Thanks Mark
> If I search for 'K' I get the 3 current entries for Kew Gardens,
> because it recognises (or assumes) the Index herbariorum code for Kew.
> Very clever - I might have assumed it would find all entries with K in
> them!
> If I search for 'kew' I get no results. It assumes this must also be
> an IH code. Not so good. Some naive users may assume Kew data isn't
> present at this stage. Our fault for having such a short name I
> suppose.
This was a last minute 'feature' that seemed like a great idea at the
time. It may still be a great idea if I knock it down to 2 letters. If
we still get problems I'll take it away altogether.
You can now search for Kew and find CABI - progress!
> If I leave the Natural Language Search option on, and search for 'kew
> gardens' I get a three-page list of entries. I can't easily tell why
> the non-Kew entries appear.
>
> If I remove the Natural Language searching on the latter, I get the
> results I would expect. Based on this slender evidence, I would prefer
> Natural Language Search to be 'off' by default.
The natural language search is voodoo supplied by the MySQL database.
When it is good it is very good. It works on the frequencies of the
words in your search string compared with the frequencies in the
records.
I think you may be right that it would be better to turn it off but
the code that does that is a little more convoluted so I'll raise it
as a bug for fixing when I am working on that area next and can think
it through properly.
Thanks for your contribution,
Roger