A Free Idea: Search Engine for Webpages

Dave Brueck dave at pythonapocrypha.com
Mon Aug 26 16:15:12 CEST 2002


Serge Boiko <boiko at demogr.mpg.de> wrote in message news:<m23ct3kl3u.fsf at boiko_linux.demogr.mpg.de>...
> An example to clarify the point. I have a list of journals
> (1000+lines) and I'm looking for those of them that contain the term
> "interaction", but I'm really interested in those matches that relates
> somehow to computer science.

Hmm... is this a common scenario though? I guess it seems that
'interaction' is too vague a search term to begin with, and rarely are
searches based entirely off of just titles. Wouldn't any modern search
engine give you a leg up on this list by letting you search for
_documents_ containing, e.g. both the terms 'computer' and
'interaction'?

> The imaginary program would return:
> 
> Chemico-Biological Interactions
> Hyperfine Interaction
> Interaction and linguistic structures
> Journal of musculoskeletal and neuronal interactions (JMNI)
> *User Modeling and User-Adapted Interaction*
> Nuclear Instruments and Methods in Physics Research Section B: Beam
> Interactions
> Symbolic interaction (-1999)
> 
> I see that only *User Modeling and User-Adapted Interaction* is what I
> need; not necessary to press Ctrl-F anymore...
> 
> Another point is that I could make searches of subsearches, also useful,
> isn't it?

What I find useful is Google's presentation of cached content (the one
you get to by clicking a search result's 'Cached' link instead of the
main link presented) because it highlights in a different color each
word from your search. It gives you an easy way to quickly and
visually scan a document searching for relevant content. I just scroll
down the page looking for concentrations of pretty colors! ;-)

-Dave



More information about the Python-list mailing list