Ivan Trundle Manager, communications, systems and publishing

Sifting through the statistics

The last time I wrote about ALIAnet's log files was back in late 1998, when our website experienced an unprecedented growth from a paltry 500 pages to a staggering (back then it was 'mind-numbing') 3500 information-packed pages. We reported then that our employment pages were the most popular, followed closely by our e-list pages.

Things have moved on since that time - the static page count alone has now reached the 9000-page mark, and there are many other parts of the site that generate page views on the fly from a range of databases. It has become a busy place, and the task of co-ordinating it all, and tying it all together, has increased in complexity. There is still much more to do (and there always will be) in making the site not just the premier resource for members of the Association, but indeed the premier online resource for the Australian library sector. Feedback to date (take a look at our guestbook, for example) shows that we are serving the needs of the members and the wider library sector, but this alone does not give us a true picture of how people use the site.

Our log files paint a very clear picture of how the site is used, and give useful insights into the machinations of the internet, too. For example, our latest log file [http://alia.org.au/alianet/logs/2001.03.html] shows that the primary web-based activity is that of external search engines caching material to serve through local proxies (plain English non-geek translation: companies like Google, Excite and AltaVista trawl the entire website and take copies of all of the data to store on their local hard drives in preparation for individuals to access the data through searches of the search-engine websites). Consequently, it is often rather more difficult to determine exactly what pages are of the most interest, since these search engines are generally less discriminating that the average user. However, once one looks more closely at the actual pages served, it is clear that a sea-change is occurring.

With the move to group-based representation within the Association, a good deal of web-based activity is taking place in the Sections and Special Interest Groups components of the site, and for once it is apparent that users are spreading themselves around the site more than ever before. This in itself is a good sign, as it shows a level of maturity in users which indicates that the ALIA 'home page' is not always the first port of call.

Some trends, though, have not diminished: our employment pages are still at the top of the popularity stakes, although the level of interest in our events calendar (served through dynamically generated pages from an internal database) indicates that our events calendar is the premier viewing location at the moment.

The recently introduced chat server is proving its worth, too - quickly rising to the top of the 'top-20' pile, ranking highly alongside such notables as our extensive links pages, education pages and our topical Library and Information Week pages. I was asked the other day about how members can 'book' the use of the chat server for activities: there is no need to book at all - just dive in and use it at will (though it helps to organise a date and time beforehand unless you prefer an existential experience). The CPD group has used it most successfully quite recently, and other groups of the Association are planning similar events in the near future.

But allow me to return to some statistics for a moment. I have mentioned that our website contains over 9000 unique static pages (and the data consumes around 150 megabytes of disk space, for those who are interested in archiving issues). Interestingly, on a daily basis, roughly 140 megabytes of data are transmitted in the form of just more than 10 pages to nearly 500 distinct hosts and, in any given month, around 4 gigabytes of data are transferred to users. In anyone's terms, this is a lot of information. It has undoubtedly contributed to a more open flow of information both between members of the Association and between the office-holders of the Association and members. If I were able to include data from the activity amongst our 60-odd electronic discussion lists, it is clear that information flow (some might call it information overload!) has increased significantly in the past few years alone.

We have no data to indicate how many members of the Association are able to access online material, but without a doubt the ratio of 'those connected' to 'those not-yet-connected' must be higher than in any other profession. However, we make no assumptions here - and will continue to offer alternative means of communication that do not involve being connected to the not-quite-all-pervasive internet.