Maybe I’ve missed this tip but is there a way to set a default to automatically train on the “subject” instead of the “from” info. I rarely get messages from the same address but get tons with the same subject. Right now, including combinations of words from those subjects as rules in the the blocklist isn’t working.
This used to work almost perfectly but now it’s only catching about 1/10 of the spam. I keep bumping up the aggression level but this just isn’t as effective as it used to be.
Since the accuracy is so bad and the blocklist rules don’t seem to be working, the most likely explanation is that your mail program isn’t showing the messages to SpamSieve. It doesn’t matter what SpamSieve thinks or what the blocklist says if it doesn’t even see the messages.
So no default to train on subject then?
Yes, I had read the “Please Read This First” and I had also referred to the manual. I just reread the top post again to see if I missed something helpful but my set-up would appear to be okay-- of course, maybe I don’t know how to determine that: I have all three SpamSieve options available in my Apple Mail menu. Does that mean it’s set up correctly? SS is catching some words I’ve set Rules for but not others. It’s totally ignoring the “debt” words in Rules. Does that mean SS is NOT seeing my messages?
A few weeks ago when I first started trying some of the trouble-shooting tips, I cleaned out all the “from” addresses in my BlockList because it had thousands of them which wasn’t helping, and I started fresh. I re-entered the most common spam words as Any Character Set/Contains because Subject/Contains was missing them. It’s picking up some words but not others. This would appear it is seeing my messages but, of course, I’m not an expert on this program so maybe I’m wrong.
The person who recommended this program to me said he also has been getting tons of messages that SpamSieve is missing, starting just a few weeks ago. Maybe it’s our Macs that suddenly developed problems.
So, your answer to my original question is “no, it won’t train on subject line.” Training on the “from” address isn’t blocking spam because I rarely get the same address twice. I need to train on the subject and body words and since those words have to be entered manually, my using the option of Train as Spam from the Mail menu is giving me false expectations. I will concentrate on entering Rules for words manually and hope SS starts catching them.
This program has been an absolute lifesaver. I couldn’t believe how accurate it was after training and adding Rules, so don’t think I don’t appreciate SpamSieve. It was so good and saved so much time that I got spoiled. Now I’m fighting spam again and it’s very frustrating.
Thank you. To my mind, the “Please Read This First” post tells exactly what you should do in this situation—but of course it’s clear to me, because I wrote it. Since you don’t seem to have understood that, I’d be grateful if you could tell me what wasn’t clear so that I can improve it.
The “Please Read This First” post says:
The log is essentially a list of all the messages that SpamSieve has seen, and what it thought of them. If you click the link above, it tells how you can interpret the log. And if the log shows that SpamSieve isn’t seeing all the messages, there’s a link to page that tells how you can check and test the setup.
“Any Character Set” is not for matching keywords. It’s for matching the character set in the message’s Content-Type header. For example, if the Content-Type is:
Content-Type: text/plain; charset=“UTF-8”
The character set is “UTF-8”. There’s more information about this in the manual.
Second, you shouldn’t need to make blocklist rules for common spam words. The point of SpamSieve is to do this stuff automatically, and with a higher level of accuracy (and greater safety) than is possible with blocklist rules. No one should ever need to use the blocklist window directly, but it’s there because some people want a bit of manual control in specific situations. If you find yourself constantly tweaking the blocklist, something is wrong—probably with the setup or training.
Blocklist rules should be 100% reliable. If you make a rule to block a word in the subject, and a message contains that word in the subject, it will be blocked. If a rule doesn’t work on every eligible message, that’s evidence that SpamSieve isn’t seeing all the messages.
(The other possibility is that SpamSieve saw the message, but it didn’t use the blocklist because a higher priority filter decided that the message was good. The log will tell you when this happens.)
My answer is that it’s the wrong question to ask. You don’t need an easier way to add subject and body words. If it’s trained properly, you won’t need to add words. And if it isn’t seeing the messages, it won’t matter how many words you add because it will never catch any messages that it doesn’t see.
Instead, you should look at the log. There shouldn’t be any guesswork in this. If the log shows that SpamSieve saw the messages, then we’ll know why it got them wrong. If it shows that SpamSieve didn’t see the messages, then we’ll know that the problem is outside of SpamSieve and that you need to fix the setup in your mail program.
Thanks so much
Thank you for the detailed response and your patience. As with most of us, a little knowledge is a dangerous thing, and from my perspective, I think I understand it but I don’t. I’ll spend some time going through all your suggestions and the manual again tonight and try to make sense of it. It’s not rocket science, I’m sure, so I need to work a little harder on comprehension on my end.
I can’t wait to get this great tool working for me again. I really came to rely on it, and it had gotten so intuitive and perfect, it was a little bit scary.
SUCCESS! I deleted the old Spam mailbox and closed my Apple mail. I used the SpamSieve menu to reinstall the Apple plugin, opened Apple mail, created a new Spam mailbox and reset the SpamSieve rule under Preferences. It’s working beautifully just like it was a few months ago. Yippee!