Search fix for words that are "too common" - phpBB3

Author Comment
User avatar

Posts: 2263

We have sports boards, and on our baseball board, we're unable to search for the team's best player because both his first and last name are "too common." Is there a fix for that?

As an aside, "photobucket" falls into the "too common" category on this board.

User avatar

Admin

Posts: 11070

ACP > General (tab) > Server Configuration > Search Settings > "Common word threshold"

A higher number will allow more occurrences. Zero will disable the feature.

The Search Index needs to be rebuilt after any change: Maintenance (tab) > Search Index > "Delete index" (then (re)"Create index").

Thanks for the heads-up on Photobucket. I think I should disable common words here. For your board, your database will expand horrifically in size. You may wish to consider offering a Google site search at the bottom of pages. Rebuilding the search index should be done at a quieter time of operation, e.g. overnight. For your board, it will certainly take hours and may take longer than a day.

User avatar

Admin

Posts: 11070

At 40 posts per second, that would take just under 6 hours. Of course, the rate will fluctuate and probably safe to say that it should be complete overnight. Sometimes indexing stops, and a page refresh is required to restart it. (The pop-up window does nothing really, so you can close this.) If that fails to kick it back into action (after a pause), then go back and click create again and it will continue from where it left off. It may cause a little slowness.

User avatar

Posts: 2263

Thanks!

When you refer to disabling, is that the same as setting the common word threshold to something other than zero?

User avatar

Admin

Posts: 11070

No. I would guess it is similar to a setting of 100%, if that were possible. I honestly do not know how best to suggest you set it to achieve the result you wish. A higher number may work and zero certainly will work. Talking about a day between trials and the amount of resources used, hardly makes trial and error a good option.

As phpBB search is not that great, you might find a Google site search better and it certainly uses less of your resources.

User avatar

Posts: 2263

The trial and error aspect occurred to me immediately. :) It's possible that a very slight bump from zero could address this; it's also possible that a significant one would be necessary.

How would we keep a Google search "under control," i.e. keeping it out of the members-only forums while allowing it to search the fully public ones?

Could we use more than one Google search, e.g. one at the forum level and one for the entire board (public forums only).

Also, there'd be the matter of subforums. We have a members-only subforum of a public forum.

User avatar

Admin

Posts: 11070

Sorry, I did not realize you already had a setting of zero. In that case, if common words are still being found, your search needs reindexing.

This is a Google site search for the keyword "Jim". http://goo.gl/RVx1p

It follows bot rules, so whatever your permissions are for bots, it follows them.

Display posts from previous:  Sort by  



Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group