Lunzer, Aran Edward (1995) Reconnaissance: A Widely Applicable Approach Encouraging Well-Informed Choices in Computer-Based Tasks. PhD thesis, University of Glasgow.
Full text available as:
PDF
Download (13MB) |
Abstract
The aim of this thesis is to propose and demonstrate a novel approach to human-computer cooperation in exploratory tasks, that encourages the pursuit of thorough explorations and thus increases the likelihood of finding the best available results. The thesis identifies a class of computer-supported tasks - such as artifact design, decision analysis, data presentation, and many kinds of retrieval - that all involve an exploration among alternative result specifications to find one whose outcome best fits the user's current needs. Such explorations cannot be automated, but must be directed by the user. Since there is typically no way of knowing in advance what results are available, the user must continually trade off the effort required for further exploration against the unknown chance of discovering results preferable to those found so far. The more arduous the exploration is perceived to be, the greater the user's temptation to accept early results without due consideration of alternatives. By using a new illumination zone model to analyse existing systems, the root of the problem is characterised as a lack of support for visualising the trade-offs between results in terms of the criteria that the user feels are significant at the time. Reconnaissance - in the familiar sense of using scouts to examine unknown territory - provides a suitable metaphor for the task of instructing a computer to generate a range of results, to summarise them according to the current set of criteria, and to collate these summaries into a single visualisation. A challenge in supporting computer-based reconnaissance is the provision of a convenient interface that integrates the requesting, viewing and comparison of result summaries. An implemented example of reconnaissance support shows that the parallel coordinates plotting technique can be extended into an interactive form that meets this challenge. Tests carried out on this implementation, which provided reconnaissance facilities for an existing document formatting system, confirmed that more thorough explorations were indeed encouraged. The tests also brought to light additional, unforeseen advantages to the use of result ranges and summarisation.
Item Type: | Thesis (PhD) |
---|---|
Qualification Level: | Doctoral |
Additional Information: | Adviser: Steve Draper |
Keywords: | Computer science |
Date of Award: | 1995 |
Depositing User: | Enlighten Team |
Unique ID: | glathesis:1995-75552 |
Copyright: | Copyright of this thesis is held by the author. |
Date Deposited: | 19 Nov 2019 19:28 |
Last Modified: | 19 Nov 2019 19:28 |
URI: | https://theses.gla.ac.uk/id/eprint/75552 |
Actions (login required)
View Item |
Downloads
Downloads per month over past year