6-825msg1.txt

运用贝叶斯网编的一个识别垃圾邮件的程序

Subject: re : 6 . 797 , comparative method : n - ary comparison

content - length : 1653 i , too , have note that there seem to be two notion of comparison here , but unlike scott delancey i do not assume that there we should distinguish between comparison for the sake of build a reconstruction and simply for the sake of determine possible relationship . to my mind , the number that jacque guy post demonstrate , not the weakness of n-ary comparison , but its strength : if we be look at a group of language of which we be uncertain of relationship , and the number of potential n-way cognate be as low as random chance would dictate , then the likelihood be against their be closely-enough relate to pursue reconstruction . i think this answer , by the bye , david power ' perhap rhetorical question regard the assumption under which janhunen 's claim could be consider a fallacy . i ' m not quite sure how to address power ' conclusion , however . the method of comparison most of us accept have build into them a check mechanism , such that acceptance of some set of match as " true " ( in power ' term ) constrain the set of further match we can accept : what he consider to be " false " match may not , under these constraint , be treat as match at all . under this methodology , n-way comparison * do * increase the ratio of signal to noise in the datum . i do not have access to janhunen 's original statement . if alexis manaster - ramer have summarize them accurately , i have to conclude either that janhunen be unfamiliar with the actual workings of the comparative method , or that the conclusion summarize by amr be disingenuous in the extreme - - and be indeed a fallacy , either way . rich alderson