Indexing & Taxonomy Creation

Building indexes and taxonomies are two sides of the same coin. Understanding each process and how they are related can yield insights that will improve the effectiveness of search and navigation.

• What are the definitions of "indexing"?
• What can the indexing process tell us about taxonomy development?
• How do "back of the book" indexes and A-Z web site indexes provide insights about taxonomy development and tagging?
• What are the best practices for tagging documents?

Indexing and Taxonomies
by Heather Hedden, Hedden Information Management

Heather Hedden’s presentation focused on the comparison of the processes of alphabetical indexing and hierarchical taxonomy development.

Heather began with definitions of indexes and taxonomies. An index points to information in a document, and a taxonomy is a system for classifying information from multiple documents or sources.

In comparing indexes and taxonomies, there two aspects to compare: the process and the resulting organizational display. The process for creating an index is a single integrated activity, whereas a taxonomy involves the two steps of developing the terms and then linking/tagging the content. The organizational display is typically alphabetical for an index and hierarchical for a taxonomy. A alphabetical arrangement of terms may be created simultaneously with the linking to content (indexing) or may be done separately to create a controlled vocabulary for later indexing. A hierarchical arrangement of terms (taxonomy) must be created prior to indexing. A controlled vocabulary (created in advance) may be alphabetical, hierarchical, or a combination.

Heather then provided a detailed comparison of the procedure and skills involved in index or taxonomy creation and also for the process of tagging/indexing content to a taxonomy. Next some of the tools used for index-creation and for taxonomy creation are listed.

A comparison of best practices for index creation, taxonomy creation, and taxonomy term tagging was also presented.

Finally, Heather presented resources with links for information on indexing in particular.

Indexes and Taxonomies: Characteristics in Contrast
Fred Leise, ContextualAnalysis, LLC

Fred Leise’s presentation also compared indexes and taxonomies but looked more at the indexes and taxonomies themselves than the process of creating them. His presentation also addressed the concepts of significance and relevance in indexes and taxonomies and the implications for index/taxonomy creation.

Fred’s definition of a taxonomy included its broader use to denote any controlled vocabulary: authority files (list of approved terms), taxonomies with a strictly hierarchical structure, and thesauri with additional associative structure (BT, NT, RT).

The aspects of comparison and contrast of indexes and taxonomies presented comprise uses, purposes, dimensions, term coordination, product-specificity, indexing granularity, method of creation, usage, and ability to support automation.

What is most important in contrasts is that an index is a dependent content access structure. An index always exists in a specific context, as an index to something: a book, a journal, or a series of journals. An index is a controlled vocabulary with links to specific content. A taxonomy, on the other hand, may exist as independent structure, without reference to a specific context.

Fred also covered a comparison of concept significance and relevance in indexes and taxonomies. For each indexes and taxonomies, he addresses what is significant and how is it determined, and what is relevant and how is it determined.

Other issues for index and taxonomy creation included repurposing, change, maintenance, interfacing with other systems, and extensibility.