[thelist] PHP Search Engine
Pringle, Ron
RPringle at aurora-il.org
Thu Mar 16 16:01:32 CST 2006
At least when I check the search reports, I won't have to wonder why
everyone stopped searching for "garbage" and started searching for
"Public Library"!
Actually, I think the single quotes are throwing that search off. If you
put it in double quotes to indicate a string, it finds the right page as
the first result. If you remove the quotes completely, results 1 and 2
are pdfs and the third is a the correct result.
I agree that you need to be careful when choosing a search product. And
I agree with Max that user behavior makes a BIG difference.
For instance, I generate a report that shows search terms entered that
returned no results. Many of them are misspelled words. Zoom lets you
create an index that maps common misspellings to their correct spelling.
I think that any search engine will probably require ongoing effort and
time in tweaking the content and results to give the most favorable
outcome. I try to do that with our site, but given that I am a web team
of one, I have limited time to devote to that, unfortunately.
Still, going from our old non-semantic tables based site with no search
at all to a semantic site with a decent search feature has improved the
average user's experience by leaps and bounds.
Ron
> -----Original Message-----
> From: thelist-bounces at lists.evolt.org
> [mailto:thelist-bounces at lists.evolt.org]On Behalf Of Paul Bennett
> Sent: Thursday, March 16, 2006 3:43 PM
> To: thelist at lists.evolt.org
> Subject: Re: [thelist] PHP Search Engine
>
>
> Good point Max,
>
> There *is* a difference between my testing and general user
> behaviour, but don't you find a search engines inability to
> find a page with an H2 of directly-quoted text a bit disconcerting?
>
> My 'tests' aren't always as arbitrary as this. I've been to a
> few site search software sites and entered header text like
> 'product information' and gotten no results, despite the fact
> that they had a page on their 10 page site with the header
> text 'Product Information'.
>
> My point was that search software is an interesting beast. It
> not only has to be easy to install and a good indexer of
> content, it also needs to return *high-quality results* to the user.
>
> Try this search: (Sorry Ron! :) )
>
> http://tinyurl.com/elxot
>
> 'Public Library'
>
> Results: 0 :(
>
> Removing the quotes gives better results, but this to me is a
> pretty basic failure of a lot of search software.
> Or am I just being picky?
>
> Paul
>
> -----Original Message-----
> From: thelist-bounces at lists.evolt.org
> [mailto:thelist-bounces at lists.evolt.org] On Behalf Of Max Schwanekamp
> Sent: Friday, March 17, 2006 10:28 AM
> To: thelist at lists.evolt.org
> Subject: Re: [thelist] PHP Search Engine
>
> Paul Bennett wrote:
> > Just a note to BEWARE of a LOT search engine software.
> > Case in point: - (No, I'm *not* picking on Ron here, just
> illustrating a rant er... Point)
> > http://tinyurl.com/o56lx
>
> Not to dispute your point Paul, which I am not qualified to
> do anyway,
> but context and user behavior should be taken into account when
> evaluating any software. When I changed your query from searching an
> exact string using quotes to a group of keywords, and
> removing the "City
> of Aurora - " part of the query (after all, why would a user
> enter that
> when they're already on the COA website?), the correct page was #1 in
> the results. E.g.:
>
> http://url123.com/dmna5
>
> Of course your query should have worked just fine, but Zoom
> isn't quite
> so bad as your demo might lead one to believe.
>
> --
> Max Schwanekamp
> http://www.neptunewebworks.com/
More information about the thelist
mailing list