MALA: Cross-Domain Dialogue Generation with Action Learning

TL;DR

This paper introduces MALA, a multi-stage adaptive latent action learning framework that improves cross-domain dialogue generation by learning semantic actions based on dialogue state transitions, enhancing task success and language quality.

Contribution

MALA proposes a novel multi-stage learning approach that models semantic latent actions through dialogue state effects, addressing language diversity issues in cross-domain dialogue systems.

Findings

Achieves better task completion rates across multiple domains.

Improves language quality in generated responses.

Outperforms baseline models on SMD and MultiWOZ datasets.

Abstract

Response generation for task-oriented dialogues involves two basic components: dialogue planning and surface realization. These two components, however, have a discrepancy in their objectives, i.e., task completion and language quality. To deal with such discrepancy, conditioned response generation has been introduced where the generation process is factorized into action decision and language generation via explicit action representations. To obtain action representations, recent studies learn latent actions in an unsupervised manner based on the utterance lexical similarity. Such an action learning approach is prone to diversities of language surfaces, which may impinge task completion and language quality. To address this issue, we propose multi-stage adaptive latent action learning (MALA) that learns semantic latent actions by distinguishing the effects of utterances on dialogue progress. We model the utterance effect using the transition of dialogue states caused by the utterance and develop a semantic similarity measurement that estimates whether utterances have similar effects. For learning semantic actions on domains without dialogue states, MsALA extends the semantic similarity measurement across domains progressively, i.e., from aligning shared actions to learning domain-specific actions. Experiments using multi-domain datasets, SMD and MultiWOZ, show that our proposed model achieves consistent improvements over the baselines models in terms of both task completion and language quality.