Government of Canada
Core Subject Thesaurus

If you change a term, what would be the way to let the users know that they should use the new term or the revised one and how they can correct the old "records" or texts where they assigned the old heading ?

Our plan is to publish a list of new terms added or terms modified between September 1 2001 to Michèle Hudon's ca February 2002 contract delivery date, on the thesaurus website. After that, we will publish such changes with a frequency to be determined by circumstances. Hopefully, a new thesaurus platform will enable users / indexers to search for and display recently modified terms with a date restricted search. We can do this internally on Multites now, but we cannot publish it dynamically to a Web interface.

Revising previously assigned terms is a policy matter for the indexing agency to determine. In practice, terms are very rarely deleted from the thesaurus. Terms modified, perhaps demoted or archived and their status as indexing terms may be changed. With fully functional online thesaurus application, users / indexers will be able to access this information.

Some terms are plural and some not, e.g. Patents and Brevet. Does it matter as long as the term is there in either singular or plural?

In general, it is common practice to represent English thesaurus terms in the plural form when they represent concrete entities that can be counted. Terms that represent abstract concepts or unique entities are generally in the form of singular nouns. In English, this results in a preponderance of plural nouns in a subject thesaurus, particularly one such as the CST that is post-coordinated and in which compound terms are avoided.

Compilers of French language thesauri and indexers working in French have generally preferred to use singular forms for thesaurus terms, with exceptions when required.

Indexers using the CST should use consistently the form of term indicated therein.

I have a question about which term to use when alternatives are given by the thesaurus for terms selected from the index list.

Upon selecting the term 'Environmental assessment' the page indicates to use 'Environmental impact assessment' and only provides the French translation for this word.

So does this mean that I should ONLY use the 'Environmental impact assessment' term in my metatag and NOT 'Environmental assessment'. Or are both allowed? What is the preferred/required method?

Thesauri are controlled vocabularies, which means that words or phrases to be used for indexing are pre-determined or selected from an indefinite number of synonyms or near-synonyms and are designated for use in indexing to the exclusion of all of the other synonyms.

The thesaurus contains so-called "Preferred terms" (PT) and "Non-preferred terms" (NPT). These are sometimes referred to respectively as "descriptors" and "non-descriptors" or "index terms" and "lead-in terms".

The thesaurus identifies a single PT to represent or index a given concept and then includes a number of synonyms or NPT's.

The PT is to be used for indexing or for populating the dc.subject tag. NPT's are not to be used for indexing or meta-tagging but are in the thesaurus to guide the indexer to the single preferred term.

For this reason only PT's have other language equivalent terms indicated.

Each PT record will have "USE FOR" (UF) terms indicated as part of the term record / display. This list of synonyms partly supplies a semantic context for the term and helps the indexer decide whether or not this indexing term is the appropriate one.

Each NPT record will have a "USE" term indicated as part of the term record / display. This tells the indexer that this term is not to be used for indexing and directs the indexer to "USE" the PT instead.

In the example that you cite, "Environmental impact assessment" is the PT and is to be used in indexing and metadata application. "Environmental assessment" is the "UF" or NPT and is not to be used.

"Environmental impact assessment" has a French equivalent indicated ("Évaluation environnementale"). THIS PT is to be applied to French language documents or in any case where the metadata is in French.

I found the term "Environmental conservation" in English, but there is no French equivalent.

The reason is that "Environmental conservation" is a "lead-in" or "Non-preferred term". There is a "USE:" note in the term record, in this case, "USE: Environmental management". In other words, "Environmental management" is to be used as the index term or metadata element, rather than "Environmental conservation". There is a French equivalent provided for "Environmental management".

Why aren't the GoC Core Subject Thesaurus and Termium combined as one product? Whatever goes into the GoC meta tags must be translated anyway.

The reason is that the CST is a controlled vocabulary (in the form of a thesaurus) whereas TERMIUM is a multilingual glossary. A thesaurus designates a single term to represent a concept, while a glossary records any number of words or phrases that may represent a concept to a greater or lesser degree. A glossary tells the user what a given word or phrase means or what its possible range of meanings are. It may provide words or phrases with closely related meanings or exact equivalents (synonyms). This is what TERMIUM does.

A thesaurus is a controlled vocabulary that identifies a single term, selected out of a range of possible ones, that is to be used to represent a given concept in an index, (or metadata tag), to the exclusion of all others. A thesaurus also provides a semantic context for its indexing terms by means of hierarchical (broader and narrower) and associative (related to) relationships. It may also contain supporting information such as scope and history notes.

A glossary provides natural language terms and therefore may be an appropriate resource for keywords. It may contain an indefinite number of entries.

The use of a thesaurus in electronic indexing limits the possible range of search terms to a small set that have been stripped of ambiguity and enhanced for semantic accuracy. Typically they are embedded in the target document and therefore provide a simple and direct link between a concept (search term) and an information resource known to include that concept in its subject matter.

Will there be any interoperability or standardization between government thesaurus's, especially between common terms such as brochures. For example, in the GC thesaurus, for the English term "Brochures" it says to use "Booklets". However in the Stats Can thesaurus it says to use "Pamphlets". Which one is right?

Both examples are "right" but they are derived from different vocabularies. Different vocabularies may have different scopes and purposes and probably are based upon the indexing of different bodies of documents. Hence differences. For the purpose of indexing under GOL metadata requirements, any registered vocabulary may be used. When <dc.subject> is populated, the schema must be identified.

E.g.

<html>

<head>

<link rel="schema.dc>" href="http://purl.org/dc/elements/1.1/">

<meta name="dc.subject scheme="gccore" content= " ">

etc., etc.