High‐level Subject Access Tools and Techniques in Internet Cataloging

Citation

Gilchrist, A. (2006), "High‐level Subject Access Tools and Techniques in Internet Cataloging", Library Review, Vol. 55 No. 3, pp. 228-229. https://doi.org/10.1108/00242530610656055

This 115 page A5 paperback was co‐published simultaneously as Journal of Internet Cataloging, Vol. 5 No. 4, 2002. The rationale of this example of re‐purposing is obvious from the title, but it is not clear whether the papers happened to arrive more or less simultaneously as submissions to the journal, or were specially commissioned with both in mind. The six articles and Introduction constitute, then, a useful snapshot of some related activities in this area for journal publication, but do not provide the more wide‐ranging view, expected in a book, of a problem that is taxing information architects; not only in the context of the Internet, but also in the many intranets being established and developed in both the public and private sectors.

All the six essays are by librarians: five are from the USA and one from the UK. One essay is solely concerned with the Dewey decimal classification, while another – the HILT Project – reports on research into cross‐database searching using Dewey as a reference spine for mapping other languages; and three are concerned with the Library of Congress Classification. Consequently, the main theme, in this book, is really how to adapt traditional bibliographic classifications as entry points to legacy data in the context of the complexities of networked storage and delivery. The main and common problem, therefore, is how to overcome the inherited system inertia arising from the continued (and economically justifiable) use of de facto standards. As Davis says in his essay: “As we studied LCC, we saw…that underneath its enumerative top‐level, there were potentially useful hierarchies [reviewer’s italics] at the second and subsequent levels as revealed by captions and the indispensable, conceptually significant indentations present in the schedules as a guide to classifiers”, but then goes on to say, “On the other hand, many LCC hierarchies do seem outdated or at least less useful in the sense that naïve users might have trouble inferring lower level subject categories from the upper levels of the hierarchy”. A quarter of a century ago, the UK Classification Research Group worked hard to promote the Bliss Classification, and to attract funding for its speedier revision. It is, with hindsight, arguable that an opportunity was missed to use a universal, fully faceted classification more widely, and that current problems of cross‐database searching and mapping might have been ameliorated.

Although the term “high‐level subject access” is used in the title, it is not explicitly defined in the book, and there is only one entry to “high‐level” in the index (missing, incidentally, the use of high level in the acronym HILT standing for high level thesaurus). This aspect of granularity deserves more thought. There is work under way on access at the collection level (and in corporate enterprises at the repository or web page level). Only the Editor refers to Topic Maps in her introduction, an approach, incorporated in an International standard that is vying with the Resource Description Framework (RDF) as a preferred tool for the Semantic Web. This is not to say that Topic Maps or RDF are silver bullets, but they are examples of techniques many of which are emerging from outside the library profession, of which digital librarians should make themselves aware.

Naturally, though, there is more interest in access at the document level; and in certain cases at the content level. This is not to imply that the approaches described in this book are invalid, but merely to point out that the problem of granularity is critical. Several of the authors draw attention to the current “paradigm” exhibited by the Yahoo! Directory approach and the superficial similarity of its menus to, for example, the Dewey classification. The Editor, in her Introduction, quotes Kirsten Risden, who in her research in this specific area has said “Though individual subject trees, especially Yahoo are frequently described in articles on Internet search engines, it is rare for a number of them to be described and compared and the prime concern of authors who do so has usually been the information content of subject trees and not their broad structure or classified organization”. Nor has much research been reported on the effectiveness of the directory approach, and it is salutary to remember that the business driver behind this approach is the knowledge that the majority of people do not know how to search with Boolean operators (see for example Stacey, 2003). The remarks made by Davis above illustrate clearly how much work will need to be done in order to translate all the clues embedded in a bibliographic classification, evident only to a trained classifier, and to re‐present these transparently on a website.

What is encouraging about this book is the realisation by all the authors that information provision is moving from being “materials‐based” to being “user‐focused” as argued by Rothman in his essay, or “task‐based” as argued by Forsythe and Shadle in theirs; a movement which will also incorporate a greater degree of personalisation. Davis proposes eight design principles from his work on developing a hierarchical interface to the Library of Congress Classification, most of which are user‐focused; while Nicholson et al. state that their research found “a perceived need for services or service staff to provide assistance to users searching or cross‐searching by subject” and go on to quote Jacso (1999), who notices that “in many cases, databases do not offer the necessary aids required to use the ‘preferred terms of the subject controlled vocabulary”. The emphasis here is on semantic, rather than notational, entry which is normally a mystery to most users. Finally, Morgan and Reade stress that much more needs to be done to satisfy the users, quoting from a gap analysis study “Of the four constructs measured, ‘Personal control’, which measures a university’s Web functionality was rated of most importance to users. However, perception of library success in this area was barely above the established minimum level” (Cook, 2001).