|
FACET
ANALYTICAL THEORY
Facet analysis in a rudimentary form was conceived by S.R. Ranganthan in the 1930s, although it had been preceded by similar analytico-synthetic approaches to subject classification and indexing, notably by Henry Bliss and Paul Otlet in the classification field, and Kaiser in indexing. It was developed post 1950, principally by members of the UK Classification Research Group, as a tool for the organization of document collections in technical, scientific and social scientific fields, where it was highly effective in the storage and retrieval of compound and complex subjects. Modern facet analytical theory contrasts with earlier views of knowledge as an integral whole (which is broken down into smaller and smaller units) in that it deals with individual terms or concepts which are clustered into categories to create a 'bottom-up' map of knowledge. A number of categories have been identified which are widely applicable to the terminologies of a range of subject fields; these categories are generally functional and/or linguistic in nature (e.g. entities, processes, properties, operations, agents). Compound and complex subjects are accommodated by combining individual concepts. Various forms of system syntax (links and rules for ordering and combination between categories) have been proposed to compound the individual concepts, most of which are based on natural language models; the method used in 'classical' UK facet theory depends on a formulaic inter-category order (citation order). Currently, facet analysis is used primarily to create classifications for the physical arrangement of documents (or document surrogates). The completed classification or knowledge structure is built up from individual terms which are analysed into categories and ordered by the application of the system syntax. The resultant structure is logical and predictable, and therefore highly effective in storage and retrieval. The only widely used classification embodying these principles is the second edition of the Bliss Bibliographic Classification, which employs standard categories and citation order. It seems probable that this methodology of facet analysis can be used in a broader and more innovative way to create much deeper and more complex knowledge structures and semantic networks, by extending the range of categories and by exploring variants on combinatorial methods. Although the faceted classification is regarded by many as a structure with specific characteristics, essentially facet analysis is a technique, and different models of the same universe of discourse can be derived to meet different local or subject-specific needs using different categories and variations on the syntax. Some
research is now required into the fuller range of categories and
relations that may be encountered across the complete range of
disciplines. Work is also required on the problems of interdisciplinarity,
and alternative approaches to the structuring of knowledge that
do not depend on traditional disciplines as the first point of
entry; in this area, classifications of phenomena (as opposed
to the more conventional aspect classifications), and the further
application of integrative level theory, require some investigation.
Additional properties of digital objects, especially non-text,
multi-media and images, can also provide data for categorical
analysis, and may affect the potential syntax of the system. The
formulae developed for the combination of terms and concepts will
generate n-dimensional structures that seem appropriate to a hypertext
environment. Structures generated from the expanded category base
may be particularly useful in handling digital objects. (Written by Vanda Broughton)
|
 |
 |
 |
 |
 |
 |
 |
|
||||||