> CNN.com: Information Visualization
>Article
>
> A very *timely* article published on the heals of my "New Age
>Navigation: Innovative Information
>Interfaces for Electronic Journals" article in _ The Serials
>Librarian_
>
>
>[http://www.public.iastate.edu/~gerrymck/NewAge.pdf ]
Thanks Gerry for a wonderfully comprehensive overview!
(this in no way detracts from your article, but I did note
having to wait over a minute for my first peek at it,
since the entire PDF object has to download before it can be
viewed! Dare I mention the magic phrase: meta-data? )
>http://www.cnn.com/2004/TECH/internet/01/05/seeing.search1.ap/index.html
>]
>
>NEW YORK (AP) --As wonderful as Internet search engines are, they have
>a pretty big flaw. They often deliver too much information, and a lot
>of it isn't quite what we're looking for.
These examples all start from the premise of searching in effect free
unstructured text, and trying to find patterns in the results.
I did not scan each in detail, but was surprised no-one has mentioned
one of the most interesting and "available" applications of the
Semantic Web (because surely this is where this is all heading?);
RDF, disguised as RSS, as a mechanism for finding the particular
nugget one is looking for.
RSS is a RDF-based site summary mechanism, which implements (or should)
proper Dublin Core metadata, but more importantly, it can carry
the content model for specific subjects as XML,
and thus renders itself able to
be scanned semantically on specific information objects
(and not just full text; we call these fuller objects datuments
rather than "full text" to emphasize the difference).
Again I did not immediately spot any overt Chemistry in
Gerry's article or the others mentioned above (which summarises
the difficulty which these articles allude to!). Using full RSS
alerting, one need have no difficulties. We have created
an RSS feed with chemical semantics
(http://www.ch.ic.ac.uk/rzepa/rss/) which enables the user
to specify tasks such as "alert me to any newly published
source which relates to molecules containing the element
Chromium" and for good measure, display the 3D
coordinates and other chemical metadata such as
formula of such molecules along with the free text.
(PS You will not actually find the above example in the
URL I cite because the article containing it is being
published in an American journal which, prestigious as it
is, rules that prior publication on the Internet invalidates
it for publication in their journal. So I cannot send any article
URL as a "public" statement, but could certainly discuss
this example individually with anyone interested!.
You could also try connecting to the
RSS feed at http://www.ch.ic.ac.uk/csdemo/feed.php
and see what happens (but to extract the chemistry, you will need
additional software!)
--
Henry Rzepa. Imperial College, Chemistry Dept.
+44 0778 626 8220 +44 020 7594 5804 (Fax)
|