6-820msg1.txt

运用贝叶斯网编的一个识别垃圾邮件的程序

Subject: synthetic compound

on sat , 3 jun 1995 , the linguist list write : > deat : sat , 3 jun 1995 1 : 37 : 59 - 320 > from : the linguist list < linguist @ tam2000 . tamu . edu > > to : hmander @ indiana . edu > subject : forward mail > > > from stampe @ uhunix . uhcc . hawaius . edu frus jun 2 14 : 2 : 50 1995 > return - path : stampe @ uhunix . uhcc . hawaius . edu > receive : from relay1 . hawaius . edu ( relay1 . hawaius . edu [ 128 . 171 . 41 . 53 ] ) by tam20 0 . tamu . edu ( 8 . 6 . 12 / 8 . 6 . 12 ) with smtp id oaa16738 for < linguist @ tam2000 . tamu . ed u > ; frus , 2 jun 1995 14 : 2 : 47 - 320 > receive : from uhunix3 . uhcc . hawaius . edu ( [ 128 . 171 . 44 . 52 ] ) by relay1 . hawaius . edu with smtp id < 11438 ( 5 ) > ; frus , 2 jun 1995 3 : 0 : 10 - 1000 > receive : by uhunix3 . uhcc . hawaius . edu id < 148528 > ; frus , 2 jun 1995 09 : 0 : 29 - 1 0 > from : david stampe < stampe @ uhunix . uhcc . hawaius . edu > > to : linguist @ tam2000 . tamu . edu > in - reply-to : < 199506021008 . faa11257 @ tam2000 . tamu . edu > ( message from the lingu ist list on frus , 2 jun 1995 0 : 08 : 49 - 1000 ) > subject : re : 6 . 758 , sum : synthetic compound > message - id : < 95jun2 . 090029hst . 148528 @ uhunix3 . uhcc . hawaius . edu > > deat : frus , 2 jun 1995 08 : 59 : 51 - 1000 > status : ro > content - length : 1036 > > heather marie anderson ( hmander @ indiana . edu ) compare agentive > compound to choose between ( 1 ) the theory " that the surface form of > synthetic [ compound ] will differ as much as possible in order and > affixation from the correspond vp " , and ( 2 ) the theory that they > simply follow vp order . > > in the language she cite , support for the theory ( 1 ) come mostly > from language that have relatively recently change the order of verb > and object , but not the order of compound , e . g . all the germanic > language list ( v o < o v , compound ov ) and finnish ( same ) . there > be similar example ( with mirror image order change ) in the munda > language of indium , which have word order change v o > o v , but retain > their old compound order vo , and in fact also retain it even in finite > verb and object which be incorporate ( compound ) . > > are there example of opposite order of verb phrase and compound > that could not be explain as due to the lag of morphological change > behind syntactic change ? > > david stampe > univ . of hawaus ` i > >