W3C home > Mailing lists > Public > www-annotation@w3.org > January to June 2002

RE: Search engines vs. annotations

From: Laurent Denoue <denoue@pal.xerox.com>
Date: Wed, 23 Jan 2002 11:48:39 -0800
Message-ID: <C988C2B023D2D511BC7900A024E0371C023145@mailbox.pal.xerox.com>
To: "'Uldis Bojars'" <captsolo@pop3.pid.gov.lv>, www-annotation@w3.org
Hello,

1- if you have a list of annotation servers and you know their protocol, you
can get the list of annotations
2- if annotations are stored and published in a specific file format (you
thought about RDF-XML),
then a search engine could crawl the Web to find them: Google does that with
PDF documents
and Cora is a search engine to find research papers.
3- if annotations are embedded into Web pages (for example an annotation
would contain an XPointer),
then you would have to crawl the Web and get only those XPointers. Search
engines could easily do that.

I personnaly like the 3rd solution: it lets users embed annotations in their
current web pages: annotations
do not have to be stored on a server or in a separate annotation file.
Furthermore, you could send search engines the URL of your Web pagse
containing annotations so they can directly process and include them into
their database...

I wrote a small proposal a few years ago about using extended URLs to encode
an annotation
(see
http://www.univ-savoie.fr/labos/syscom/Laurent.Denoue/publications/TR1999-01
.pdf)

The idea was quite similar to what was also proposed later by Phelps and
Wilensky for robust hyperlinks
(see http://www.cs.berkeley.edu/~phelps/Robust/papers/RobustHyperlinks.html)

I can see uses of requesting annotations left by people on a specific text
in a Web page (like a book you are about the buy on Amazon.com). What were
you motivations? Do you have usage scenarios in mind?

Laurent.


> -----Original Message-----
> From: Uldis Bojars [mailto:captsolo@pop3.pid.gov.lv]
> Sent: Tuesday, January 22, 2002 10:43 PM
> To: www-annotation@w3.org
> Subject: Search engines vs. annotations
> 
> 
> Hello!
> 
> The ability to annotate any WWW resource is wonderfull, but
> I have a question (possibly a FAQ):
> 
> Is there a way to search for all web pages, which match certain
> conditions AND have annotations [on a set of annotation servers that
> this search engine accesses]? I.e. something like Google, but with a
> flag in advance search "Is annotated".
> 
> While the amount of annotated pages is small compared with the total
> amount of the information on Web, I would very much prefer to be able
> to filter out pages that are annotated as opposed to look at all pages
> and see if they have annotations.
> 
> P.S. The same question in broader scope would be: is there a search
> engine [like Google] to find all resources with RDF metainformation
> that matches selected criteria (i.e. describes object Person from a
> certain schema, etc.)?
> 
> -- 
> Best regards,
>  Uldis                          mailto:captsolo@pop3.pid.gov.lv
> 
Received on Wednesday, 23 January 2002 14:48:44 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Friday, 25 March 2005 11:19:17 GMT