<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2900.3268" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT size=2>If anybody's started on HW3 and has comments / answers to
these questions, it would be appreciated. I'm going to try and move ahead
later this evening.</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Thanks!</FONT></DIV>
<DIV><FONT size=2>Ben</FONT></DIV>
<BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
<DIV style="FONT: 10pt arial">----- Original Message ----- </DIV>
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black"><B>From:</B>
<A title=bcipolli@cogsci.ucsd.edu href="mailto:bcipolli@cogsci.ucsd.edu">Ben
Cipollini</A> </DIV>
<DIV style="FONT: 10pt arial"><B>To:</B> <A title=ligncse256@ling.ucsd.edu
href="mailto:ligncse256@ling.ucsd.edu">ligncse256@ling.ucsd.edu</A> </DIV>
<DIV style="FONT: 10pt arial"><B>Sent:</B> Saturday, March 01, 2008 9:00
AM</DIV>
<DIV style="FONT: 10pt arial"><B>Subject:</B> [Ligncse256] More HW3
questions</DIV>
<DIV><BR></DIV>
<DIV><FONT size=2>Hey all,</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Some more questions from HW3:</FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>1. Does anybody get performance on the BaselineParser like
in the HW (F=30%)? I'm getting F1=16.43% on validation and 16.16%
on test trees.</FONT></DIV><FONT size=2>
<DIV><BR>2. Is there any purpose to constructing a Grammar object? After
being constructed, it doesn't seem used by the BaselineParser class at
all. Commenting it out seems to have no effect.</DIV>
<DIV><BR>3. Did we cover "Agenda-Based" parsers in class? If so, did we
call them agenda-based parsers? If not, what did we call them?</DIV>
<DIV><BR>4. You might think the BaselineParser would do better if it used
HW2's HMM Tagger instead of the "unigram" tagger (which I think is really just
another version of HW2's most frequent scorer + greedy decoder trellis
combo). </FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Tagging performance (overall/unknown):</FONT></DIV>
<DIV><FONT size=2> HW2 HMM perf:
95.8%/65%<BR> HW unigram perf: 92%/40% </FONT></DIV>
<DIV><FONT size=2></FONT> </DIV>
<DIV><FONT size=2>Instead I got a drop in performance from 15% to under
5%. Was anybody else able to hook up their HW2 tagger code to HW3?
If so, what stats do you see?<BR></FONT></DIV>
<P>
<HR>
<P></P>_______________________________________________<BR>Ligncse256 mailing
list<BR>Ligncse256@ling.ucsd.edu<BR>http://pidgin.ucsd.edu/mailman/listinfo/ligncse256<BR></BLOCKQUOTE></BODY></HTML>