https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6863
Priority: P2
Bug ID: 6863
Assignee: [email protected]
Summary: weight message body words according to visibility
Severity: normal
Classification: Unclassified
OS: Linux
Reporter: [email protected]
Hardware: PC
Status: NEW
Version: 3.3.2
Component: Learner
Product: Spamassassin
Spam that tries to defeat bayes learning sometimes appends volumes of normal
text to the spam at the top of the message. It seems as though the bayes
scoring should somehow weight larger those text tokens that are likely to be
seen by the user: i.e., the first blocks of text/plain bodies, or (if SA can do
so) top rendered+visible parts of html.
--
You are receiving this mail because:
You are the assignee for the bug.
