Template-assisted Contrastive Learning of Task-oriented Dialogue Sentence Embeddings

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Oh, Minsik, Li, Jiwei, Wang, Guoyin
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910121306619904
author Oh, Minsik
Li, Jiwei
Wang, Guoyin
author_facet Oh, Minsik
Li, Jiwei
Wang, Guoyin
contents Learning high quality sentence embeddings from dialogues has drawn increasing attentions as it is essential to solve a variety of dialogue-oriented tasks with low annotation cost. Annotating and gathering utterance relationships in conversations are difficult, while token-level annotations, \eg, entities, slots and templates, are much easier to obtain. Other sentence embedding methods are usually sentence-level self-supervised frameworks and cannot utilize token-level extra knowledge. We introduce Template-aware Dialogue Sentence Embedding (TaDSE), a novel augmentation method that utilizes template information to learn utterance embeddings via self-supervised contrastive learning framework. We further enhance the effect with a synthetically augmented dataset that diversifies utterance-template association, in which slot-filling is a preliminary step. We evaluate TaDSE performance on five downstream benchmark dialogue datasets. The experiment results show that TaDSE achieves significant improvements over previous SOTA methods for dialogue. We further introduce a novel analytic instrument of semantic compression test, for which we discover a correlation with uniformity and alignment. Our code is available at https://github.com/minsik-ai/Template-Contrastive-Embedding
format Preprint
id arxiv_https___arxiv_org_abs_2305_14299
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Template-assisted Contrastive Learning of Task-oriented Dialogue Sentence Embeddings
Oh, Minsik
Li, Jiwei
Wang, Guoyin
Computation and Language
Artificial Intelligence
Learning high quality sentence embeddings from dialogues has drawn increasing attentions as it is essential to solve a variety of dialogue-oriented tasks with low annotation cost. Annotating and gathering utterance relationships in conversations are difficult, while token-level annotations, \eg, entities, slots and templates, are much easier to obtain. Other sentence embedding methods are usually sentence-level self-supervised frameworks and cannot utilize token-level extra knowledge. We introduce Template-aware Dialogue Sentence Embedding (TaDSE), a novel augmentation method that utilizes template information to learn utterance embeddings via self-supervised contrastive learning framework. We further enhance the effect with a synthetically augmented dataset that diversifies utterance-template association, in which slot-filling is a preliminary step. We evaluate TaDSE performance on five downstream benchmark dialogue datasets. The experiment results show that TaDSE achieves significant improvements over previous SOTA methods for dialogue. We further introduce a novel analytic instrument of semantic compression test, for which we discover a correlation with uniformity and alignment. Our code is available at https://github.com/minsik-ai/Template-Contrastive-Embedding
title Template-assisted Contrastive Learning of Task-oriented Dialogue Sentence Embeddings
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2305.14299