[MKSearch-dev] Future directions

[Chris Croome]

Hi

On Thu 10-Nov-2005 at 10:08:06AM -0000, Phil Shaw wrote:
> 
> Mark assumes our ultimate target is free text search. My feeling is to
> keep the focus on structured metadata because it has explicit meaning,
> and this allows us (ultimately) to make better use of the semantics
> and inferencing provided by Sesame.

Yeah, there are a lot of tools out there that do full text searches, I
don't know if trying to compete with them is worth while...

> The other is the idea of a personal aggregator. MKSearch works just as
> well for intranets as for the public Web, but could also index
> metadata in Open Office and PDF, possibly Microsoft Office documents
> on the local file system.

That would be cool, I'm not sure of the need to be able to index a file
system, I think OS specific tools is probably the way to go for this,
but if you want to add support for this I wouldn't object...

Being able to get the metadata from file formats other than xml/html
ones would be good, there is also the metadata in jpegs.

> Another thing I have been thinking about is to create a public
> demonstration service that indexes the UK Information Asset Register:
> 
> http://www.opsi.gov.uk/iar
> 
> http://www.opsi.gov.uk/cgi-bin/searchIAR.pl?DB=iar
> 
> That also has a category-based browse interface,  which Adam has
> talked about in a different context.
> 
> The query/results system is designed so that results can be cached
> server side for improved performance.

Sounds good :-)

Chris

-- 
Chris Croome                               <chris at webarchitects.co.uk>
web design                             http://www.webarchitects.co.uk/ 
web content management                               http://mkdoc.com/