The Road to Quality is Paved with Good Revisions: A Detailed Evaluation Methodology for Revision Policies in Incremental Sequence Labelling

Brielen Madureira, Patrick Kahardipraja, David Schlangen


Abstract
Incremental dialogue model components produce a sequence of output prefixes based on incoming input. Mistakes can occur due to local ambiguities or to wrong hypotheses, making the ability to revise past outputs a desirable property that can be governed by a policy. In this work, we formalise and characterise edits and revisions in incremental sequence labelling and propose metrics to evaluate revision policies. We then apply our methodology to profile the incremental behaviour of three Transformer-based encoders in various tasks, paving the road for better revision policies.
Anthology ID:
2023.sigdial-1.14
Volume:
Proceedings of the 24th Annual Meeting of the Special Interest Group on Discourse and Dialogue
Month:
September
Year:
2023
Address:
Prague, Czechia
Editors:
Svetlana Stoyanchev, Shafiq Joty, David Schlangen, Ondrej Dusek, Casey Kennington, Malihe Alikhani
Venue:
SIGDIAL
SIG:
SIGDIAL
Publisher:
Association for Computational Linguistics
Note:
Pages:
156–167
Language:
URL:
https://aclanthology.org/2023.sigdial-1.14
DOI:
10.18653/v1/2023.sigdial-1.14
Bibkey:
Cite (ACL):
Brielen Madureira, Patrick Kahardipraja, and David Schlangen. 2023. The Road to Quality is Paved with Good Revisions: A Detailed Evaluation Methodology for Revision Policies in Incremental Sequence Labelling. In Proceedings of the 24th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 156–167, Prague, Czechia. Association for Computational Linguistics.
Cite (Informal):
The Road to Quality is Paved with Good Revisions: A Detailed Evaluation Methodology for Revision Policies in Incremental Sequence Labelling (Madureira et al., SIGDIAL 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.sigdial-1.14.pdf