Learning Connective-based Word Representations for Implicit Discourse Relation Identification

Chloé Braud 1 Pascal Denis 2
2 MAGNET - Machine Learning in Information Networks
Inria Lille - Nord Europe, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189
Abstract : We introduce a simple semi-supervised approach to improve implicit discourse relation identification. This approach harnesses large amounts of automatically extracted discourse connectives along with their arguments to construct new distributional word representations. Specifically, we represent words in the space of discourse connectives as a way to directly encode their rhetorical function. Experiments on the Penn Discourse Treebank demonstrate the effectiveness of these task-tailored representations in predicting implicit discourse relations. Our results indeed show that, despite their simplicity, these connective-based representations outperform various off-the-shelf word embeddings, and achieve state-of-the-art performance on this problem.
Complete list of metadatas

Cited literature [40 references]  Display  Hide  Download

https://hal.inria.fr/hal-01397318
Contributor : Pascal Denis <>
Submitted on : Thursday, December 15, 2016 - 8:37:54 PM
Last modification on : Friday, March 22, 2019 - 4:46:03 PM
Long-term archiving on : Monday, March 20, 2017 - 10:42:24 PM

File

emnlp16.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01397318, version 1

Citation

Chloé Braud, Pascal Denis. Learning Connective-based Word Representations for Implicit Discourse Relation Identification. Empirical Methods on Natural Language Processing, Nov 2016, Austin, United States. ⟨hal-01397318⟩

Share

Metrics

Record views

377

Files downloads

260