The linguistics of authorship recognition

[a_b_desert_king]

No: HPFGUIDX 120609


> snip

> 4. There are some recognized ways to do authorship recognition 
> studies.  In my own graduate work I use my faculty advisor's 
> Vocabulary Management Profiles program to check for word usage 
> (hapax legmena, type/token ratio, etc).  The website is an 
academic 
> one (that is, not for profit) and is:
> www.missouri.edu/~youmansc/vmp/
> You might consider getting some mooseming text, then finding a bit 
> of comparable wording from JKR, say, transcript from a chat or 
> interview, then comparing relevant statistics.  

snip

> 
> Merry Christmas! Happy Holidays!
> Brian Brinkman

Hi Brian,

Since I have nada to do this Christmas Day, I decided to give your 
site a try.  I'm not sure I'm doing it right tho...

I found all of mooseming's posts that I could (very interesting read 
for anyone who would like to see them) and took some of her text 
from her website Rumours and News sections.  Now, this was what I 
got:

VMP2.2: "mooseming files.txt" Interval: 55 
TotalTypes=857  TotalTokens=2413  Types/Tokens=0.3552
AvgR = the ratio of Types / Tokens over the moving interval.
           

Mean avgR = 0.35530
Standard Deviation = 0.06603
Fractal Dimension = 1.480432  

If I did that right, that says to me that there is a pretty close 
match....  But I'm not sure I did it right....

Anyway, if you want to try it yourself I have the files here - just 
e-mail me and I'll happily send it to you to try.  I'm really 
curious now....

Heather - Merry Christmas to you too!