Friday, March 15, 2013

Gmail now explains why an email was spammed






If you use Google and its assorted products every day, and you are the least bit like us here at Geek.com, you relish the idea of a peak behind the curtains. After letting us watch a search engine quality meeting last week Google, seems to be in a particularly generous mood. Next stop on the transparency train is Gmail, which now explains why a message was spammed.
Above you can see a few sample message from Google regarding the algorithm’s reasoning for sending something to your spam folder. None are particularly insightful, in fact they cover the basic ground for why any email would be spammed: it’s something you’ve manually picked out as spam, something the filters have previously identified as spam, something new that uses a word or phrase that is typically found in spammed messages, or it is an email (or is like an email) that the crowd has identified as spam.


Google will also spam something if it appears to be an unsafe email or a phishing scam, as seen above. It’s worth noting that of all the phishing scams I received not all were marked with this message, so Google seems to be examining a number of signals within each email and then tagging the the one that is the strongest or that was flagged first. Something might also be marked as spam if it came from an unconfirmed sender.

These are all that I’ve been able to find, but there could be others (post them if you find them).
So why give people, potentially spammers, insight into the reasons why messages are spammed? My guess would be that Google is confident in its spam filtering system and that they fully realize a false positive is much more of a problem for people than a spam message appearing in their inbox. If this is the case, then it’s in their interest to ensure that real emails can be tracked and despammed. And let’s not forget that only you have access to the messages in your spam folder and someone that was spammed for you might not be for another user. So if you are seeing false positives, this gives you a way to identify and remedy that — either by creating a filter or hitting “Not Spam” until the system learns your preferences.

No comments:

Post a Comment