> On 8/31/07, James Weinheimer <j.weinheimer_at_aur.edu> wrote:
> > The opposite case of CS ignoring the experience of library science is just
> > as prevalent.
>
> Why? What has LIS got that the rest of the world could benefit from?
It's nice to think that there is no need to learn from others, and they just need to learn from you. With such an attitude, there is nothing at all to learn from the people who have been organizing information for the past 2000 or so years.
> > I still say that the old standards of finding resources *consistently and
> > reliably* by its authors, titles, and subjects is still what most people
> > want.
>
> Hmm. I constantly deal with people who searches that neither of these
> three can help me. Why are these three what we should aim for? (Apart
> from it being traditional)
I never claimed that this is all that people want, but they do want to know these things at a minimum. The new tools have not made the result of the question "What do you have by Fyodor Dostoyevsky?" completely obsolete.
> > When this can be proven to be done--using scientific methods of expert
> > analysis, control groups, and so on--then I will be convinced.
>
> Don't we already do well on title, author and subject searches? *puzzled*
No. You don't do well with these types of searches, as I have tried to show in many messages on this list. But, it depends on what you mean by "well." Does the search work and not break in the middle? Yes, the search functions. Do you get a result? Yes. Is the result completely off-base? No, many of the results may have something to do with the materials someone wants.
But, when you search on google for the works of Leo Tolstoi, do you really think you are looking at the works by and about Leo Tolstoi that are available in Google? If so, then you are wrong. This I can prove. Can a system be built that will allow for all works by Tolstoi to be found? I grant the possibility, but it hasn't been built yet.
Show me how we can do a concept search that works. No theory, no trusting, no shoulds, no ifs, no maybes. Show me a tool that searches full text and can do what the traditional catalog at least attempts to do: to search the *concept* "World War I." I have *never ever ever* seen it. Not even close. These things search text, not concepts.
Or lacking this, convince us that our users no longer want to search concepts. I submit they do, and the Semantic Web is all about precisely that. Do you think people no longer want this? If so, what is your evidence?
Regards,
Jim Weinheimer
Received on Fri Aug 31 2007 - 18:12:14 EDT