Metadata Standards
Background
The Australian Government, in conjunction with the National Archives of Australia has created the Australian Government Locator Service (AGLS) metadata standards to improve the visibility and accessibility of government departments' and agencies' services over the Internet. In other words, they want the individual organisations to use a common standard for describing the content and characteristics of their documents. The CSIRO is included in the agencies who are supposed to adhere to these standards and, as a unit within the CSIRO, so are we.
The broad idea behind this is that a user should be able to discover the capabilities and knowledge stored across the different areas of the government by going to a central search engine. However, other search engines can also use the metadata. Capturing certain information such as author, date created, subject can also be of assistance in maintaining the website. Suppose that a new astronomy software package Z was very different to the previous package X and that the procedural documents which must be updated were scattered across the website. Searching only for documents with a subject containing package X rather than the main text should eliminate documents such as research papers where package X may have been mentioned in passing.
Metadata Structure
The National Archives of Australia provide a list of metadata elements for the AGLS, based upon the Dublin Core standards. The CSIRO has already adopted much of the AGLS standard for their corporate website as well as adding additional elements particular for their setup. The CSIRO metadata is listed in full at http://www.csiro.au/itsb/csiroonline/metadata.html. We can base most of the ATNF metadata on the compulsory elements of the AGLS. The use of additional ATNF metadata is recommended for internal management purposes. For example, if ATNF.AccessPermission is labelled Internal, then the maintainer will know not to move it to a public access section of the website.
The proposed elements are listed in the tables below. Note that only those components considered useful are listed. The User Entered field indicates if the contents of the metadata element require updating by a page author or if they can be stored in the template.
Dublin Core Based Elements
|
Element
|
Components
|
Description
|
User Entered
|
| DC.Creator |
personalName, email1 |
Page author details |
Yes
|
| DC.Publisher | corporateName, jurisdiction, contact, address, email2 | The name and contact details for the organisation responsible for making the resource available (the ATNF) |
No
|
| DC.Rights | Link to copyright page |
No
|
|
| DC.Title | The title of the document |
Yes
|
|
| DC.Title.alt | An alternative title or subtitle for the document |
Yes
|
|
| DC.Subject | Keywords for the document |
Yes
|
|
| DC.Description | Description of the document contents |
Yes
|
|
| DC.Language | Language the document is written in (en - English) |
No
|
|
| DC.Date.created | Date document created |
Yes
|
|
| DC.Date.modified | Date of last modification to the document |
Yes
|
|
| DC.Identifier | The url of the document |
Yes
|
1The other components are: corporateName, jurisdiction, sector, contact, address. However, much of this information belongs to DC.Publisher, is sensitive information, or is inappropriate for the ATNF.
2The same components are available as for DC.Creator.
ATNF Elements
|
Element
|
Components
|
Description
|
User Entered
|
| ATNF.Site | The website at which the document belongs - Marsfield, Parkes, Narrabri |
Yes
|
|
| ATNF.Maintainer | contactName, email | The person (or group) responsible for maintaining the page |
Yes
|
| ATNF.MetaVersion | The version number for our metadata scheme |
No
|
|
| ATNF.AccessPermission | Public or Internal |
Yes
|
Example
<link rel = "schema.AGLS" href="http://www.naa.gov.au/recordkeeping/gov_online/agls/1.2"> <meta name="DC.Creator" lang="en" content="personalName=Wright,Andrew; email=Andrew.Wright@atnf.csiro.au">
<meta name="DC.Publisher" lang="en" content="
corporateName=CSIRO Australia Telescope National Facility;
jurisdiction=Commonwealth;
contact=+61 2 9372 4100 (phone), +61 2 9372 4310 (fax);
address=PO Box 76 Epping NSW 1710 Australia;
email=atnf@atnf.csiro.au">
<meta name="DC.Rights" scheme="URI" lang="en" content="http://www.csiro.au/aboutCSIRO/disclaimer.html">
<meta name="DC.Title" lang="en" content="Web Design Specifications">
<meta name="DC.Title.alt" lang="en" content="Structure and Navigation">
<meta name="DC.Subject" lang="en" content="HTML; navigation; usability">
<meta name="DC.Description" lang="en" content=" The structure of the ANTF website should be modified to improve usability">
<meta name="DC.Language" scheme="RFC1766" content="en">
<meta name="DC.Date.created" scheme="ISO8601" content="2001-04-27">
<meta name="DC.Date.modified" scheme="ISO8601" content="2001-05-11">
<meta name="DC.Identifier" scheme="URI" content="http://www.atnf.csiro.au/computing/web/spec12.html"> <meta name="ATNF.Site" content="Epping">
<meta name="ATNF.Maintainer" content="contactName=webmaster; email=webmaster@atnf.csiro.au"> <meta name="ATNF.MetaVersion" content="0.8">
<meta name="ATNF.AccessPermission" content="internal">
Implementation
The amount of user entered metadata in the proposed scheme is quite large. Some of the metadata is currently entered through tags such as <title> and in the "maintained by" link at the foot of the pages. Much of the data can be reused for each individual. I suggest that a online form be used to generate page templates. Repeated items such as titles and maintainer names can be entered once and automatically assigned to the correct page sections. The generated code can then be (for now) cut and pasted into their preferred HTML editor. It should be possible to setup a mechanism, say with cookies, whereby the user preferences do not require re-entering every time a new template is generated. There may still be problems associated with certain programs used for web page creation, but those are general issues which will be considered elsewhere.