Lars Stavholm wrote, on 16. mar 2007 13:43:
[...]
> In addition, just looking at an excerpt from a debug message:
>
> [snip]
> [burton] [1.000000] ^M*Office (1frq, 129s, 0i)
> [burton] [1.000000] ^M*software (1frq, 122s, 0i)
> [burton] [1.000000] ^M*Acrobat (1frq, 119s, 0i)
> [burton] [1.000000] ^M*Premiere (1frq, 118s, 0i)
> [burton] [1.000000] ^M*Suite (2frq, 118s, 0i)
> [snip]
>
> I guess this might hold information that I need. Can anyone tell
> me what the numbers in the parenthesis "1frq, 122s, 0i" means?
> frq is frequency I guess, but how about the other two?
Got a bit of it! 122s is 122 times occurrence of the chained token pair
as spam, 0 times as innocent, 1frq means that the token is unique, 2frq
means that it occurs twice, etc. If I search for the token
16463589117999081304 (1frq, 2s, 5i)in my MySQL DB (using phpMyAdmin, I
get (sorry for the folding, this is copy'n paste):
Full Texts uid token spam_hits innocent_hits last_hit
1 16463589117999081304 2 5 2007-03-17
--Tonni
-- Tony Earnshaw Email: tonni at hetnet dot nlReceived on Sat Mar 17 17:50:47 2007
This archive was generated by hypermail 2.1.8 : Mon Mar 19 2007 - 00:00:06 CET