Catalog Home Page

A proposed taxonomy of media collections and its implications for design and management of multimedia databases.

Hobbs, V.J. and Pigott, D. (2004) A proposed taxonomy of media collections and its implications for design and management of multimedia databases. In: 12th International Conference on Information Systems Development, 29 - 31 August, Melbourne.

Google Books Preview: http://books.google.com.au/books?id=bd1Tv00URuEC&p...
*Open access. Some pages may not be available

Abstract

The extension of the principles of database management to encompass collections of media artefacts has led to a new range of solutions, and a matching extensive literature covering both research (including content-based retrieval, metadata standards and cataloguing, best practices for media acquisition, and digital preservation) and implementation (from web page content management to online museums, digital libraries, media archives and satellite and medical imaging). To make use of, and contribute to, this literature (whether by further research, design or new implementations) there must first be an overview of the domain. There are two ways in which this overview can be attained: one is to have a continual survey of what exists, and the other is to establish a comprehensive system of classification.

Surveys are found either as enumerative assessment of problem domains, or as justification for comprehensive general solutions. The former (such as Martinez and Mouaddib, 1999; Ozden et al., 1997; Vogel et al., 1995; Yoshitaka and Ichikawa, 1999), while invaluable when published, cannot hope to be exhaustive, and rapidly become obsolete; and we are still left with the (recursive) problem of referencing the surveys themselves. Moreover, we can have no reason to expect that a set of general principles will emerge from these efforts (Quicke. 1993). Even where such surveys begin with a checklist in the form of a taxonomy (such as Martinez and Mouaddib, 1999) the result is still an enumeration with regularised attributes rather than a classification proper. On the other hand, surveys that are made for the purposes of justifying general solutions (e.g. Del Bimbo, 1996; (Grosky, 1997) cannot necessarily be treated as comprehensive coverage of the domain, nor will a collection of such surveys ever amount to a complete coverage.

Classifications of multimedia collections should begin with an a priori taxonomy informed by surveys of the literature, but with the intent to comprehensively map out the entire multimedia domain. Such approaches have proved successful in the fields of visual programming (Burnett and Baker. 1994), data visualisation (Schneiderman, 1996) and media immersion interfaces (Gabbard and Hix, 2003). Taxonomies have the advantage that while the systems that are classified may become obsolete, the framework can deal with new examples and still serve its purpose. This paper proposes such a taxonomy, and describes its implications for design and management of multimedia databases.

Publication Type: Conference Paper
Murdoch Affiliation: School of Information Technology
Publisher: Springer
Notes: Appears in: Henry Linger, Julie Fisher, Wita Wojtkowski, Jože Zupančič, Kitty Vigo, Josie Arnold (eds)Constructing the infrastructure for the knowledge economy: methods and tools, theory and structure. 2004. Springer pp 519-530
URI: http://researchrepository.murdoch.edu.au/id/eprint/1893
Item Control Page