[cs241] data (I think this is correct)

[Hoang]

I got the same results as Tim.
Also, for calculating P(path/VP1,VP0), should we multiply
[#(VP1,VP0/path) / #(VP1,VP0)] by P(path), i.e.:
 
P(path/VP1,VP0) = P(path) * [#(VP1,VP0/path) / # (VP1,VP0)] ?
 
Thanks,
 
Hoang.

-----Original Message-----
From: cs241-bounces at list.cs.brown.edu
[mailto:cs241-bounces at list.cs.brown.edu] On Behalf Of Tim St. Clair
Sent: Monday, October 29, 2007 1:52 PM
To: cs241 at list.cs.brown.edu
Subject: [cs241] data (I think this is correct)


Note: the number in the path column is the number of the path, when you
order them descending from most common:

0    .VP*.SBAR.S.VP    128041
1    .VP*.S.VP    125588
2    .VP*.VP    93718
3    .VP*.NP.SBAR.S.VP    35535 
4    .VP.S*.S.VP    32972
5    .VP.S*.SBAR.S.VP    25879
6    .VP*.PP.NP.SBAR.S.VP    21643
7    .VP*.PP.S.VP    19329
8    .VP.S*.NP.SBAR.S.VP    17293
9    .VP.SINV*.S.VP    11499
10    .VP*.NP.VP    10465 

(* proceeds the root)
 <http://spreadsheets.google.com/ccc?key=p3coYwZqOPPyNxYIj9z7ZWA&hl=en> 
http://spreadsheets.google.com/ccc?key=p3coYwZqOPPyNxYIj9z7ZWA
<http://spreadsheets.google.com/ccc?key=p3coYwZqOPPyNxYIj9z7ZWA&hl=en>
&hl=en

-- 
Tim St. Clair

(617) 460 - 6497 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.cs.brown.edu/pipermail/cs241/attachments/20071029/2d804d54/attachment.htm