Diable: Efficient Dialogue State Tracking as Operations on Tables

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lesci, Pietro, Fujinuma, Yoshinari, Hardalov, Momchil, Shang, Chao, Benajiba, Yassine, Marquez, Lluis
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909351767179264
author Lesci, Pietro
Fujinuma, Yoshinari
Hardalov, Momchil
Shang, Chao
Benajiba, Yassine
Marquez, Lluis
author_facet Lesci, Pietro
Fujinuma, Yoshinari
Hardalov, Momchil
Shang, Chao
Benajiba, Yassine
Marquez, Lluis
contents Sequence-to-sequence state-of-the-art systems for dialogue state tracking (DST) use the full dialogue history as input, represent the current state as a list with all the slots, and generate the entire state from scratch at each dialogue turn. This approach is inefficient, especially when the number of slots is large and the conversation is long. We propose Diable, a new task formalisation that simplifies the design and implementation of efficient DST systems and allows one to easily plug and play large language models. We represent the dialogue state as a table and formalise DST as a table manipulation task. At each turn, the system updates the previous state by generating table operations based on the dialogue context. Extensive experimentation on the MultiWoz datasets demonstrates that Diable (i) outperforms strong efficient DST baselines, (ii) is 2.4x more time efficient than current state-of-the-art methods while retaining competitive Joint Goal Accuracy, and (iii) is robust to noisy data annotations due to the table operations approach.
format Preprint
id arxiv_https___arxiv_org_abs_2305_17020
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Diable: Efficient Dialogue State Tracking as Operations on Tables
Lesci, Pietro
Fujinuma, Yoshinari
Hardalov, Momchil
Shang, Chao
Benajiba, Yassine
Marquez, Lluis
Computation and Language
Machine Learning
Sequence-to-sequence state-of-the-art systems for dialogue state tracking (DST) use the full dialogue history as input, represent the current state as a list with all the slots, and generate the entire state from scratch at each dialogue turn. This approach is inefficient, especially when the number of slots is large and the conversation is long. We propose Diable, a new task formalisation that simplifies the design and implementation of efficient DST systems and allows one to easily plug and play large language models. We represent the dialogue state as a table and formalise DST as a table manipulation task. At each turn, the system updates the previous state by generating table operations based on the dialogue context. Extensive experimentation on the MultiWoz datasets demonstrates that Diable (i) outperforms strong efficient DST baselines, (ii) is 2.4x more time efficient than current state-of-the-art methods while retaining competitive Joint Goal Accuracy, and (iii) is robust to noisy data annotations due to the table operations approach.
title Diable: Efficient Dialogue State Tracking as Operations on Tables
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2305.17020