Wang, Di and O'Keefe, Simon orcid.org/0000-0001-5957-2474 (2021) Memory state tracker : A memory network based dialogue state tracker. In: Rocha, Ana Paula, Steels, Luc and van den Herik, Jaap, (eds.) ICAART 2021 - Proceedings of the 13th International Conference on Agents and Artificial Intelligence. 13th International Conference on Agents and Artificial Intelligence, ICAART 2021, 04-06 Feb 2021 ICAART 2021 - Proceedings of the 13th International Conference on Agents and Artificial Intelligence . SciTePress , Virtual, Online , pp. 533-542.
Abstract
Dialogue State Tracking (DST) is a core component towards task oriented dialogue system. It fills manually-set slots at each turn of an utterance, which indicate the current topics or user requirement. In this work we propose a memory based state tracker that includes a memory encoder which encodes the dialogue history into a memory vector, and then connects to a pointer network which makes predictions. Our model reached a joint goal accuracy of 49.16% on MultiWOZ 2.0 data set (Budzianowski et al., 2018) and 47.27% on MultiWOZ 2.1 data set (Eric et al., 2019), outperforming the benchmark result.
Metadata
Authors/Creators: |
|
---|---|
Copyright, Publisher and Additional Information: | © 2021 SCITEPRESS |
Keywords: | Dialogue state tracker, Memory network |
Dates: |
|
Institution: | The University of York |
Academic Units: | The University of York > Faculty of Sciences (York) > Computer Science (York) |
Depositing User: | Pure (York) |
Date Deposited: | 19 May 2023 13:40 |
Last Modified: | 13 Jan 2024 00:21 |
Published Version: | https://doi.org/10.5220/0010385705330538 |
Status: | Published |
Publisher: | SciTePress |
Series Name: | ICAART 2021 - Proceedings of the 13th International Conference on Agents and Artificial Intelligence |
Refereed: | No |
Identification Number: | https://doi.org/10.5220/0010385705330538 |
Related URLs: |