Hello, I have a question regaring retraining and groups. I have setup my primary account to be classification group for everyone else, so if a user's account is not mature enough then it looks at my account to work out if the mail is spam/ham. Now, lets say one of these users is getting a lot of spam. They can retrain all these spams but dspam still will look to my account as their account is not mature enough to learn on its own yet.
What's the best way to solve this? I am thinking, get the user to forward their mail to me then I can retrain that as spam. But that would mean that my account would probably think that any email from that user is then going to be spam (I think). So how can I work around that problem? One final question. In the docs, it says that an account is considered mature if it has more than 1000 hams or 250 spams. Well, there is an account that has almost 1000 mails, but dspam_stats shows this: TP True Positives: 35 TN True Negatives: 986 FP False Positives: 2 FN False Negatives: 39 SC Spam Corpusfed: 9 NC Nonspam Corpusfed: 0 TL Training Left: 1512 SHR Spam Hit Rate 47.30% HSR Ham Strike Rate: 0.20% OCA Overall Accuracy: 96.14% It says it still has 1512 mails to go before it's passed its learning stage? I actually came across an account that had more than 1000 hams and it also still had training to do. Thanks again for help. Am using latest Dspam, v3.8.0.
