Difference between revisions of "S. Patwardhan and E. Riloff. EMNLP 2009"
PastStudents (talk | contribs) |
PastStudents (talk | contribs) |
||
Line 14: | Line 14: | ||
such as event extraction, a larger field of view is | such as event extraction, a larger field of view is | ||
often needed to understand how facts tie together. | often needed to understand how facts tie together. | ||
− | This paper proposed a new model for event extraction. | + | This [[Category::paper]] proposed a new model for [[AddressesProblem::event extraction]]. |
To determine whether a noun phrase should be extracted as a filler for an event role | To determine whether a noun phrase should be extracted as a filler for an event role | ||
the new model computes the joint probability that NPi : | the new model computes the joint probability that NPi : | ||
Line 20: | Line 20: | ||
# is a legitimate filler for the event role. | # is a legitimate filler for the event role. | ||
To compute the probability of a sentence describing a relevant | To compute the probability of a sentence describing a relevant | ||
− | event, they use SVM, which is not a probabilistic classifier. The authors | + | event, they use [[UsesMethod::SVM]], which is not a probabilistic classifier. The authors |
used the margin as an indicator of confidence. It worked well for them. | used the margin as an indicator of confidence. It worked well for them. | ||
Named entities, lexico-syntactic pattern features, | Named entities, lexico-syntactic pattern features, | ||
Line 28: | Line 28: | ||
whether a noun phrase can be a legitimate | whether a noun phrase can be a legitimate | ||
filler for a specific type of event role based on | filler for a specific type of event role based on | ||
− | its local context, the authors used Naive Bayes classifier. | + | its local context, the authors used [[UsesMethod::Naive Bayes classifier]]. |
Revision as of 04:17, 30 November 2010
Citation
S. Patwardhan and E. Riloff. A unified model of phrasal and sentential evidence for information extraction. in EMNLP 2009
Online version
Summary
Previous IE systems make decision only based on immediate context around a phrase. The authors argue that for more complex tasks, such as event extraction, a larger field of view is often needed to understand how facts tie together. This paper proposed a new model for event extraction. To determine whether a noun phrase should be extracted as a filler for an event role the new model computes the joint probability that NPi :
- appears in an event sentence, and
- is a legitimate filler for the event role.
To compute the probability of a sentence describing a relevant event, they use SVM, which is not a probabilistic classifier. The authors used the margin as an indicator of confidence. It worked well for them. Named entities, lexico-syntactic pattern features, sentence length, bag of words, and verb tense are used as features.
To determine whether a noun phrase can be a legitimate filler for a specific type of event role based on its local context, the authors used Naive Bayes classifier.