M<Galeti a écrit : > Hello Guys, > > I am trying to realize if most of you work in companies like where I work. > Usually anyone works in my company during the weekends and our customers and > supplier ""usually"" not work to. Then I started to think: who will send > email to us in weekends? I started to check logs and saw 70%, 80% of emails > coming in weekends are spam (in my case). > > Is this something that listen good to you? I mean: this makes sense to you > too? Maybe score messages in weekends with high points? >
you could, but you need testing to find a score that doesn't cause FPs (you may also need to exclude some portions of the mail flow, such as mailing lists or people you know send over the week end...). note that the day "name" is already seen by Bayes. and while it is hard to quantify its "action", I'd say Bayes is probably safer here. > Only thinking...... > you could do the same for the hour of the day... > Have a nice weekend... > M<Galeti >
