately handle a new interaction. The implications of these findings for the design and implementation of the user modelling function in an intelligent interface, and for the design and implementation of the interface as a whole, are outlined.

Publication Type:	Thesis (Doctoral)
Subjects:	Z Bibliography. Library Science. Information Resources > Z665 Library Science. Information Science Z Bibliography. Library Science. Information Resources > ZA Information resources Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4050 Electronic information resources Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4450 Databases
Departments:	School of Communication & Creativity > Media, Culture & Creative Industries > Library & Information Science School of Communication & Creativity > School of Communication & Creativity Doctoral Theses Doctoral Theses

Export

Downloads