VMs: VMs best-fit 2-state PFSMs

["Ben Preece" <ben.preece@xxxxxxxxxx>]

I'd generated best-fit 2-state PFSMs for various VMS transcriptions, for both language A and B.  None of them show a close correlation between language A and language B.

FSG-A
[0,0] : C D F H P Q S T 4 5 8 
[0,-] : Z X Y 
[1,1] : A E G I K L N O R 2 6 
[1,-] : V 
[-,1] : M 7 

FSG-B
[0,0] : D E F G H K M N P R 2 6 8 
[0,-] : Z 
[-,0] : Y 
[1,0] : L 7 
[1,1] : A C I O S T 4 

EVA-A
[0,0] : a c g i l m n o r s y
[0,-] : j 
[-,0] : v 
[1,1] : d e f h k p q t 
[-,1] : x 

EVA-B
[0,0] : a g i m n o y 
[0,1] : l 
[0,-] : j 
[1,0] : x 
[1,1] : c d e f k p q r s t 
[1,-] : h 

I notice that these groups stay together:  { a g i m n o y },  { d e f k p q t }, and { c l r s }, and the last group switches allegiance between the other two groups.

This last encoding is one I used sometimes.  It maps EVA cph, cth, ckh, cfh onto P, T, K, and F; it maps ch and sh onto c and ç; and it maps ee, eee, ii, and iii onto é, ê, í, and î.  I should also have mapped eeee onto something but never did.

BHP-A
[0,0] : a g i î m n o r y 
[0,1] : é ê l s 
[0,-] : í j
[-,0] : v
[1,1] : c ç d e F f K k P p q T t 
[-,1] : x

BHP-B
[0,0] : a e é o p q 
[1,0] : f i í j 
[1,1] : c ç d F g î K k l m n P r s y T t 
[1,-] : x 

Again, at least this one group stays together: { c ç d F K k P T t }, which includes the symbols which map onto EVA ch, sh, cph, cth, ckh, and cfh.

At Rene's suggestion, I'll see what Curva looks like, too.


-Ben
______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list