Oliver Cole wrote:
>Hi!
>
>I've recently got back into Piggy Bank now that I've got Firefox 1.5.0,
>and I also discovered Solvent. The idea behind Solvent is so cool that I
>almost regret submitting the first bug on the project :(
>
>Inspired by something I saw last time I used Piggy Bank, I wanted to
>write a screen scraper for UKC faculty... turning pages on:
>
>http://www.cs.kent.ac.uk/people/index.html
>
>into FOAF.
>
>UKC also has the module web pages... unfortunately the external view has
>virtually no useful information, which may prove to be a practical
>problem later. Anyway, take a look:
>
>http://www.onesteprevolution.com/temp/CO522.png
>
>The three staff links at the top all link to the same people pages that
>are linked to from above, and hence, once I've finished my people page
>-> FOAF scraper, I wondered if it was possible to begin inferring
>http://purl.org/vocab/relationship/worksWith from the module web pages?
>
>Or do I have the wrong idea?
>
>
You have the right idea--that of pulling information together from
several web pages and several web sites :-) But right now we have not
yet implemented RDF inferencing in Piggy Bank so you have to do the
"inferencing" yourself during scraping. That is to say, right now it's
not possible to scrape only
A teaches course X, B teaches course X
and then specify the rule
If M teaches course P and N teaches course P then M works with N.
Rather, your scraper has to produce these statements
A teaches course X, B teachers course X, A works with B.
David
Received on Sat Dec 10 2005 - 13:58:01 EST
