False positives, even though I train as good...

I hadn’t pruned or deleted my corpus for years and there were some irregularities with Spamsieve so last week I decided to start anew and I deleted the corpus.

Ever since, there have been a couple contacts of mine that repeatedly get classified as Spam no matter how many times I train them as ‘good.’ I actually went and trained more than 30 past emails from those people as good yet the new ones always end up in my ‘Junk’ folder. What’s weird is that they are never highlighted. It’s like Spamsieve recognizes they are not spam but still puts them in the junk folder.

Any suggestions?

You can check to see what SpamSieve predicted for these messages by looking in the log. My guess is that your mail program moved them directly to the junk folder and never asked SpamSieve to analyze them. So you should check the setup in your mail program to make sure it’s as described in the manual and that no other rules are set to move messages to the junk folder.