Area 51 » Technology

definition commitment
116
followers

Natural Language Processing and Computational Linguistics (NLP/CL) [closed]

Proposed Q&A site for NLP/CL students and faculty in the academic research community as well as industry professionals
flag

Announcements

1  
Linguistic question belong on lingistics.stackexchange.com. Our cs.stackexchange.com site was created for these type of algorithm Q's. – Robert Cartaino♦ Apr 11 at 15:20
@Robert Cartaino You wrote "lingistics" :)... About the questions, please read my answer here – Alenanno May 15 at 9:21

closed as duplicate of Computer Science by Robert Cartaino♦ Apr 11 at 15:21

This proposal would tend to drain audience from an existing Stack Exchange site.

This proposal is now in the Commitment phase — example questions are locked!

45 Example Questions (8 closed)

active oldest votes
up vote 23 down vote
WordNet has really poor coverage for my dataset. Are there any unsupervised methods for extracting the hyponym/hypernym relationships?
added by percent20 Jun 3 '10 at 6:58
link|flag
up vote 22 down vote
Is there a dependency parser that can be used for very large corpora (1Tbyte)?
added by percent20 Jun 3 '10 at 6:57
link|flag
up vote 22 down vote
What are some of the methods for stemming?
added by percent20 Jun 3 '10 at 6:59
link|flag
up vote 21 down vote
Where can I find a Dutch corpus?
added by Michel de Ruiter Jun 15 '10 at 20:37
link|flag
up vote 21 down vote
What algorithms exist to split words combinations (very common in Dutch/German) into separate words?
added by Michel de Ruiter Jun 23 '10 at 8:23
link|flag
up vote 14 down vote
How do I tell what language is a plain-text file written in?
added by warren Jul 7 '10 at 21:31
link|flag
up vote 13 down vote
Are there any algorithms for an automatic construction of a dictionary?
added by Yarm Aug 5 '10 at 15:59
link|flag
up vote 10 down vote
Is there an NLP parser out there which accepts input with some words already tagged? As in "Bob/NP walked the dog." where Bob is a known proper noun.
added by Heath Hunnicutt Jul 13 '10 at 1:34
link|flag
up vote 5 down vote
Do the state-of-the-art machine translation algorithms parse the sentence for preprocessing?
added by Yarm Aug 5 '10 at 15:57
link|flag
up vote 4 down vote
What are the best programming languages for exploring NL parsing and language analysis (and why)?
added by Bruce Alderson Jun 22 '10 at 16:28
link|flag
up vote 4 down vote
Are there any well known algorithms to detect the presence of names?
added by warren Jul 7 '10 at 21:17
link|flag
up vote 3 down vote
How do you test for statistically significant improvements in machine translation performance?
added by dmcer Jun 23 '10 at 7:21
link|flag
up vote 3 down vote
Where can I find POS Taggers for Portuguese text?
added by Hbas Jul 21 '10 at 0:35
link|flag
up vote 3 down vote
What's the computational complexity of the Universal Recognition Problem (URP) for syntactic theories (HPSG, LFG…) ?
added by anno Sep 22 '10 at 15:33
link|flag
up vote 2 down vote
Can you write a capable NL parser using only regex?
added by Bruce Alderson Jun 22 '10 at 16:27
link|flag
I think it's a great question. Although the answer is No, interesting is Why not? – Michel de Ruiter Jun 23 '10 at 8:17
I think the answer could be yes ... depends exactly how capable g -- regex implementations in common programming languages aren't really limited to regular languages! What's difficult is the feature structures needed for semantic theories – Felix Dombek Dec 23 '10 at 23:59
up vote 2 down vote
Is their any principled way of filtering the unreasonable 5-grams in the Google n-gram corpus?
added by dmcer, edited by WikiSpeedia hang-around Mar 6 '11 at 10:03
link|flag
(n-gram: en.wikipedia.org/wiki/N-gram) – WikiSpeedia hang-around Mar 6 '11 at 10:03
up vote 2 down vote
What algorithms for negating sentences exist? For example, "Bob is good" -> "Bob is not good" rather than "Not Bob is good".
added by warren Jul 7 '10 at 21:21
link|flag
up vote 2 down vote
How to evaluate a machine translation system?
added by trampj, edited by WikiSpeedia hang-around Mar 19 '11 at 19:03
link|flag
up vote 2 down vote
What are some simple algorithms for automatic question generation from a given text?
added by Peter Apr 4 '11 at 23:25
link|flag
up vote 1 down vote
What sort of careers can I look forward to when pursing CL? What kind of projects could I work on? What companies do work in this field? [closed]
added by Charlie Jun 16 '10 at 22:59
link|flag

closed as off topic by dmcer, Dervin Thunk, warren, James Tauber, morfu Aug 22 '10 at 11:54

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 1 down vote
How do I do unsupervised grammar induction?
added by forefinger Jul 7 '10 at 6:58
link|flag
this should be specified further – Dervin Thunk Jul 29 '10 at 15:41
I agree -- that's too general to be answered – rmalouf Mar 18 '11 at 0:01
up vote 1 down vote
Corpus/data set of English words with syllabic stress information?
added by warren Jul 7 '10 at 21:18
link|flag
up vote 1 down vote
Are there neurological findings that support the reality of Universal Grammar? [closed]
added by Yarm Aug 5 '10 at 16:34
link|flag

closed as off topic by James Tauber, user13466, morfu, Daniel Trebbien, Edgar Nov 22 '10 at 7:01

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 1 down vote
Any good papers on avoiding text classifiers using embarrassing accidental characteristics of a training set (the "dead Jesuits" problem)?
added by DavidDLewis Oct 20 '10 at 18:44
link|flag
What is the "dead Jesuits" problem? – bjoernz Nov 20 '10 at 13:04
The MUC-3 evaluations involved doing text classification and information extraction on texts about terrorist incidents in Central America. There was a great deal of redundancy in the reports, including many mentions of a massacre of Jesuit priests. A number of participants systems' learned that the word "Jesuit" was a good predictor for the MURDER category. – DavidDLewis Nov 24 '10 at 17:52
up vote 1 down vote
Other than The Gutenburg Project, what major sources of "correct" natural language are available to learn from?
added by warren Jan 22 '11 at 2:11
link|flag
up vote 1 down vote
What is a good algorithm for identifying idioms in a sentence, even when declined/negated/used with artistic license?
added by Carmi Mar 5 '11 at 17:57
link|flag
What do you mean by "sorting"? Maybe better wording would be: "What is a good technique for identifying the use of idioms in a sentence, even when etcetc..." – rmalouf Mar 18 '11 at 0:12
Edited to reflect rmalouf's suggested change. – Carmi Mar 29 '11 at 11:55
up vote 0 down vote
When should one use "which" as opposed to "that"? [closed]
added by Meidan Alon Jun 22 '10 at 16:01
link|flag

closed as off topic by Bruce Alderson, dmcer, Michel de Ruiter, Heath Hunnicutt, Mark Davidson Jul 16 '10 at 11:26

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 0 down vote
Why did Microsoft/Bing buy the NLP based search engine Powerset? [closed]
added by dmcer Jun 23 '10 at 7:22
link|flag

closed as off topic by warren, Heath Hunnicutt, Hbas, peeles, Yarm Aug 5 '10 at 16:28

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 0 down vote
Where can I find a website that teaches English? [closed]
added by Yarm Aug 5 '10 at 16:26
link|flag

closed as off topic by James Tauber, warren, anno, DavidDLewis, kepler Nov 26 '10 at 13:57

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 0 down vote
Have there been any objective studies of the impact of licensing conditions on the use of natural language datasets by researchers? [closed]
added by DavidDLewis Oct 20 '10 at 18:36
link|flag

closed as off topic by Bala, Tangurena, BenV, nhaugo, Carmi Mar 5 '11 at 18:02

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote 0 down vote
How should one trade off length of segment vs. number of segments in sampling a speech data set (with many speakers) to estimate ASR accuracy?
added by DavidDLewis Oct 20 '10 at 18:41
link|flag
up vote 0 down vote
Why could as of now no complete grammar for English be compiled (say, 99.9 % coverage of well-formed sentences?)
added by Felix Dombek Dec 24 '10 at 0:32
link|flag
Perhaps this is more a linguistics question than an NLTK question? – hippietrail Mar 29 '11 at 9:33
up vote 0 down vote
What kind of parser is most likely used in our brains to parse natural language and why is this assumed?
added by Felix Dombek Dec 24 '10 at 0:36
link|flag
up vote 0 down vote
Does the nltk module of Python have any feature to identify nouns from a given English text?
added by Barun Feb 22 '11 at 16:03
link|flag
up vote 0 down vote
How can I implement Karttunen's priority union operation for finite state transducers using OpenFST?
added by rmalouf Mar 18 '11 at 0:05
link|flag
up vote 0 down vote
How does Kneser-Ney smoothing work?
added by rmalouf Mar 18 '11 at 0:09
link|flag
up vote 0 down vote
Are there any known formal definitions of information that might relate to natural languages as carriers of information?
added by Ricky Mango Mar 24 '11 at 10:10
link|flag
up vote 0 down vote
How much grammar knowledge do modern statistical translators like Google Translate use?
added by hippietrail Apr 26 '11 at 12:05
link|flag
up vote -1 down vote
What institutions/universities provide a good program in computational linguistics? Which are well known for their work in CL? [closed]
added by Charlie Jun 16 '10 at 23:01
link|flag

closed as off topic by dmcer, Dervin Thunk, warren, morfu, Edgar Nov 22 '10 at 7:01

This question does not relate to the topic of the proposal.
It was closed as part of an automated migration of off-topic to close votes on September 29, 2011.
up vote -1 down vote
What do people view as being the most compelling practical successes of semi-supervised learning in NLP?
added by DavidDLewis Oct 20 '10 at 18:37
link|flag
Sounds like a poll question, so not really considered "proper" anywhere on the SE network. – Carmi Mar 5 '11 at 17:59
1 2