Understanding Trends Across a Community
Back at UKOLN’s IWMW 2006 event Andy Powell gave a plenary entitled “Reflections on 10 years of the Institutional Web” in which he summarised the trends he had observed during the ten years in which UKOLN had been hosting its annual event aimed at members of institutional Web management teams.
Andy used the JICSMail archives for the web-support list in order to search for early occurrences of technologies (such as RSS and CMS). This, however, was a time-consuming process; due to the lack of APIs to the JISCMail Web archives and anyone wishing to apply further analyses would have to start from scratch.
The Role of RSS
We first provided a news page about the event at IWMW 2005 and ensured that this was also available as an RSS feed. We later realised that RSS could be used not only for providing news but also as an open format which would allow content to be reused by other applications and since then have provided a wide range of RSS feeds for the events, including details of the speakers and abstracts of the plenary talks and workshop sessions.
But what might we learn if we make available RSS feeds for even further back? Might this approach make it much easier for those who wish to gain a better understanding of how the topics addressed at the events have developed over the years? This was the question we were looking to answer.
The Community Now and Then
A Wordle display based on the RSS feed of the abstracts of the parallel workshop sessions at this year’s IWMW 2010 event is shown. We can see that topics such as ‘mobile’ and ‘location’ will be addressed in a number of the sessions. It is also interesting to see that ‘HTML5′ and ‘RDF’a’ also feature fairly prominently in the display as does ‘social’ – which relates to social networks and not the social aspects of the event. I should also add that this data is based in 17 parallel sessions which last for 90 minutes, from which participants can attend two sessions.
But what was being discussed back at the IWMW 2000 event? I have created an RSS feed of the abstracts of the workshop sessions for that event, which was held at the University of Bath 6-8 September 2000. The format of the event was slightly different back then as, for IWMW 2000 only, participants could choose one from parallel sessions which lasted for three hours or two from four sessions which lasted for 90 minutes.
The Wordle display based on the RSS feed for the IWMW 2000 parallel workshop sessions is shown. This time we see that there is a strong interest in CMSs, which seems to have disappeared from this year’s event. There was also an interest in ‘VLEs’, ‘payment’, ‘ecommerce’ and ‘security’ which, again, do not seem to be being addressed this year (or, to my recollection, in recent years).
A better understanding of changes since 2000 would be seen if redundant words (such as ‘institutional’, ‘web’, ‘management’, ‘workshop’, ‘sessions’ and ‘participants’) which probably occur in every abstract were removed. And in additional to the graphical capabilities provided by Wordle I wonder if more sophisticated text mining tools could be used to explore the changes in the topics which the community has been addressing over the 14 years which the IWMW event has been held.
We have RSS feeds containing providing information on the plenary talks and workshop sessions for IWMW 2000-2010 together with biographical details for the plenary speakers and workshop facilitators since the event was started. Note that the RSS feed for the plenary speakers contains geo-location information of the host institution which means that we can also display a map showing the location of the host institutions of these contributors.
I have some ideas of how this structured data about the event (as opposed to HTML pages designed for reading by human) could be used – and it is useful to have data available for use with tools such as Yahoo Pipes. I also wonder if others have any suggestions on ways this data could be used?
I also feel that other events should be providing RSS feeds of their event information in a similar fashion – especially those events which are well-established within the community. If the abstracts of the talks given at events such as national ALT, JISC and UCISA conferences held over the years were provided in RSS this should provide a valuable open and reusable resource to facilitate data-mining. And what about the Eduweb series of conferences which provide a similar role to IWMW for university Web team in US higher educational institutions? Shouldn’t such high profile conferences aimed at technically advanced user communities be taking a lead in providing structured information about the events? Isn’t there a danger that in only focussing on the future we fail to learn lessons by looking at out past?