We had a crack at using the SOLR spell check handler, but the quality of the results just wasn't that great. Admittedly, we didn't spend a great deal of time on it, but that's just because we're lazy :)

In the end, we created a specialised aspell dictionary based off all the non-numeric terms in our Title, Author, Subject and Publisher indexes. We found this worked pretty well, as the main issue we had with the spell check was giving suggestions for people's names and such. I knew that creating an extra dictionary would be pretty easy, and the results are much better.

Definitely not as neat a solution as getting the solr spell checker to work, but good enough for the time being :)

On 27/11/2008, at 2:15 AM, Andrew Nagy wrote:

Sory Daniel - the spell checker problem has been much more of a problem than expected and will take some more time to figure out the solution.  Hopefully we can spend some time in the next few weeks to get it solved, but there is no quick fix.  Other than enabling the old spellchecker.


On Mon, Nov 24, 2008 at 12:49 PM, Lovins, Daniel <daniel.lovins@yale.edu> wrote:

Hi all.


I realize that the Solr spellchecker issue is being tracked on jira: http://vufind.org/jira/browse/VUFIND-13


and that it's slated to be fixed in RC2 …


but could anyone suggest some kind of RC1 workaround that would allow us to get it working this week? I ask this because we're in an evaluation period at Yale, and having the spell-checker in place would make a big difference.


Thanks for your help.




This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
VuFind-General mailing list


Steven McPhillips <smcphillips@nla.gov.au>
IT Business Systems
National Library of Australia
Try our new catalogue - http://catalogue.nla.gov.au