Re: Bogus Search Results

[Golda Velez <gvelez@iwhome.com>]

You can try using the new (slightly experimental) RankHits.pm module from 

	http://webglimpse.net/allusers/rankhits-beta.tar.gz

Using that module, you can customize the ranking formula, using # of times the
keyword appears, whether it is in the title, age of the page, and a few other
factors.

To completely eliminate the footer hits, you'd have to write a filter to delete
the footers on the fly before indexing.

--G

On Fri, 11 Aug 2000, Brauchli, Stephan P wrote:
> We are having some issues with the results WG returns.
> Lets say I do a search for 'employer', the first result shows like this:
> 
>                    TWC to Present $367,270 Grant to Enhance Skills, Jul 13
> 2000
> 
>                         line 89:About TWC | Job Listings | Worker Services |
> Employer Services | Welfare Reform | Education/Training 
> 
> As you can see WG found the word 'employer' in the footer as a link and so
> returned this file as the first result.
> I know there are other files such as an employer home page where this word
> appears multiple times.  
> WG does not seem to care it just looks at its index file and returns any
> file (in seemingly a random order) that contains the keyword.
> 
> Has anyone had the same problem and resolved it?
> Is it possible to exclude footer information from the search
> (wgfilterindex)?
> 
> I have tried adding META keyword tags to certain pages and had those rank
> first.
> 
> Thanks,
>    Stephan