(1) I want Mollom to remove/identify spam, but I also want to manually inspect and approve ham.

Background: I look after a Drupal site on which comments are currently pre-moderated: "post comments without approval" is unticked for regular users and we regularly check the approval queue to publish genuine comments and delete spam & comments which break the site rules.

I'd like to start using Mollom to automatically remove/flag spam and profanity, both to reduce our workload and to remove human error from the spam-identifying process.

However, I can't work out a way to incorporate it into our workflow. Spam and profanity are not the only issues we deal with; the site prohibits other types of comment. I want Mollom to deal with the spam/profanity, but still leave ham in the approval queue to await manual inspection.

Is that possible? Something like http://drupal.org/node/938846 would work, but that appears to have stalled. Is there another way to achieve what I need?

(2) I'd like to be able to see the comments Mollom discards.

Background: When first running Mollom, I want to ensure the false positive rate is acceptably close to zero. We have many users in countries which produce a lot of spam and I don't want them to be blocked for sharing an IP address with a spammer.

The normal way to do this is to tick 'Retain the post for manual moderation', but that results in spam going into the approval queue, which (as explained above) also contains our genuine messages. There seems therefore to be no way to check Mollom's false positive rate without ticking "post comments without approval", which we are unable to do.

Is there any way to inspect which messages are flagged by Mollom without selecting "post comments without approval" for users?

Thanks in advance for your help on these two questions!

Comments

On 1)

The module already allows you to combine the automated moderation with manual moderation. It actually is kind of the default configuration. Basically, you just enable the Mollom protection for a form, setting spam posts to be discarded. If you leave your comment user permissions as-is, then all new comments that were not blocked/caught by Mollom, will still end up as unpublished/unapproved on your site, ready for manual moderation.

On 2)

Unfortunately, at this point, the module does not support a "dry-run" mode yet. If that mode would exist, then I guess you'd additionally want something along the lines of #537008: Add debugging information to comments and nodes on top of it, so you were able to review the classification for every comment, before disabling the dry-run mode.

These two features are not really trivial and we haven't had time to work on them yet. The closest comparable functionality that exist, but which you already mentioned, is to set the form protection to retain spam posts. Implementing those features is not at the top of our priority list currently, but I guess there'd always be room for changing priorities, in case you'd be interested in funding the development. (If you are, then we can discuss in the Mollom support ticket, which has been created in parallel to this issue.)

Status:Active» Fixed

Status:Fixed» Closed (fixed)

Automatically closed -- issue fixed for 2 weeks with no activity.