[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Sukhotin's Algorithm




Mark Perakh wrote:
> 
> Why would anybody try Sukhotin's algoritm on Hebrew which has no letters
> for vowels at all?  I wonder how it can work on any language given the
> fact that the Sukhotin'sd premise seems to be obviously wrong.

It's a statistical thing, I would guess. A weak signal coming out 
through high noise.

One thing I vaguely recollect from the earlier days is that with each
identified vowel there is also a confidence value. 
And I equally vaguely remember that the confidence value for the first
vowel in Voynichese was already lower than the second or third in 
English.

Cheers, Rene