bug-gne
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug-gnupedia] Re: [Bug-gnupedia]Data Model


From: Patrick Gearman
Subject: [Bug-gnupedia] Re: [Bug-gnupedia]Data Model
Date: 17 Jan 2001 23:09:32 -0800

On Wed, 17 January 2001, Bob Dodd wrote:

> It strikes me that one of the first things we need is some way to wrap
> encyclopedia entries to give cataolgers some chance in the future.
> 
> Thinking of what I would need to help catalog the information, I'd want
> to get at: title, abstract (if it's a long entry), authors/main
> authors, citations, references to external peer reviews, keywords,
> synonyms, target audience/assumed knowledge, publication date, source
> journal if this material has been published before, copyright text. Oh,
> and the content itself, mustn't forget that...

I'd drop the target audience/assumed knowledge category. It's too subjective 
based on the knowledge of the poster of the article. Everything else looks
to be very good choices, although for keywords, if it hasn't appeared in the
first X hundred words of the article (X = 2 - 4) then it is probably only 
tangentially
related to the article.
 
> Since the approach seems to be web-based, it would seem to make sense
> for our HTML to in fact be XML (or at least to be wrapped in XML), and
> to use XML DTD entries to describe the fields we need.
> 
> Not that I would expect any "normal" contributer to know XML, but
> perhaps we could offer a simple tool to enter the information (e.g. a
> small Web page hosted at GNU?) that would generate the XML wrapper for
> the contributer. We could go as far as to use that wrapper tool as part
> of a "submissions" page rather than submitting by e-mail.

Personally, I'd prefer to see a submissions page, but if what the site
experienced after the post on /. is any indication, that page could get
easily overwhelmed. You want to insure that any material on that page
(like the article) is not lost if it does not go through the first time due to
overuse.
 
> Such a tool is needed relatively soon if we want to capture this sort
> of information a standardised way; the more entries that are sent
> ad-hoc, the more difficult it will be to go back and fix this later. I
> know cataloging is much further down the line, but we need to get the
> information format, and the content itself, standardised from the
> beginning, or our catalogs will be no better than the current web
> search engines.

True. Although that could be a function of the highest level of editors.
Some editing will be necessary, if nothing else, then for spelling and
grammatical errors.

Also, just to avoid the kind of crapola that is rampant on /. and other boards,
there has to be a system wherein junk articles are removed. Whether this
is accomplished through a system of editing out posts, negative reviews,
whatever, it need to be done. There is little point to an online encyclopedia
full of "Furst posst on Nuclear reactors!"

Just my 2 shekels.

Patrick Gearman


Find the best deals on the web at AltaVista Shopping!
http://www.shopping.altavista.com



reply via email to

[Prev in Thread] Current Thread [Next in Thread]