Do you think you have a good HAM corpus?

[Giampaolo Tomassoni <g....@libero.it>]

Great!

So, would you please have a check to FPs produced by these rules:

body      __TRUF_1      m'\Wr.+servation\W'i
body      __TRUF_2      m'\W(?:carte\s+(?:internationales?\s+)?de\s+cr.+dit|carte\s+bancaire|visa\s+ou\s+mastercard|cr
edit\s+card|bank\s+card)\W'i
meta      __TRUF_3      __FRAUD_QXX
body      __TRUF_4      m'(?:\+|00)[ /-]?225[ /-]'
body      __TRUF_5a     m'\W(?:agence|agency)\W'i
body      __TRUF_5b     m'\W(?:voyage|travel)\W'i
body      __TRUF_6      m'\W(?:avion|air)\W'i
body      __TRUF_7      m'\Wdistance\W'i
body      __TRUF_8      m'\W(?:chambres|rooms)\W'i

meta      TRUF  3*__TRUF_1 + 2*__TRUF_2 + 2*__TRUF_3 + 5*__TRUF_4 + (__TRUF_5a && __TRUF_5b) + __TRUF_6 + __TRUF_7 + _
_TRUF_8 > 6
describe  TRUF          I Toto' francofoni (experimental)
score     TRUF          0.001

meta      TRUF_POS      (3*__TRUF_1 + 2*__TRUF_2 + 2*__TRUF_3 + 5*__TRUF_4 + (__TRUF_5a && __TRUF_5b) + __TRUF_6 + __T
RUF_7 + __TRUF_8 > 3) && !TRUF
describe  TRUF_POS      Possibili Toto' francofoni (experimental)
score     TRUF_POS      0.001


They are regarding advance frauds sent by someone from Ivory Coast. I'm occasionally receiving them with a low score and I would like to best SA behaviour about them, but I don't have an ham corpus against which test them.

Some of them are in french, some in frenchish and few in frenchtalian. I'm concerned mostly on french and frenchish versions: my customer can easily "detect" frenchtalian... So, I would like to have them checked either against a good english and french ham corpus.

FRAUD_QXX if from the 20_advance_fee.cf, which should came with SA. TRUF and TRUF_POS are, of course, the two possible results (fraud in Italian is said "truffa", thereby "TRUF"). The first should trigger only on frauds, the latter may occasionally trigger on ham.

Thank you!

-----------------------------------
Giampaolo Tomassoni - IT Consultant
Piazza VIII Aprile 1948, 4
I-53044 Chiusi (SI) - Italy
Ph: +39-0578-21100

MAI inviare una e-mail a:
NEVER send an e-mail to:
 rainbowl@tomassoni.eu