On Fri, 1 Mar 1996, Fytton Rowland wrote:
> (1) The use of Alta Vista or other search engines *does* present a quality
> control problem: you find information on your subject of interest but you
> have no idea what (if any) form of quality control has been applied to it.
Which is very similar to my experiences with the paper based abstracting
services. I know when I started my PhD lit survey, when archie and
veronica were just being developed and the web consisted of a few tens of
servers, I followed lots of dead ends into crappy paper based resources.
The one thing this taught me real quickly was what was likely to be a good
resource and what was likely to be chaff. The paper based abstracting
services have abstracts from both the good resources and the chaff. Why?
Because my chaff is another man's good resource and vice versa. Of
course, following chaff on paper takes far more time and effort than it
does electronically.
> But information professionals
> (including those in training to become information professionals) still
> recognise the need for selectivity, and it is to be hoped that they will
> succeed in finding means of helping clients to cope with the information
> explosion even when those clients are at a remote terminal and are never
> seen face to face.
I think in the long run we'll need to let the end user's _software_ be
swamped with information and then let that do the sorting out of wheat
from chaff before presenting it to the user. Preselecting what people
will find interesting _is_ imposing views of what interesting is according
to the information professional onto the end user. The end user has to
have the tools that they can configure (or that can configure themselves)
to the user's preferences and tastes and I can imagine information
professionals lending a helping hand with new users in setting up initial
profiles and such like.
However in the short term, we don't have those tools and there isn't a
consensus on how they should work (if you don't believe me, pop over to
the agents mailing list and ask what they think an intelligent agent is.
The question appears with a periodicity of about two months and everybody
has a different answer). So to help the user out we've got databases
coming on stream with pointers to "quality" reviewed resources. These
might be the first port of call for a user, but I think its important to
recognise that end users are likely to use them in addition to and in
conjunction with the massive web indexing services like Lycos and Alta
Vista. The web indexes aren't going to go away tomorrow because they
fulfill a very useful role - they provide a much wide trawl over the net
and a much more timely indexing. People want this and we see this in the
number of accesses these services get.
> Having said this, to condemn the WWW because there is too much information
> of dubious quality on it is like saying that paper is a bad thing because
> pornography is sometimes printed on it.
Hear, hear. Ban paper I say :-)
Tatty bye,
Jim'll
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Jon "Jim'll" Knight, Researcher, Sysop and General Dogsbody, Dept. Computer
Studies, Loughborough University of Technology, Leics., ENGLAND. LE11 3TU.
* I've found I now dream in Perl. More worryingly, I enjoy those dreams. *
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|