UMSLG lecture on datasets

Lecture to UMSLG: 12th November 1990

I've been asked to bring you up to date on the ISI deal and also to tell you a little about what lies ahead in the hope of stimulating some discussion. The first thing I want to do before bringing you up to date is to stress that what we are engaged in is a giant experiment, one of the largest experiments ever conducted, but none the less an experiment. In terms of size its also a world first and I hope that at least some of you share my excitement at the possibilities which are being opened up.

As you will know, the deal was set up through a piece of opportunism about a year ago and after the data had been bought there was a call for tenders for sites to run the data. Three were received and that from the University of Bath accepted. They are to network the data on JANET using STATUS as a software frontend.

Through a variety of accidents Peter Stone and I had finished up as the library advisers of the CHEST team and we made it clear from the start that we wanted this to be the first of a whole series of data deals bringing information directly to the end user and by-passing the intermediary in the library or computer centre. When the Computer Board realised that it had unwittingly taken a radical initiative it decided to set up the National Task Group on Datasets which was retrospectively to rewrite the record to show that this was simply a development of existing policy rather than an innovative step. The report of that group which included Peter and myself was taken to the Computer Board last week and I'll come back to its reactions in a few minutes.

Meanwhile, back at Bath it has been realised that they have taken on a major undertaking. Two groups have been set up, a steering group which is to try and monitor the experiment and report to the Computer Board on how any future initiatives should be managed in the light of the Bath experience. Then there is a working group which is composed of representatives from half a dozen test sites and which is helping in the design of the user interface. Finally a meeting will soon be called for next month partly with the purpose of setting up a user group to make sure that the thing actually works.

The service is scheduled to start on 14th January and so far 47 sites have signed up. A newsletter has already been issued on JANET and it is intended to use that as a means of communication. There is a mailing list based on the mailserver at Newcastle which anyone can join to keep up to date. As part of the project the Bath team will be producing a training pack to train the trainers within the next couple of months. The general idea is to spend the first six months of next year reaching key product champions in each institution and making sure that there is a cadre of trainers on each site. The real mass input is then likely to be with the 1991 intake of students. Of course any site can go faster if it is able.

However, to revert to the idea that this is an experiment a whole host of practical questions remain to be answered, some for the community, some for the site and some for the Computer Board. For example,

- Is the model of using Bath the right one, or do we want several sites

- Should we buy time on commercial services

- Do we want Status or BRS or neither

- what will be the effect on ILL (in the US requests have doubled)

- How will the average university train 6000 people by next November and 2000 a year thereafter

- Has the computer centre the ability to register 6000 people for e-mail

- how many concurrent users can you handle

- what changes will you have to make in the registration process

- how will you balance resource between CD hardcopy and this

- have you done anything about the institutional IT strategy

Its not good enough just to drop this chatrge on the library which then cancels the hardcopy. That may be ok for an isolated item but this is the start of a trend and what happens when the bill gets to 50,000 a year rather than 5000

I say this is the start of a trend, but we have hit the first potential hiccup. The Board met last week and approved the report of the National Datasets Task Group but has allocated no money immediately. Next year the Computer Board merges with the UFC and will be given responsibility for library computing. They are about to set up a committee to examine library computing needs. A couple of years ago that would have confined itself to library housekeeping systems. Now it has to worry about international telecommunications needs and information services and data. This will no doubt take quite a few months to report and so although the Board is committed to spending up to half a million a year on data, it wont do it for a while. Of course this does not prevent CHEST negotiating deals which we can be asked to find 100% funding for.

So what is in the pipeline. Discussions are going on with INSPEC, BIOSIS, COMPENDEX and MEDLINE. From next month STN will be connected to something called IXI which in turn is connected to JANET. This will remove all telecommunications charges although there is no saving on data.

There are severe problems with Medline I'm sorry to say. NLM is happy to arrange a deal, although it would prefer not to network the data in the UK until it is networked in the USA. There is also the issue of the British Library service and its close relations with NLM. However NLM has no option but to sell the data and as it is at the top of everyone's list thanks to the noise you have all been making, we can expect it to be bought by the Board. However, because of the complications I have been describing it would be fair to say that the pessimistic assumption is that the data will not be available until late 1992.

One other possibility exists which I would like your advice on. Excerpta Medica is about to launch a new abstracts service using the data which goes into Excerpta, but issued as a sort of weekly current contents but with abstracts. For reasons I won't go into given the time available it may be possible to get this free early next year, with the chance of carrying it on in subsequent years at a very cheap price of a hundred or two pounds a year. But is there a market. They claim not to compete directly with Medline, to have reduced overlap and to be particularly strong in drug and pharmacological areas. So we can have that early next year, eighteen months to two years ahead of a possible Medline deal but with no idea how the Computer Board will react to the notion of buying another medical related database. We are assured it wont prejudice a medline deal - but who knows. I and CHEST will be very interested to hear your view of priorities.

So in closing, can I re-emphasise a couple of points. Firstly we are all guinea pigs in a giant and exciting experiment. Secondly we are not well prepared yet and have a lot to do and at least part of that lies in alerting registries and computer centres to the time-bomb ticking away. Thirdly at least in my opinion we are entering a quite new world where the role of the librarian will change fundamentally to that of information broker and information skills teacher to the masses and not just the few who darken our doors, what has been called the library without walls.

As Woody Allen said in his address to the graduates

We face a world full of pitfalls and yet a world full of opportunities. The secret of our success will be to avoid the pitfalls to seize the opportunities, and yet still get home by six o' clock.