Search the web
Sign In
New User? Sign Up
LingPipe
? Already a member? Sign in to Yahoo!

Yahoo! Groups Tips

Did you know...
Hear how Yahoo! Groups has changed the lives of others. Take me there.

Best of Y! Groups

   Check them out and nominate your group.
Having problems with message search? Fill out this form to ensure your group is one of the first to be migrated to the new message search system.

Messages

  Messages Help
Advanced
Messages 166 - 195 of 777   Oldest  |  < Older  |  Newer >  |  Newest
Messages: Simplify | Expand   (Group by Topic) Author Sort by Date ^
166
... Here are some other presentations available online. I'll add links and add the papers to the distro in the next release. In all cases, the LingPipe...
Bob Carpenter
colloquialdo...
Offline Send Email
Sep 1, 2005
5:26 pm
167
Hello, I'm replying to an older email from August 4th, 2005, because it contains some relevant context. ... e/lucene/analysis/StopFi ... Here here :)...
Otis Gospodnetiæ
otis_gospodn...
Offline Send Email
Sep 19, 2005
6:37 pm
168
... I'm working on that now. I have the code that I'll post as a demo with the next release. ... That's a good idea -- I can wrap a TokenStream to create a...
Bob Carpenter
colloquialdo...
Offline Send Email
Sep 19, 2005
6:50 pm
169
I've just discovered LingPipe; it has some great features! I have one question about usage, though: if I build a TokenizedLM, then compile it to a file and...
Brian Pinkerton
BriPink
Offline Send Email
Sep 26, 2005
8:36 pm
170
... Nope -- you're absolutely spot on correct in both the way the lm package does behave and how it should behave. Having just reviewed the code after reading...
Bob Carpenter
colloquialdo...
Offline Send Email
Sep 26, 2005
9:09 pm
171
LingPipe Version 2.1 Released New York, 3 October 2005 LingPipe 2.1.0 is out. This release is backward compatible with 2.0 (and in almost all cases, with 1.0...
Bob Carpenter
colloquialdo...
Offline Send Email
Oct 4, 2005
2:53 am
172
Greetings. I was trying to run experiments with a very big model file (56.8MB) generated by a 69.9MB iob file. The training was OK (I had to increase the...
Andreas Vlachos
RiderOnTheSt...
Offline Send Email
Oct 11, 2005
3:34 pm
173
... file=/homes/av308/data/FlybaseAbstracts_17_06_2005/tests/input/mcolosimo_goldstandard.tokens-noDOC ... ...
reckb
Offline Send Email
Oct 12, 2005
5:57 pm
174
... Hi Breck, Thanks for getting back to me. Indeed, I should have provided more details. So here we go: I am using the NE detection model, using your command...
Andreas Vlachos
RiderOnTheSt...
Offline Send Email
Oct 13, 2005
5:43 pm
175
Hello everybody, My name is Philip and i am a student at the School for applied science in Hagenberg Austria. My problem is that i should write a...
spitzph
Offline Send Email
Oct 17, 2005
9:14 am
176
... This could be due to a lot of things. If you'd like me to try to debug it for you, could you put the BIO training file and the test file up somewhere on a...
carp@...
colloquialdo...
Offline Send Email
Oct 17, 2005
10:02 pm
177
... Hi Bob, I can put it on my webpage, but unless you prefer to do it this way, let's see if it can be solved without doings so. ... Thanks for the...
Andreas Vlachos
RiderOnTheSt...
Offline Send Email
Oct 18, 2005
11:28 pm
178
... Longer's more helpful for debugging. Thanks. ... It's a problem when the decoder can't find any viable paths. I'm going to rewrite the decoder so it can...
carp@...
colloquialdo...
Offline Send Email
Oct 19, 2005
7:41 pm
179
Found it. The problem isn't the size of the model per se, but rather the high-skew (see description below) of a transition in the model interacting with...
carp@...
colloquialdo...
Offline Send Email
Oct 19, 2005
8:45 pm
180
... Yes, that's it probably! I think the way I create my training data results in skewed distributions, apart from large models. Thanks for sorting this out, ...
Andreas Vlachos
RiderOnTheSt...
Offline Send Email
Oct 21, 2005
11:38 am
181
Final report on null returns from named-entity decoding. It does turn out to be the case that all possible continuations are being pruned. I fixed that, but...
carp@...
colloquialdo...
Offline Send Email
Oct 24, 2005
8:22 pm
182
Hi all, I've found recently about LinpPipe. I implemented a naive Bayes classifier that follows the approach from Mitchell (Machine Learning) and try it on the...
Eduard Barbu
eduard_barbu
Offline Send Email
Nov 9, 2005
3:55 pm
183
... Thanks for writing to the list -- this is a great case study. The bottom line answer to question two is that LingPipe's classifier is better because of...
carp@...
colloquialdo...
Offline Send Email
Nov 9, 2005
7:40 pm
184
Hi all, First I want to thank Bob for his detailed and competent answer. It really has been useful. However I want to make some clarifications: 1. I do not...
eduard barbu
eduard_barbu
Offline Send Email
Nov 11, 2005
9:05 pm
185
... That's an unusual approach, but mathematically sound. (In fact, more sound than many attempts to deal with unknown tokens.) ... This can skew the result...
carp@...
colloquialdo...
Offline Send Email
Nov 14, 2005
7:01 pm
186
LingPipe 2.1.1 is out (see the Download Page). This release patches all bugs of which we are aware. Major bug fixes include: (a) some MEDLINE handler...
carp@...
colloquialdo...
Offline Send Email
Nov 18, 2005
11:04 pm
187
... Most of the tools out there, like LingPipe, can be trained on arbitrary data. The reason we don't distribute models is that neither the CoNLL licenses nor...
Bob Carpenter
colloquialdo...
Offline Send Email
Nov 20, 2005
7:43 pm
188
Sorry -- I somehow fat fingered a one-year old draft out of my mailer. I even confused myself with that one. - Bob Carpenter...
Bob Carpenter
colloquialdo...
Offline Send Email
Nov 20, 2005
7:52 pm
189
Hello, My name is Dina Conti and I am student in computer science. We have got as homework to implement a naïve base classifier for text classification. We...
dinaconti1985
Offline Send Email
Nov 24, 2005
6:40 pm
190
Hello everyone! First of all, thank you for making Ling-Pipe available! I have one question related to the coreference module: how difficult would be to...
Nicolae Mihalache
xpromache
Offline Send Email
Dec 13, 2005
8:39 am
191
... Your welcome. It is a pleasure to see it being used. ... That depends on your skills/resources so it is hard to really say. Writing a tutorial for this...
reckb
Offline Send Email
Dec 13, 2005
4:07 pm
192
... That's my fault. I somehow distributed a version without the null constructor. I've attached the latest version as a tarball, because it's only 37KK. You...
carp@...
colloquialdo...
Offline Send Email
Dec 14, 2005
8:20 pm
193
Take 2. This time with the attachment. Sorry about that. - Bob...
carp@...
colloquialdo...
Offline Send Email
Dec 14, 2005
8:26 pm
194
Hello, ... one ... compile it ... call ... I just poked around Brian's code and see what he was talking about. Judging from the Javadocs for 2.1.1, it's still...
Otis Gospodnetiæ
otis_gospodn...
Offline Send Email
Dec 16, 2005
7:10 pm
195
... Just to remind everyone of why this came up without repeating two long messages, the advantage of this generalization is that it allows the background ...
carp@...
colloquialdo...
Offline Send Email
Dec 16, 2005
8:16 pm
Messages 166 - 195 of 777   Oldest  |  < Older  |  Newer >  |  Newest
Advanced
Add to My Yahoo!      XML What's This?

Copyright © 2009 Yahoo! Inc. All rights reserved.
Privacy Policy - Terms of Service - Guidelines - Help