|
At one time I had asked around what others were using for
their max files. If I remember correctly, Fritz responded to me that he had
his set @ 3000, and considering his organization is about the same size as
mine, I went ahead and set it at that. For some reason later on, I decided to
up it to 30k. Seems to me, if you have too many, thus stale entries for
the current trend in spam, or too few, thus not a broad enough selection of
words/samples to get an accurate account of all current spam, then your bayes
filter is going to be much less accurate. If you have too large of a setting
for how much mail you process per day, then obviously it is going to take
longer to refresh it then a site that gets 10x the amount of daily mail. I knocked it back down to 3000 and deleted all old entries
until there were that many left in spam/notspam, and am going to watch the
results as emails are reported to the email interface. There has to be a general guideline for ratios to best feed
the bayes filter. Such as “10k mail per day, set to 3max files, 15k per
day 3500”, etc etc. Has anyone taken the time to test this? Perhaps if
people chipped in their configs that have been extremely accurate for them, and
their daily/weekly stats, then we could build listing for documentation and
come to a consensus of the best place to start for settings. I considered tweaking mine up and down once a month and
seeing how that goes, but that would take me months and months to tweak out,
and would only produce the ‘best’ config for my volume of mail. -Paul |
------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
