You are here: Home > FAQs
Not necessarily. Metadata embedded in one page or document may facilitate the discovery and retrieval of other documents or files.
In the first instance a document that is divided into multiple web pages and has an introductory or top-level page. In these cases, metadata applicable to the ‘whole’ document ought to be assigned to the introductory or top-level page and not to each individual web page. The introductory or top-level page does not constitute a ‘covering’ document (see below); it is rather an integral part of the document.
A second instance is where metadata might be deployed at the aggregate-level (see below).
Examples might include reports, strategy statements, white papers or legislative works where the publication has been split into different web pages (files). Presentation is often the reason for such division, and the need to avoid large page sizes.
Likely candidates:
Department of Foreign Affairs Strategy Statement – contents-listing page plus
five content pages:
http://www.irlgov.ie/iveagh/aboutus/strategy/default.asp
Department of Public Enterprise Report of Integrated Ticketing Committee -
contents-listing page plus nine content pages:
http://www.irlgov.ie/tec/transport/reports/integrated.htm
Department of Public Enterprise Annual Report and Financial Statements 1999
- contents-listing page plus seventeen content pages:
http://www.irlgov.ie/tec/report99/
Attorney General’s Irish Statute Book: Finance Act 1998 – contents-listing
page plus 138 content pages:
http://193.120.124.98/ZZA3Y1998.html
Speeches and press releases ought to be assigned their own metadata.
Metadata for non-HTML files can either be stored separately in a database (with a link to the resource) or embedded in a cover HTML page (if an HTML version of the document does not exist). The one covering document will suffice where two or more formats exist. If an HTML version does exist (alongside non-HTML formats), metadata embedded in the HTML document will suffice for discovery of the non-HTML format(s) (making use of the DC.Relation.hasFormat element. SEE DC.Relation).
A covering document is a document that contains a summary or abstract of another along with other details of that document. It is often designed to present the user with either enough information for them to determine the usefulness, accessibility or otherwise of the described document, or contain sufficient information that it might act as a substitute for the described document. 'Where it is not possible to produce a large document in HTML then a summary HTML version should be provided' (Web publication guidelines, October 1999).
Where a covering document is used, metadata ought to be applied to the covering document only, not to both it and the document covered.
A resource should not have more than one covering document.
The URL for the document being described.
Examples of the use of ‘cover’ documents include;
http://www.irlgov.ie/justice/Publications/Equality/KingsInn.htm – links to (pdf) paper delivered at a Bar Council seminar on refugee and asylum law in Ireland.
http://www.doh.ie/publications/imprec.html – includes publication details of a report on the implementation of recommendations of The Commission on Nursing with a link to a (pdf) copy.
http://www.gov.ie/bills28/bills/2000/6900/default.htm - summary page for Bill with link to Bill in pdf format.
http://www.gov.ie/educ/educationno2bill.htm – links to (pdf version of) Education (No.2) Bill, 1997 and separate (pdf version of) explanatory memorandum.
A covering document usually relates to one other resource, not multiple resources.
However, a covering document may relate to more than one resource where it acts as a linking page to like-subject documents only (as distinct from multiple or different subjects). Metadata applied in this manner may be referred to as aggregate-level (or collection-level) metadata. Assigning metadata in this manner allows groups of resources to be retrieved. It is dependent on like material being arranged in a file structure which brings the covered material together and therefore facilitates levels of aggregation and the maintenance of aggregate-level metadata records. Aggregate-level metadata does not necessarily replace item-level metadata, and its deployment is dependent on the requirements the metadata is meant to serve, and in light of any IPSMS (Irish Public Service Metadata Standard) compliance requirement.
Example 1: The Department of Agriculture has grouped all documents relating to foot and mouth disease together, conveniently, in one folder with an introduction page (www.gov.ie/daff/AreasofI/FMD/default.HTM). All these documents share the same theme or subject matter, namely foot and mouth disease.
Example 2: The Department of Enterprise, Trade and Employment has grouped all documents on import licensing together, conveniently, in one folder with an introduction page (www.entemp.ie/tcmp/export.htm). It has similarly treated content on export licensing in the same manner.
It is important to remember that the aggregate-level metadata record is one record (though referring to more than one document), not an amalgam of a number of metadata records. (see Can I include multiple metadata records covering different documents in the <HEAD> of one document? below).
If an aggregate-level metadata record is used, it is important to add as much information as possible to ensure that the documents covered can be retrieved. This means using subject terms covering all the content covered by the metadata record.
Pdfs can be converted to HTML using third-party products such as the Magellan plug-in for Acrobat or Iceni Gemini or the free online converter Gohtm.com.
With documents that are held in more than one location, it is only necessary to have metadata attached to one of the records, with due mention of the two or more locations (via repeat use of the DC.Identifier element).
This scenario is not dissimilar to that where documents exist in more than one format. (SEE How do I assign metadata to pdfs, rtfs and word documents? above). This helps avoid the return of duplicate search results, and inconsistency where two metadata records for the same resource differ. The difficulty is that where there are two separate parties involved, an arrangement needs to be made and adhered to as to which copy contains the metadata record.
No. Though visually it might seem that one has separate records when viewing the source of a document, it does not mean that one has separate records. In fact it constitutes one record with repeat use of the elements. One merely ends up with a 'confused' metadata record.
Only one IPSMS DC metadata record can be associated with any one document.
The attached template allows you to create metadata records using a simple copy and paste procedure. The data pertinent to a record is entered as the content (content=" "). The text must reside between the double straight quotes " ".
If an element is to be repeated (e.g. document matching two document types), just copy the appropriate line and paste in below. The date entry should match the format given.
When the record has been completed, copy and paste it into the <HEAD> of the document, placing it just before the closing </HEAD> tag.
The description ought to be a textual summary of what a document is about. You may also wish to indicate intended audience, purpose of document, or any such information as assists the user in determining content and suitability. Whereas a summary may contain within it keywords that may be suitable for indexing a document, it is not appropriate for a description to consist purely of keywords. The DC.Subject entry is the appropriate element for capturing subject terms.
Ideally, the description ought be kept to a maximum of 3-4 lines.
| << Previous Section |