Handbook   Overview   Lectures   Tutorials   Readings   Assessment   Staff

IMS5048: Information Continuum -- Lecture 3

Topic 9: Metadata

21 September 2005.

Contents:

1. Introduction.
2. Functional requirements for accessibility.
3. Review of metadata and technology features of Information Continuum Model.
4. Examples of international and national metadata initiatives.

4.1. Example 1: Dublin Core.
4.2. Example 2: Australian Government Locator Service.
4.3. Example 3: NAA Recordkeeping Metadata Standard.
4.4. Example 4: SPIRT Recordkeeping Metadata Project.

5. The Resource Description Framework.
6. Metadata Registries.
7. Metadata Creation.
8. Automating metadata.
9. Strategic directions.

Readings

Australian Government Locator Service Site: http://dublincore.org/archives/1999/dc7/AGLS.htm.

Buckland, M., ‘Vocabulary as a Central Concept in Library and Information Science'. See http://www.sims.berkeley.edu/~buckland/colisvoc.htm .

Debbie Campbell, ‘Dublin Core Metadata and the Australian Metaweb Project', Sept 1999, http://www.nla.gov.au/nla/staffpaper/dcampbell1.html.

‘ Demystifying Metadata', at http://mappa.mundi.net/trip-m/metadata/

Dublin Core Site: http://dublincore.org/.

Metadata Standards and Resources http://www.egov.vic.gov.au/Research/WebSiteIssues/Metadata/metadata.htm

Meta Matters: http://dcanzorg.ozstaging.com/mb.aspx. This site provides an introduction to metadata and the Australian metadata community.

Metaweb Project: http://www.dstc.edu.au/Research/Projects/metaweb This site contains information about metadata initiatives and tools.

Metaweb Project: Analysis of Metadata Creation Tools: http://www.dstc.edu.au/Research/Projects/metaweb/toolpost-orig.html.

Sue McKemmish, Adrian Cunningham and Dagmar Parer, ‘Metadata Mania', paper presented by Sue McKemmish and Adrian Cunningham to the ASA Conference in Fremantle, August 1998, http://rcrg.dstc.edu.au/publications/recordkeepingmetadata/sm01.html.

Eric Miller, 'An Introduction to the Resource Description Framework', D-Lib Magazine , May 1988, http://www.dlib.org/dlib/may98/miller/05miller.html.

National Archives of Australia, Recordkeeping Metadata Standard for Commonwealth Agencies: http://www.naa.gov.au/recordkeeping/control/rkms/contents.html.

Resource Description Framework Site: http://www.w3.org/Metadata/RDF/.

SPIRT Recordkeeping Metadata Project Site: http://rcrg.dstc.edu.au/research/spirt/.

Stu Weibel, ‘State of the Dublin Core', D-Lib , April 1999: available via http://www.dlib.org/dlib/april99/04weibel.html.

1.Description and relationship to ICM.

Metadata Attributes =

Data about data, i.e., data which identifies or categorises other data. These are the data structures and data entry actions and technologies which enable us to store, recall and disseminate information. The attributes cross refer to the data needs for storage or memory, our technological capacities to assign data, and data about the action, structure, content, and context of the communication itself.

What does descriptive metadata do? It

Metadata has existed for a long time in many forms. Language is a form of metadata, when defined as a set of rules for articulating feelings and ideas for communications and use. There have been attempts to pursue this line of thinking by using computing to create a lexical database for English, by designing a system of relational ideas based on the use of words. WordNet at the Cognitive Science Laboratory at Princeton University has developed a semantic network. See: http://wordnet.princeton.edu/.

Another metadata project related to words is Lexical FreeNet, where users can specify whether they want words that are related in meaning to other words, different in meaning, words that rhyme and have similar meaning, and words that are spelt in a similar manner. See: http://www.lexfn.com/.

Other forms of popular metadata include abstracts, citations, barcodes, ISBNs, personal names, coats of arms, flags, ID numbers, signatures, PINs.

Metadata means many things to many people, including

These categories overlap. The focus in this topic is on descriptive metadata.

Metadata initiatives are responding to increasing opportunities for information accessibility and transacting business in distributed networked environments. Metadata adds value to data, which we use to categorise and manage information (‘Categorisation/Metadata') in the ICM. The ICM also highlights the systems and tools which we use to communicate (‘Technology'), linked to metadata. The Model draws attention to how these factors operate and inter-relate within and across four Dimensions -- Create, Capture, Organise, and Pluralise. It is possible to categorise the roles of information professionals across the Dimensions:

Each of these roles can be assisted by the use of metadata.

Where is descriptive metadata captured and stored? It can be found:

The emergence of generic and sector specific metadata sets requires assessment of some of the technological issues relating to attributing and managing metadata in the electronic world, and the strategic alliances within and between metadata communities. The metadata initiatives are occurring at a time when IT professionals, librarians, information managers, knowledge managers, cultural heritage players and other stakeholders are working together to develop coherent information architecture and metadata regimes to support

  1. Document management,
  2. Document discovery, and
  3. Document delivery,

in electronic networked environments. Although the main impetus for these developments has related to information sharing and accessibility, there is an emerging imperative to develop architecture and regimes that will support and evidence the transaction of business in networked environments. Thus far the main national and international efforts aim to build a global infrastructure of rules and standards in the virtual world which is equivalent to the regimes which manage recorded information of all kinds in the paper world in order to provide better access to information.

Systems which parallel in the virtual world of cyberspace, the kind of rules and protocols we are familiar with in the paper world, are beginning to emerge, e.g., the global information community's initiatives relating to developing a consensus on metadata regimes to help manage and make accessible document-like objects (digital identifiers as part of Digital Information Objects) in distributed networked environment, particularly the Dublin Core project which defines a common core of metadata. Work is being undertaken by specific information communities to standardise specialised sets of metadata that are interoperable with common core sets.

At the same time, the increasing use of the Internet for electronic commerce, e-government and business activity of all kinds requires the development of new policies, frameworks and structures to regulate and facilitate business activity on the ‘net. It has caused radical change.

2. Functional requirements for accessibility in networked environments.

So far much attention has been paid to the development of standardised metadata sets to enable information resource discovery and retrieval, i.e., the emphasis has been on information sharing and information access. Related developments in the recordkeeping sector in Australia are focusing on what metadata regimes are needed to support and evidence the transaction of business in a virtual world. It is useful to explore what principles drive current national and international initiatives to establish controls over metadata management in distributed networks within organizational and global domains.

Reflection

Reflect on the following attempt at a preliminary statement of the functional requirements for accessibility in networked environments. Keep these requirements in mind as you look at the Examples of national and international metadata initiatives and consider which of the functional requirements they address.

What does descriptive metadata aim to achieve? Here is a list of some aims:

It is easy to relate this list to parts of the ICM.

A Few Functional Requirements for Accessibility in Networked

Environments.

Visible You know what is there.
Searchable What is there can be searched.
Retrievable What is there can be found and seen.
Useable What is there is complete, accurate and reliable; it represents itself identically to any user every time it is retrieved; and its meaning is clear. It can be re-used.
Available What is there can be delivered to whoever is authorised to access and use it under the applicable regime of access and user permissions, wherever it is located and wherever the user is.
Restrictable What is there is only retrievable, and available to authorised users under the relevant terms and conditions; the system secures information from unauthorised access.
Interoperable

What is there is visible, searchable, retrievable,

useable, available and restrictable

through common user interfaces.

NOTE: In order to meet some of the general requirements outlined above, further specialised requirements might need to be developed for different kinds of documents. For example, the functional requirements for useable records are different from those for other kinds of documents.

3. Review of metadata and technology features highlighted by the Information Continuum Model.

What are the aims of using individual metadata elements? They can be applied for:

Before considering the examples of metadata-related initiatives, you might find it useful to refer again to the Information Continuum Model, particularly the third and fourth dimensions, and the metadata and technology attributes.

The Third and Fourth Dimensions

3D Organise

The dimension in which captured communicative actions are organised to meet the needs of an organisation or information community, including the materials and systems involved, the identification and categorisation schemes used, the structures in place, and the way an organisation or information community draws from, and contributes to, stored memory.

4D Pluralise

The dimension in which captured and organised communicative actions are brought together and shared beyond and among organizations and societies, including the materials and systems involved, the identification and categorisation used, the structures in place, and the way in which organisations and societies draw from, and contribute to, stored memory.

Metadata Attributes.

Element (used at Individual Level).

Aspect of a communication that identifies or categorises it at a level meaningful to individual participants, but not necessarily involving group consensus.

Controlled Element (used at Collaborative Level).

Aspect of a communication that identifies or categorises it at a level of meaning which commands consensus within a collaborating group.

Corporate or Organisational Domain.

Aspect of a communication that identifies or categorises it at a level of meaning which commands consensus throughout an organisation or an information community with a particular realm of interest or knowledge.

Societal or Global Domain.

Aspect of a communication that identifies or categorises it at a level of meaning supported by consensus beyond individual organizations or information communities.

Technological Attributes (e.g., Networks).

Available communication and information technologies in terms of systems and materials which produce communications. These attributes cross refer to technical means of structuring and assigning metadata, the techniques and media for storing, recalling and disseminating memory, and the technologies for communicative actions and structures.

The communication and information technologies encompassed by the model are analogue as well as digital. They are the means: the end is communication whether local and immediate, or across space and time.

Instrument (used at Individual Level).

Systems and materials which enable individual participants to engage in communicative action.

System (used at Collaborative Level).

Systems and materials which enable communications within a collaborating group.

Organisational System (operating at Corporate Level).

Systems and materials which enable communications within organisations and information communities with a particular realm of interest or knowledge.

Inter-organisational System (operating at Societal Level).

Systems and materials which enable communications among organisations and information communities.

Reflection.

Reflect again on the metadata/categorisation and technology attributes represented in the Information Continuum Model. In relation to the metadata attributes, note in particular the emphasis on ‘consensus'. Particular importance resides in consensus about the assignment of data identification and categorisation tags.

When looking at the examples of collaboration in relation to the development of metadata regimes, pay special attention to the ways in which ‘consensus' is being negotiated by the various information/metadata communities involved. Consider how, in order to provide for the maximum effectiveness of organisational and societal memory, the attribution of data within data structures to ‘document-like information objects' that capture communicative transactions of all kinds needs to be controlled.

But also keep in mind how these controls and the consensus they reflect help to build ‘structures of remembering and forgetting' as we explored in the Topic on ‘Memory'.

4. Some examples of international and national metadata-related initiatives.

What are metadata schema? Metadata schema define the meaning, structure and syntax of a set of listed metadata elements. In addition, metadata schemas often specify schemes which define metadata element values.

In the following Examples we will consider efforts to:

Identify and reach agreement on generic or core sets of metadata elements for attribution to all DIOs in distributed networks for information discovery purposes.

-- Example 1: Dublin Core. Standardise sector specific sets and ensure their interoperability with generic sets.

--Example 2: Australian Government Locator Service. Develop interoperable recordkeeping sets.

-- Example 3: National Archives of Australia Recordkeeping Metadata Standard for Commonwealth Agencies.

-- Example 4: SPIRT Recordkeeping Metadata Project.

Before exploring the Examples, read/browse through the following material for an introduction to metadata, the metadata communities, and library and

recordkeeping perspectives on metadata issues and initiatives.

Readings.

Meta Matters: http://www.nla.gov.au/meta/. This site provides an introduction to metadata and the Australian metadata community.

Debbie Campbell, ‘Dublin Core Metadata and the Australian Metaweb Project', Sept 1999, http://www.nla.gov.au/nla/staffpaper/dcampbell1.html.

Sue McKemmish, Adrian Cunningham and Dagmar Parer, ‘Metadata Mania', paper presented by Sue McKemmish and Adrian Cunningham to the ASA Conference in Fremantle, August 1998, http://www.sims.monash.edu.au/research/rcrg/publications/recordkeepingmetadata/sm01.html.

4.1.Example 1: Dublin Core .

The Dublin Core initiative aims to establish a generic metadata set to be applied to all DIOs on the Internet. This core set is designed to be embedded or persistently linked to atomic level document-like information objects. Its primary objectives relate to information resource discovery and interoperability, i.e., improving search capability in global networks.

The set is deliberately designed to be simple, flexible, and 'extensible'. This means that each of its 15 elements can be extended by adopting specialised sets of metadata elements to provide more information, eg the basic subject descriptor could be extended by using Library of Congress subject headings, provided these were standardised in such a way that they were Dublin Core compliant.

An associated project is the development of the Warwick Framework in which generic and cross-sectoral specific metadata sets can be applied.

Reading.

Check out the Dublin Core site: http://purl.oclc.org/dc/.

Check out how metadata is defined in this initiative.

Dublin Core elements Version 1.1 [ANSI/NISO Z39.85-2001]

http://dublincore.org/documents/dces/ .

The complete set, including all legal DC terms:

http://dublincore.org/documents/dcmi-terms/ .

Look in particular at the definitions of the 15 elements of the Dublin Core:

Note who is involved in the Dublin Core metadata community and the kinds of strategic partnerships involved in the development and implementation. Note the recommended implementation strategies, including the deployment of technology.

Read Stu Weibel, ‘State of the Dublin Core', D-Lib , April 1999: available via http://www.dlib.org/dlib/april99/04weibel.html , for information on the state-of the-art with the DC initiative.

4.2. Australian Government Locator Service.

Information locator systems provide knowledge structures for representing, identifying, locating and delivering information resources. The National Archives of Australia is the lead agency for the development of the Australian Government Locator Service (AGLS), an outcome of the work of the Information Management Steering Committee. This Office of Government Technology committee recommended frameworks for government information policy and the deployment of technology into the 21st century including the development of a Government Locator Service (see AGLS Victoria: Metadata Implementation Manual http://www.egov.vic.gov.au/Victoria/StrategiesPoliciesandReports/Reports/AGL S/aglsvic.htm ).

What does AGLS do? It:

The objectives of AGLS relate to promoting accessibility of government information and enabling individuals and organisations to transact business electronically with government agencies at all three levels, and to support the related initiatives in the Investing for Growth package of 1997, and its successors (http://www.noie.gov.au/projects/egovernment/Archive/online_survey/ProgressReport2000/major_ini.htm).

A key part of the AGLS is a standard set of metadata to be attributed to all Australian government documents made accessible through the Internet. In spite of the overall objectives of AGLS, the AGLS metadata specification is essentially an information discovery, retrieval and delivery set.

Reflection.

Browse the AGLS Usage Guide of the Metadata Element Set , the Australian Government Locator Service (AGLS) manual (available at http://www.naa.gov.au/recordkeeping/gov_online/agls/metadata_element_set.html).

Check out how metadata is defined in this initiative. Note its objectives and compare these purposes with the objectives of the Dublin Core set. Look in particular at the 19 metadata elements that make up the standardised set developed by AGLS:

Additional elements:

Note the additional elements:

Check out the definitions of these elements, noting that the set includes the 15 elements that make up the Dublin Core plus ‘function', ‘availability', ‘audience' and ‘mandate'. Note the interoperability of the AGLS set with the Dublin Core -- the strategy followed was to ‘inherit' the DC set and extend upon it.

Note:

1.that AGLS elements can be attributed to (used to describe) documents, agencies and government services -- does the DC set have a similar application?

2.that AGLS elements can be attributed to any level of aggregration of the resource being described (in any particular instance, the ‘type' element enables specification of what type of entity is being described, at what level of aggregration) -- how the ‘relation' element allows you to link together the information resources you are describing.

Note who is involved in the AGLS metadata community and the kinds of strategic partnerships involved. Note the recommended implementation strategies, including the deployment of technology.

4.3.Example 3: N.A.A. Recordkeeping Metadata Set.

The National Archives of Australia is currently developing a minimalist set of metadata to be attributed to all Commonwealth records in current recordkeeping systems. The set has a number of common elements with the Australian Government Locator Service set and the stated intention is to be

compatible with AGLS, as well as meeting the requirements of the Australian Records Management Standard, AS 4390. See the Australian Government Locator Service (AGLS) set, which is largely Dublin Core with an additional ‘Function' element, available for download here: http://www.naa.gov.au/recordkeeping/gov_online/agls/metadata_element_set.html .

Reflection.

Read/browse through National Archives of Australia, Recordkeeping Metadata Standard for Commonwealth Agencies: http://www.naa.gov.au/recordkeeping/control/rkms/contents.html.

Check out how recordkeeping metadata is defined in this initiative.

Reflect on the objectives of this set of metadata and compare them with the objectives of the DC and AGLS sets.

Check out the definitions of each of the elements.

Identify which elements are similar to/overlap with the elements in the DC and AGLS set.

Consider whether the set is ‘interoperable' with the DC and AGLS set.

Note the kinds of strategic partnerships involved in its development and implementation.

Note the recommended implementation strategies, including the deployment of technology.

4.4.ARMS: Recordkeeping Metadata Project.

The main deliverable of the 1998-9 Strategic Partnership with Industry -- Research & Training (SPIRT) Support Grant, 'Recordkeeping metadata standards for managing and accessing information resources in networked environments over time for government, social and cultural purposes' (Chief Investigators: Sue McKemmish, Monash University, and Ann Pederson, UNSW; Partner Chief Investigator: Steve Stuckey, Australian Archives) is the Australian

Recordkeeping Metadata Schema. The follow-on project is Clever Recordkeeping Metadata, or ‘Create Once, Use Many Times - The Clever Use of Metadata in eGovernment and eBusiness Processes In Networked Environments' (at http://www.sims.monash.edu.au/research/rcrg/research/crm/index.html). The Schema provides:

a standardised set of structured recordkeeping metadata elements,

a framework for developing and specifying recordkeeping metadata standards targeted for application in specific sectors,

a framework for reading or mapping other metadata sets in ways which can enable their semantic interoperability,

a classification of recordkeeping metadata according to functionality of purpose,

input to an Australian Framework Standard for Recordkeeping Metadata,

input to research and development in the broader metadata community nationally and internationally.

The follow-on research project is developing a proof-of-concept prototype to demonstrate how standards-compliant metadata can be created once in particular application environments, then used many times to meet a range of business purposes. The prototype will be implemented in a test-bed site to provide a model for best practice.

Reflection.

Browse the following Web Site for more information about this project: http://www.sims.monash.edu.au/research/rcrg/research/spirt/index.html.

Check out how recordkeeping metadata is defined in the initiative known as the Australian Recordkeeping Metadata Schema.

Reflect on the objectives of this set of metadata and compare them with the objectives of other sets.

Identify which elements are similar to/overlap with the elements in the DC, AGLS and NAA sets.

Consider whether the set is ‘interoperable' with the other sets.

Note the recommended implementation strategies, including the deployment of technology.

5. Resource Description Framework and Metadata Registries (Inter-operability).

There is a wide range of reasons as to why it is worth standardizing:

Recent years have seen the evolution of technical standards and frameworks to ensure that metadata sets can be interoperable to the maximum degree possible. A key to their interoperability is a mechanism which will enable like concepts expressed with different metadata tags to be identified, mapped and linked.

While mapping and 'crosswalks' (as they are known) between metadata sets is certainly a major plank in the intellectual framework needed to allow this level of interoperability, actual implementation of this depends on consistency of presentation and structure of the tags, as well as the language used within them. Consistency of structure and making sure that there is a universally understood way of locating tags is the purpose behind the development of the RDF (Resource Description Framework).

RDF is a new framework developed by the World Wide Web Consortium (sometimes known as W3C). W3C is a major player in the Internet environment, not least because one of its participants is Tim Berners-Lee, the architect and developer of the initial world wide web structures. W3C became interested in the whole area of metadata through its work with the PICS protocols. PICS (or Platform for Internet Content Selection) was all about mechanisms to filter material, creating selective views. In developing this controversial area of metadata mechanisms, the W3C devised the notion of ratings associated with information resources. These ratings could be applied for a variety of purposes, not just filtering content, but also quality and status of documents. This led them into the world of metadata and seeing the buoyant activity occurring in this arena in turn led to the recognition that different communities were defining metadata sets and structures that might not be interoperable. The RDF framework is the result of work to overcome these problems with interoperability of metadata sets. It is about imposing interoperable structures on metadata -- not the content of the metadata nor how that content is expressed. As such, it provides a standard which is gaining rapid acceptance and most metadata projects will be compliant with the RDF framework in the very near future.

If you would like to explore this Framework further, check out the following reading and web site:

Eric Miller, 'An Introduction to the Resource Description Framework,' D-Lib Magazine , May 1988, at http://www.dlib.org/dlib/may98/miller/05miller.html.

Resource Description Framework Site: http://www.w3.org/RDF/.

6. Metadata registries.

A few metadata specifications in the networked world have been described. The Dublin Core set is designed to be a generic set of metadata; whereas AGLS and the recordkeeping metadata sets are defined as sector specific, that is, they are metadata sets that have been adopted for a particular community in order to satisfy the metadata needs of that community. Other examples of sector-specific sets are the sets for educational material proposed by Educom in the USA and EdNA in Australia; and the set for geographical information, the Australian ANZLIC set. See: assorted metadata standards linked to the EdNA site: http://www.edna.edu.au/edna/go/pid/1.

Each set has been developed to meet the special needs and specific requirements of a community. A major challenge for the metadata community as a whole is: How do we know what metadata sets there are and which ones to apply in a particular context?

An answer emerging relates to the establishment of metadata registries. A metadata registry is defined as 'a formal system that records the semantics, structure, and interchange formats of any type of data'.

It is envisaged that metadata registries will be nominated bodies who are responsible for the management of the specific metadata sets. They will be charged with ensuring that the set they are responsible for is maintained and kept up-to-date and with being the authoritative source for information about the specific community set. In this way, we can see that OCLC would be the appropriate body to ‘register' the Dublin Core and NAA the AGLS set. The concept seems sound, but, at this stage, it there are no firm specifications of what a registry might do or standardised ways of mapping between metadata sets. It is also envisaged that all the metadata registries would be linked to one site, forming one central node for access to all metadata sets.

A 1997 Workshop on Metadata Registries defined the functionality required:

Registries of multiple sets of metadata may include formal mappings of overlapping metadata sets where appropriate. One of the essential characteristics of a metadata registry is the existence of a formal Authority agency that manages development and evolution of the registry and is responsible for policies pertaining to contents and operation of the registry. An attribute of each set of metadata should be the authority agency for that particular set. In addition, a metadata registry should include rules for extensibility if such rules are defined. A metadata registry should include sufficient information to facilitate conformance to the standard metadata that it includes.

(Joint Workshop on Metadata Registries, Workshop Report, July 1997).

7. Issues with metadata creation and use.

How the metadata specified in standardised sets can be associated with information objects has been explored in the Australian metadata community through initiatives such as the MetaWeb project of the National Library of Australia ( Guidelines for the National Library of Australia, http://www.nla.gov.au/guidelines/metaguide.html ) and the Distributed Systems Technology Centre (DSTC).

Linking metadata to the object it depicts can occur within the document, through the establishment of metadata repositories which maintain metadata separate from the object, or by encapsulating the document with metadata. The balance between immediate implementability and the long term durability of various options is also being explored.

What do you do with metadata?

Once there is agreement within a specific community on a metadata set, how do we apply it? At present most of the metadata ascription is a tedious business of filling in the fields manually. Nobody thinks that this is brilliant. But are there any alternatives to tedious manual processes of humans attaching metadata tags? And what do we do with the tags once allocated - how do we search the metadata?

At present there are a number of options for managing metadata and the web-object it describes. The AGLS User Manual makes note of three major possibilities:

1.Embedded metadata: where metadata is included in the 'header' information of the document. This information is not presented to the viewer as they normally access web documents, but can be seen by viewing the full text surrounding the document. Here the tags are formulated within html (hypertext mark-up language), using the tag. Using this strategy, the metadata becomes a part of the resource itself. As it notes in the AGLS User Manual, one current problem with this approach is that all resources need to have HTML tags, even where the resource itself is not in HTML form which restricts its application to text based resources. The expansion of the framework to include XML and the RDF framework will overcome some of HTML's restrictions because those languages and structures will deal with other types of resources, such as audio or video, as well as text.

2.Metadata repositories: databases into which the descriptive elements are placed. The repositories are external to the resource. The resource sits in one place and its description sits in the repository. The two are linked. The AGLS User Manual favours this approach, suggesting that its flexibility and capacity to cope with change is a major advantage. We look a little further at metadata repositories in the next section.

3.Metadata in document management, records management and other information resource management software. As the ALGS User Manual acknowledges, document management, records management and other information resource management software actually contains heaps of metadata associated with individual records. These systems are actually repositories of data which need to be kept associated with the object itself.

At present, this type of systems link is not particularly well articulated.

Activity.

Check out a few Websites to see if they include metadata tags.

Click on the View menu, and select ‘View source' and click to look at the metadata (this should work with both Netscape and Internet Explorer).

We suggest the sites referred to in these Topics, but you might also like to try a state or Commonwealth government site and something else, like Dilbert or a sports site or newspaper site.

Compare the metadata you find, and see if you can work out what they are doing to manage it in line with the technical management options described above.

8. Automating metadata creation/capture.

There are a variety of software tools which have been devised to commence the process of automating the collection or even attribution of metadata. The MetaWeb project is an Australian project which has been involved with devising and evaluating such metadata creation tools. The generic types of software identified by that project are:

  1. Gatherers: tools that sweep across web pages and are programmed to 'pick up' and copy all metadata contained within nominated framing tags, at present usually those elements of the document header that are between the tags. The gatherers then dump the data back into a nominated place, usually known as a 'repository'.
  2. Brokers: tools which process the data behind the scene and index metadata within the repository, so making it available for future use. Brokers are an intrinsic architectural element of distributed computing. Brokers and resolution services have a close relationship.
  3. Repositories: metadata repositories are the currently favoured architectures into which the metadata garnered by the gathers gets placed. Repositories are databases which consist of the metadata only, with links to where the document can be found. The documents themselves don't live within the repositories.

This class of software tools is generic to the computing architecture of the web. The tools are devised to work across platforms and can be used for a variety of applications, of which metadata management is just one.

There are a number of prototype tools for actually creating the metadata for web documents. Two types of tools are commonly discussed, metadata generators and metadata editors. Metadata generators extract metadata from existing html-encoded documents and place the data elements within the META tags. Editors are tools which provide a template for entering new metadata content. The supporting software converts the elements from the template into the META tags.

Browse.

Metaweb Project: http://www.dstc.edu.au/Research/Projects/metaweb. This site contains information about metadata initiatives and tools.

Metaweb Project: Analysis of Metadata Creation Tools: http://www.dstc.edu.au/Research/Projects/metaweb/toolpost-orig.html.

9.Strategic directions; the future.

There are a number of aspects of metadata under intense scrutiny at the moment. These include:

Related Activity:

Select ONE of the example metadata initiatives in Section 3.

Analyse this initiative in terms of the Pluralise Dimension of the Information Continuum Model, with particular reference to Metadata Attributes in the Societal or Global Domain and the Technology Attributes of Inter-Organisational Systems.

You should address:

  1. the organisations, groups or people directly responsible for the initiative.
  2. the scope and coverage of the metadata set.
  3. the objectives and purposes of the set.
  4. the metadata community/communities involved.
  5. relationships with other metadata communities.
  6. the strategies involved in achieving consensus on the metadata set.
  7. the strategic partnerships forged in its development and/or implementation.
  8. the interoperability issues -- what frameworks, tools and techniques are
  9. used to ensure interoperability?
  10. the recommended implementation strategies, including the deployment of technology.
Graeme Johanson, 21 September 2005.