TU Darmstadt / ULB / TUbiblio

A Mention-Ranking Model for Abstract Anaphora Resolution

Marasovic, Ana and Born, Leo and Opitz, Juri and Frank, Anette :
A Mention-Ranking Model for Abstract Anaphora Resolution.
[Online-Edition: http://www.aclweb.org/anthology/D/D17/D17-1021.pdf]
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP)
[Conference or Workshop Item] , (2017)

Official URL: http://www.aclweb.org/anthology/D/D17/D17-1021.pdf

Abstract

Resolving abstract anaphora is an important, but difficult task for text understanding. Yet, with recent advances in representation learning this task becomes a more tangible aim. A central property of abstract anaphora is that it establishes a relation between the anaphor embedded in the anaphoric sentence and its (typically non-nominal) antecedent. We propose a mention-ranking model that learns how abstract anaphors relate to their antecedents with an LSTM-Siamese Net. We overcome the lack of training data by generating artificial anaphoric sentence--antecedent pairs. Our model outperforms state-of-the-art results on shell noun resolution. We also report first benchmark results on an abstract anaphora subset of the ARRAU corpus. This corpus presents a greater challenge due to a mixture of nominal and pronominal anaphors and a greater range of confounders. We found model variants that outperform the baselines for nominal anaphors, without training on individual anaphor data, but still lag behind for pronominal anaphors. Our model selects syntactically plausible candidates and -- if disregarding syntax -- discriminates candidates using deeper features.

Item Type: Conference or Workshop Item
Erschienen: 2017
Creators: Marasovic, Ana and Born, Leo and Opitz, Juri and Frank, Anette
Title: A Mention-Ranking Model for Abstract Anaphora Resolution
Language: German
Abstract:

Resolving abstract anaphora is an important, but difficult task for text understanding. Yet, with recent advances in representation learning this task becomes a more tangible aim. A central property of abstract anaphora is that it establishes a relation between the anaphor embedded in the anaphoric sentence and its (typically non-nominal) antecedent. We propose a mention-ranking model that learns how abstract anaphors relate to their antecedents with an LSTM-Siamese Net. We overcome the lack of training data by generating artificial anaphoric sentence--antecedent pairs. Our model outperforms state-of-the-art results on shell noun resolution. We also report first benchmark results on an abstract anaphora subset of the ARRAU corpus. This corpus presents a greater challenge due to a mixture of nominal and pronominal anaphors and a greater range of confounders. We found model variants that outperform the baselines for nominal anaphors, without training on individual anaphor data, but still lag behind for pronominal anaphors. Our model selects syntactically plausible candidates and -- if disregarding syntax -- discriminates candidates using deeper features.

Title of Book: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Uncontrolled Keywords: AIPHES_area_a3
Divisions: DFG-Graduiertenkollegs
DFG-Graduiertenkollegs > Research Training Group 1994 Adaptive Preparation of Information from Heterogeneous Sources
Event Location: Copenhagen, Denmark
Date Deposited: 03 Jul 2017 21:49
Official URL: http://www.aclweb.org/anthology/D/D17/D17-1021.pdf
Identification Number: TUD-CS-2017-0149
Export:

Optionen (nur für Redakteure)

View Item View Item