Dear all, I have succesfully installed the lingpipe software but I would like to create a dutch model voor LingPipe. What am I to do? How can I train the ...
For Dutch, you're in luck. There's a freely distributed corpus for newswire data and LingPipe is distributed with the filters to train it. Here's a response...
Bob Carpenter
carp@...
Jan 8, 2004 5:28 pm
12
Dear all, I am currently writing my master thesis about named entity recognition and its impact on cross-lingual information retrieval. I wanted to evaluate...
Here's a how-to file that I'll include with the next release of LingPipe. Please let me know if there are any further questions. Training Dutch and Spanish...
Bob Carpenter
carp@...
Jan 22, 2004 8:57 pm
14
Hi, I am interested in training the NE-recognizer for German and I have some questions I would like to ask: - what exactly is the role of a dictionary...
The dictionary in the AbstractNECommand is used heuristically. It works left-to-right by token, labeling the longest match from the dictionary at each point....
Bob Carpenter
carp@...
Jan 29, 2004 4:34 pm
16
Hi Bob, I'd like to find out to what extent can the actual implemented co-reference algorithm be applied to other languages!? Is there any language specific...
... Coref depends on a few things. The current implementation for corefeference within documents is pretty sparse. English pronouns are hard-coded in the...
Bob Carpenter
carp@...
Jan 29, 2004 6:27 pm
18
Your success with German coref depends as well on what your needs are. If you need pronouns then you are in for some work since the English coref is indeed ...
Hi, I have finally succeded in creating the German model for the NE using the ConLL 2003 data, but the annotation doesn't seem to work at all: I'm getting as...
Dear fellow LingPipe-ians, we have been starting to experiment with the new lingpipe 1.0.3 as part of our university research. We have encountered a problem ...
Hi everybody! We have followed the instructions in the getting_started file and Bob's email sent on Jan. 22nd to create models for Spanish and Dutch based on...
hi: i have just downloaded lingpipe and tried to run it, i got the following error message (see dump below), does anyone have any idea what may have caused it....
... The exception is thrown when the file is not valid UTF-8. ... This pretty much confirms it. The basic picture with character encodings is: character ->...
Bob Carpenter
carp@...
Feb 2, 2004 5:25 pm
24
Looks like a Windows XP / 1.4.2 JVM problem. See, for instance, the first hit on Google for: Unexpected Signal : EXCEPTION_FLT_STACK_CHECK ...
Bob Carpenter
carp@...
Feb 2, 2004 5:51 pm
25
I was able to recreate your bug, so it's definitely a problem with LingPipe and not with the way you're using it. Thanks for giving me enough info to recreate ...
Bob Carpenter
carp@...
Feb 2, 2004 11:01 pm
26
A new and improved version of LingPipe is available from: http://www.alias-i.com/lingpipe/download.html The bug in CoNLL training has been fixed (see below)....
Bob Carpenter
carp@...
Feb 3, 2004 1:20 am
27
The ConLL TagParser for the 2003 data format, as posted on the list on 30th of January, is buggy - SORRY. In the mean time I have sent Bob a newer version, ...
Dear Bob, as I wasn't able to run WithinDocCorefAnnotateCommand with the new ConLL '03 generated English model, I had a quick look at the code and found the ...
Thanks for the updated models. I used the one you sent me and didn't realize it differed from the one posted earlier. You're absolutely right in your...
Bob Carpenter
carp@...
Feb 3, 2004 5:25 pm
30
Hi everybody! I'm trying to use LingPipe for named entity annotation and I wonder if it's possible to give the program some a-priori constraints. For example,...
... Yes there is--turns out that such a capability is crucial to cover the bald spots in NE annotation output. LinpPipe has two mechanisms for NE annotation...
... Run-time User-Defined Dictionaries ============================================= Yes, you can define your own dictionaries to be applied at run time. Like...
Bob Carpenter
carp@...
Feb 13, 2004 6:48 pm
33
Hi, I am trying to run Lingpipe over the Reuter's corpus. However, I am getting the following error. Processing file=data\raw\reut2-021.sgm [Fatal Error]...
Hi, I am trying to run Lingpipe over the Reuter's corpus. However, I am getting the following error. Processing file=data\raw\reut2-021.sgm [Fatal Error]...
Hi, I downloaded 'The Count of Monte Cristo' from www.gutenberg.org. I inserted appropriate tags and used lingpipe process it. Here is the error i got. [Fatal...
Hi, If the email occurs as '<aa@...>', the system crashes. [Fatal Error] grexp10.txt:176:22: Element type "hart" must be followed by either attribute...
Am I wrong to conclude that LingPipe detects named entities and anaphora? I input the following sentence to the LingPipe demo. The LingPipe detects only words ...
... LingPipe does do limited coreference and named entity detection. But like all such technologies it makes mistakes. The named entity detection was not...
... Navendu, The input format has to be in valid xml. You need to get an xml validator--generally findable on the web for free and use that. To check your...
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.
