Earlier this year DCMI, the Dublin Core Metadata Initiative, celebrated 15 years of Dublin Core. The UK higher education community has had a significant role to play in the development of Dublin Core, with colleagues and former colleagues at UKOLN having been involved since 1995.
Much of the discussions related to the development of Dublin Core standards and related activities has taken place on a series of JISCMail lists. But how has use of these lists developed over time? This is a question which relates to work I am involved in in exploring ways of analysing and interpreting data related to use of networked services. I have previous described trends related to the growth in use of Facebook within UK universities and have captured evidence of early use of institutional use of iTunesU and YouTube Edu in order that future analyses will have benchmark figures to make comparisons with.
In order to understand the trends in use of JISCMail lists by those involved in standardisation, deployment and use of Dublin Core metadata I used the DCMI’s list of mailing lists as my starting point. I used the JISCMail search facility to obtain information on the numbers of posts on each list by carrying out a search for messages containing an ‘@’ in the sender’s email address sent between 1990 and 2010. I have also included information on the date on which the mailing lists were established.
|List||Established||Total Nos. of Posts|
|DCMI General Mailing List|
The broadest of mailing lists related to the international Dublin Core effort. Unlike other lists, which relate to the tasks of specific working groups or special interest areas, this list is for discussion of all issues relevant to the development, deployment, and use of Dublin Core metadata.
|DCMI Architecture Mailing List|
This list, which supersedes dc-datamodel, dc-schema, and dc-implementors, is intended for discussion of a technical architecture for the Dublin Core.
|DCMI Communities Mailing Lists|
The DCMI Accessibility Community is a forum for individuals and organizations involved in implementing Dublin Core in a context of accessibility, with the objective to enhance interoperability of accessible resources through the use of Dublin Core metadata.
|Collection Description: This list is intended for discussion of issues related to the use of the Dublin Core (DC) for describing collections of resources.||February 2002||602|
|Education: Electronic discussion list to support the efforts of the international Dublin Core effort’s Educational metadata working group in exploring issues directly related to deployment of Dublin Core for the description of educational materials.||August 1999||689|
|Environment: This list supports discussion of deploying Dublin Core metadata in environmental applications.||February 2002||151|
|Government: This list is intended for discussion of the uses to which the Dublin Core Element Set might be put in describing government and public sector resources.||December 1999||501|
|Identifiers: A Dublin Core Metadata Initiative (DCMI) forum to discuss identifiers||October 2007||32|
|Libraries: A mailing list for the DCMI Libraries group focussing on issues from the library sector.||December 1999||433|
|Kernel: A list to support the work of the DCMI Kernel Group||September 2003||60|
|Knowledge Management:A forum to support the work of the Dublin Core Knowledge Management Community||December 2007||48|
|Localization and Internationalization: This list supports the efforts of the DCMI Localization and Internationalization group exploring issues directly related to deployment of Dublin Core metadata in multiple languages.||January 1998||275|
|Preservation: The DCMI Preservation Community is a forum for individuals and organisations involved in implementing Dublin Core metadata in a context of long-term digital preservation, with the objective to promote the application of Dublin Core in that context.||December 2003||191|
|Registry: The DCMI Registry Community is a forum for service providers and developers of both metadata schema registries and controlled vocabulary registries to exchange information and experience.||December 1999||661|
|Science and Metadata: A forum for individuals and organisations to exchange information and knowledge about metadata describing scientific data||February 2009||103|
|Social Tagging: Dublin Core social tagging discussion list||October 2006||157|
|Standards: List to support discussion on issues related to standarization of DCMI specifications||February 1999||107|
|Tools: This list supports discussion of building and using software tools related to the Dublin Core.||April 2002||61|
|DCMI Task Groups Mailing Lists|
|Collection Description Application Profile Task Group: A list to support work on the Dublin Core Collection Description Application Profile||January 2007||74|
|Kernel Application Profile Task Group: A list for developing the Dublin Core Kernel Application Profile||November 2007||127|
|Metadata Provenance Task Group: The list will support the Dublin Core Task Group on Metadata Provenance||June 2010||26|
|DCMI/NKOS Task Group: Dublin Core Metadata Initiative Task Group developing a Dublin Core Application Profile for KOS (Knowledge Organization System) Resources||August 2010||3|
|DCMI/RDA Task Group: List to support discussion on Resource Description and Access (RDA)||December 2005||532|
Note that these statistics were collected on Friday 11 December 2010.
The Need For Trend Analysis
Such figures are pretty meaningless taken in isolation. We might expect the general discussion lists to be more popular than more specialised lists, and well-established lists to have had more traffic than those which have only been set up recently. Of more interest should be the trends showing usage of the lists.
As can be seen the number of posts to DCMI JISCMail lists peaked in 2002 and has dropped sharply since. The number of lists has grown with sharp rises in 1999, 2002 and 2007. However the average number of posts to the lists has also seen a sharp decline.
The details for the individual lists are shown in the following chart.
I should also add that the data I collated in order to produce these charts is available as a Google Spreadsheet.
Why the interest in metrics on usage of mailing lists? In part such evidence can be used to identify whether technologies, in this case mailing lists, are still being actively used – as I described earlier this year in a post on The Decline in JISCMail Use Across the Web Management Community University Web managers seem to no longer be using mailing lists to the extent they did previously.
Mailing lists used to develop standards, as opposed to those used by practitioners to address routine queries, may be valuable for historical analyses, such as observing discussions on decisions taken. The MarkMail service, for example, provides access to over 36,000 messages posted on the W3C’s www-html list. But it seems that several lists, such as dc-identifiers and dc-knowledge-management, have failed to attract significant traffic, with only a total of 32 and 48 messages having been posted to these lists. Have the discussions taken place in other fora, I wonder?
It also seems to me that there is a need for popular services, such as JISCMail, to provide simple ways in which usage statistics along the lines I have illustrated in this post, can be produced. I wonder whether this needs to be done by developments to the JISCMail Listserv software itself or could be layered on externally? Any thought?