RE: VMs: VMS Word context similarities

["Marke Fincher" <MarkeFincher@xxxxxxxxxxxxxxxxxxxxxxx>]

Randomising the word order of an input file 
changes the adjacency relationships of words
in two different ways.   It breaks up the strong 
meaningful relationships, but it also creates
a weak background context match between all 
words.

I tried this with the bible sample, looking at 
the grouping of the word 'was'.

In the unscrambled file it was strongly grouped 
as: 
(were,be,was,is,are)  ... which makes sense.

In the scrambled file it was weakly bundled into
a large group: 

(was,in,yahweh,from,do,shall,god,you,which,to,they
is,his,all,that,a,for,be,into,your,my,them,him,said
its,has,who,us,as,this,he,her,i,when,by,one,up,will,
me,come,house,on,have,were,their,out,we,with,are,
son,king,it,men,go,but,not,came,people,before,because,
at,day,so,or,if,man,let,no,went,saying,had,israel,
there,against,also...)  !!

Something similar happens with the VMS.  When the
threshold is set high, randomisation reduces the
number of relationships found,  when the threshold
it set low, it increases the number.
 
Marke

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