I’ve been noticing comment spam recently in my blog. I suppose it was only a matter of time before the spammers discovered this new form of obnoxious behavior. A quick Google search this morning lead me to two MT Plug-ins from the same person.

First up is a plug-ins that displays an image with a security code that must be entered to verify the commenter is human. This seems easy but has some issues in accessibility; the blind and people using text only browsers. The latter group I personally don’t care much about. If someone can’t be bothered to get into the 21st century and use a graphical browser I’m probably not overly interested in what they have to say. 🙂 The first group though could be a problem, if I have some blind visitors and they want to post a comment.

The second option is a plug-ins that uses a Bayesian filter to block spam. This option removes the accessibility problems but does require a training process to teach the filter what constitutes spam. Also this solution probably will require routinely updated training sessions as spammers learn ways around the Bayesian filter, just like in email.

Right now I’m leaning towards the first option and then if I get any email from a blind visitor I’ll think about implementing the Bayseian filter.