Identifying attack and support argumentative relations using deep learning
File(s)
Accepted version
Author(s)
Cocarascu, Oana
Toni, Francesca
Type
Conference Paper
Abstract
We propose a deep learning architecture to
capture argumentative relations of
attack
and
support
from one piece of text to an-
other, of the kind that naturally occur in
a debate. The architecture uses two (uni-
directional or bidirectional) Long Short-
Term Memory networks and (trained or
non-trained) word embeddings, and al-
lows to considerably improve upon exist-
ing techniques that use syntactic features
and supervised classifiers for the same
form of (relation-based) argument mining.
capture argumentative relations of
attack
and
support
from one piece of text to an-
other, of the kind that naturally occur in
a debate. The architecture uses two (uni-
directional or bidirectional) Long Short-
Term Memory networks and (trained or
non-trained) word embeddings, and al-
lows to considerably improve upon exist-
ing techniques that use syntactic features
and supervised classifiers for the same
form of (relation-based) argument mining.
Date Issued
2017-09-09
Date Acceptance
2017-07-01
Citation
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, 2017, pp.1374-1379
Publisher
Association for Computational Linguistics
Start Page
1374
End Page
1379
Journal / Book Title
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017
Copyright Statement
© 2017 Association for Computational Linguistics
Identifier
https://aclanthology.info/papers/D17-1144/d17-1144
Source
2017 Conference on Empirical Methods in Natural Language Processing
Notes
timestamp: Tue, 30 Jan 2018 13:42:04 +0100 biburl: http://dblp.org/rec/bib/conf/emnlp/CocarascuT17 bibsource: dblp computer science bibliography, http://dblp.org
Start Date
2017-09-09
Finish Date
2017-09-11
Coverage Spatial
Copenhagen, Denmark, September 9-11, 2017