Use this URL to cite or link to this record in EThOS:
Title: Visualising multiple overlapping classification hierarchies
Author: Graham, Martin
Awarding Body: Edinburgh Napier University
Current Institution: Edinburgh Napier University
Date of Award: 2001
Availability of Full Text:
Access through EThOS:
Access through Institution:
The revision or reorganisation of hierarchical data sets can result in many possible hierarchical classifications composed of the same or overlapping data sets existing in parallel with each other. These data sets are difficult for people to handle and conceptualise, as they try to reconcile the different perspectives and structures that such data represents. One area where this situation occurs is the study of botanical taxonomy, essentially the classification and naming of plants. Revisions, new discoveries and new dimensions for classifying plants lead to a proliferation of classifications over the same set of plant data. Taxonomists would like a method of exploring these multiple overlapping hierarchies for interesting information, correlations, or anomalies. The application and extension of Information Visualisation (IV) techniques, the graphical display of abstract information, is put forward as a solution to this problem. Displaying the multiple classification hierarchies in a visually appealing manner along with powerful interaction mechanisms for examination and exploration of the data allows taxonomists to unearth previously hidden information. This visualisation gives detail that previous visualisations and statistical overviews cannot offer. This thesis work has extended previous IV work in several respects to achieve this goal. Compact, yet full and unambiguous, hierarchy visualisations have been developed. Linking and brushing techniques have been extended to work on a higher class of structure, namely overlapping trees and hierarchies. Focus and context techniques have been pushed to achieve new effects across the visually distinct representations of these multiple hierarchies. Other data types, such as multidimensional data and large cluster hierarchies have also been displayed using the final version of the visualisation.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: QH301 Biology ; QA76 Computer software Computer software Pattern recognition systems Pattern perception Image processing