1. LING/C SC 581: Advanced Computational Linguistics
Lecture Notes
Feb 1st

2. Today's Topic Have partially graded Homework 2 …
Assume you have installed tregex as requested last time
Homework 5 out today

3. tregex Select the PTB directory:TREEBANK_3/parsed/mrg/wsj/
Deselect any unwanted files
The_Wonderful_World_of_Tregex.ppt by Galen Andrew
Help button
Browse:

4. tregex
Adjust Max displayed trees if needed:

5. tregex

6. tregex Help: tregex expression syntax is non-standard wrt bracketing
S < VP
S < NP

7. tregex
Help: tregex boolean syntax is also non-standard

8. tregex
Help

9. tregex x <, y, 1st child y; x <- y, last child y;
x $+ y, x immediate left sister of y

10. tregex
same node
Pattern:
<, $+ (/,/ $+ $+ /,/=comma))) <- =comma)
Key:
<, first child
$+ immediate left sister
<- last child

11. tregex
Help

12. tregex

13. tregex Different results from:
@SBAR < /^WH.*-([0-9]+)$/#1%index << < (/^-NONE-/ < /^\*T\*-([0- 9]+)$/#1%index))

14. tregex Reason for difference Example:
WHADVP also possible (not just WHNP)

15. Treebank Guide Parsing guide 1, prsguid1.pdf (318 pages):
prsguid2.pdf: addendum for the Switchboard corpus

16. Homework 5 Usual rules, due next Wednesday (midnight)
One PDF file to me

17. Free Relatives Taken from Caponigro, I. (proceedings of WECOL 2002).
Some background:

18. Free Relatives
Free vs. Headed relatives:

19. Free Relatives

20. Free Relatives
I asked what he asked

21. Free Relatives

22. /^SBAR-NOM/
/^SBAR-NOM/ 775

23. Homework 5
What words or phrases head free relatives? List as many as you can think of.
Using tregex, devise search strings to build a count/frequency table of the words/phrase that comprises the 775 reported examples
Submit your tregex search strings and your table
Type
Count
Freq
[Ww]hat
500*
65%* …
*sample numbers only…