W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > March 1999

Re: word counts

From: Al Gilman <asgilman@iamdigex.net>
Date: Sun, 14 Mar 1999 10:10:02 -0500
Message-Id: <199903141506.KAA180966@relay.interim.iamworld.net>
To: WAI-ER <w3c-wai-er-ig@w3.org>
At 11:39 AM 3/14/99 +0000, jonathan chetwynd wrote:
>> 3   word counts in meta tags.
>
>I had been trying to implement a browser, linked to a database, with
>lookahead. It is a search aid, but not in the sense of altavista, much more
>a browser.
>It does not use text, anyway not more than 30 words per page.
>However it is very difficult to find pages on the www that fit this
>criteria. Pages need to be accessible to every variety of ability.

If every browser is running its own search indexer, then we rapidly exhaust
the communication capability of the Web.

On the other hand, if we get word count into the indexing databases of the
existing search services, and in the advanced preferences section of their
query formulation interface, you have a service which will search according
to the preferences of your client group and it is updated by processes that
are already working.

To sell the attributes into the search databases, we need to find
attributes which give them a competitive edge with some larger market
segment.  That may take some thought, but in fact I expect that the unmet
needs of non-readers is a gold mine of market edge discriminators, just as
the unmet needs of blind or deaf consumers is a rich lode of ore in which
to find ways to make the user interface better.

The other intermediate step is to prototype the indexing and querying
strategies with a pilot-scale robot activity.  But that is exactly what the
Web Characterization activity in W3C is about to do.  We need to get our
robot experiments on their agenda.

Al
Received on Sunday, 14 March 1999 10:06:51 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:10:32 GMT