First, thanks to Alex for pointing this out. I’ve been reading a lot about Wolfram Alpha and wondering what it is, and now we can see.
While I can’t class myself as a naysayer, I do see that this project does not really overlap with librarianship (or it doesn’t have to, anyway). From what I see of Wolfram Alpha, it is designed to be a giant “answer machine” which will respond to natural language questions to provide someone an answer.
With libraries and especially library catalogs, they are designed with different purposes in mind. The library catalog is *supposed* to allow users to get an idea of what are the intellectual holdings of a specific collection. Therefore, it should allow someone who doesn’t know what is in the local collection, to browse *intellectually* the contents of that collection. The system was designed originally to work with printed book catalogs and then was transferred to cards. Therefore, if someone browses a concept such as, Caesar, Julius, they should see a number of things that they would never have thought of. Here is the search for Caesar, Julius in the Princeton University catalog (I hope this works):
and through the subdivisions, you see concepts such as “Adversaries,” “Cult,” “Death and burial,” “Language” and so on. This is what is I mean by an “intellectual browse:” the user can get an idea of the richness of a specific collection, and become aware of things that he or she would never have thought of independently. This goal is quite different from Wolfram Alpha (an answer machine).
The traditional library methods can be very powerful (even though it can be done through physical means through the arrangement of cards) and, I suspect, this power is what Bernard has in mind when he asks whether browsing is necessary or not. I have not seen any system that attempts to replicate this automatically anywhere, except possibly Vivisimo, and the results are bizarre (IMHO).
Now, does the library catalog succeed today? Back in the old days, transferring the original methods from the book catalog to the card catalog resulted in several problems. Then, transferring that same browsing capability onto the web has not succeeded (and probably will not succeed) in my opinion. If it is assumed that we need to retain the power of traditional browsing–which is my hope–different methods must be found.
There is a more important question though: is the access provided by a library catalog still necessary today? Especially with competition such as Wolfram Alpha out there, where people will tend to start with more and more often (or to Google-type search engines) and they will not continue to seek out our tools, so consequently, ours will remain ignored. Obviously, I think our tools are necessary but those tools, and librarianship as a whole, must be reimagined, repurposed, redesigned, and rethought in all kinds of ways. Essentially, we must find how to fit ourselves into the larger world of information instead of waiting for everyone else to come tp us. For example, we should be thinking how could we work with Wolfram Alpha whether we like it or not?
But I’ve been over these points in many other messages.
>�Count me in as a Wolfram Alpha naysayer. A “computable almanac” isn’t
>�that useful, and its not a “step” to anything. Seemingly minor
>�increases in the intellectual complexity of a question require
>�astronomically better data, data models and algorithms, if they are
>�even possible to “compute.” The answer, if there is an answer, lies in
>�Norvig’s “unreasonable effectiveness of data,” not in any
>�combination of “curated” data and “millions of lines of
>�Just an opinion!
>�On Sun, May 3, 2009 at 9:13 PM, Alexander Johannesen
>�> Another nail in the library coffin, especially the academic ones ;
>�> � http://www.youtube.com/watch?v=5TIOH80Qg7Q
>�> Organisations and people are slowly turning into data producers, not
>�> book producers