A pipeline for extracting abstract dependency templates for data-to-text natural language generation
Mille, SimonORCID: 0000-0002-8852-2764, Ricci, Josep, Shvets, AlexanderORCID: 0000-0002-8370-2109 and Belz, AnyaORCID: 0000-0002-0552-8096
(2023)
A pipeline for extracting abstract dependency templates for data-to-text natural language generation.
In: Seventh International Conference on Dependency Linguistics (Depling, GURT/SyntaxFest 2023), 9-12 Mar 2023, Washington, D.C..
We present work in progress that aims to address the coverage issue faced by rule based text generators. We propose a pipeline for extracting abstract dependency template(predicate-argument structures)from WikipediatexttobeusedasinputforgeneratingtextfromstructureddatawiththeFORGe system. The pipeline comprises three main components:(i) candidate sentence retrieval, (ii)clause extraction, ranking and selection, and (iii) conversion to predicate-argument form. Wepresentanapproachandpreliminaryevaluationfortherankingandselectionmodule.
Proceedings of the Seventh International Conference on Dependency Linguistics (Depling, GURT/SyntaxFest 2023).
.
Association for Computational Linguistics (ACL).
ADAPT/DCU by the MSCA-PF-EF2021 grant awarded for the action 101062572, UPF by the EC-funded research and innovation programme Horizon Europe under the grant agreement number 101070278 and by the Erasmus + programme
ID Code:
28661
Deposited On:
04 Jul 2023 10:38 by
Anya Belz
. Last Modified 04 Jul 2023 10:38