Task-Oriented Dialog Systems that Consider Multiple Appropriate Responses under the Same Context

By Yichi Zhang, Zhijian Ou, Zhou Yu. 2019

198 citations recorded by the source corpus.

Conversations have an intrinsic one-to-many property, which means that multiple responses can be appropriate for the same dialog context. In task-oriented dialogs, this property leads to different valid dialog policies towards task completion. However, none of the existing task-oriented dialog generation approaches takes this property into account. We propose a Multi-Action Data Augmentation (MADA) framework to utilize the one-to-many property to generate diverse appropriate dialog responses. Specifically, we first use dialog states to summarize the dialog history, and then discover all possible mappings from every dialog state to its different valid system actions. During dialog system training, we enable the current dialog state to map to all valid system actions discovered in the previous process to create additional state-action pairs. By incorporating these additional pairs, the dialog policy learns a balanced action distribution, which further guides the dialog model to generate diverse responses. Experimental results show that the proposed framework consistently improves dialog policy diversity, and results in improved response diversity and appropriateness. Our model obtains state-of-the-art results on MultiWOZ.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

A Simple Language Model for Task-Oriented Dialogue (2020, methodology): “Decoding Initialized from pre-trained weights, SimpleTOD does not need to employ an advanced, more costly decoding strategy such as beam search, diverse beam search, and top-k sampling as opposed to HDSA [9] and DAMD [60].”
A Survey of Data Augmentation Approaches for NLP (2021, background): “Zhao et al. (2018) attempt to mitigate gender bias in coreference resolution by creating an augmented dataset identical to the original but biased towards the underrepresented gender (using gender swapping of entities such as replacing "he" with "she") and train on the union of the two datasets.”
Dynamic Dialogue Policy for Continual Reinforcement Learning (2022, methodology): “Similar to existing work (Shu et al., 2019; Zhang et al., 2020a) we separately predict domains, intents and slots for action prediction.”
ConvLab-2: An Open-Source Toolkit for Building, Evaluating, and Diagnosing Dialogue Systems (2020, background): “ConvLab-2 also integrates DAMD (Zhang et al., 2019) which obtains state-of-the-art results on MultiWOZ.”
Conversation Graph: Data Augmentation, Training, and Evaluation for Non-Deterministic Dialogue Management (2020, methodology): “Related to our work is a recent paper on Multi-Action Data Augmentation (Zhang et al., 2019, MADA).”
Advances and Challenges in Conversational Recommender Systems: A Survey (2021, methodology): “The BLEU score [128] is used to measure the fluency of these mod ls mimicking hum n conversations [10, 209].”
Simulated Chats for Task-oriented Dialog: Learning to Generate Conversations from Instructions (2020, methodology): “In contrast to existing methods that modify existing conversations to generate additional data (Zhang, Ou, and Yu 2020; Gao et al. 2020), we propose a new augmentation framework that harnesses the strength of such large scale language models, to simulate the crowd source data collection process and…”
Dynamic Dialogue Policy Transformer for Continual Reinforcement Learning (2022, background): “Similar to existing work (Shu et al., 2019; Zhang et al., 2020a) we separately predict domains, intents and slots for action prediction.”
SOLOIST: Few-shot Task-Oriented Dialog with A Single Pre-trained Auto-regressive Model (2020, methodology): “To tackle this, several models were proposed to handle the multi-domain dialog response generation (Pei et al., 2019; Budzianowski and Vuli´c, 2019; Mehri et al., 2019; Zhao et al., 2019; Wu et al., 2019; Zhang et al., 2019a).”
MinTL: Minimalist Transfer Learning for Task-Oriented Dialogue Systems (2020, methodology): “…FJST (Eric et al., 2019), HyST (Goel et al., 2019), SUMBT (Lee et al., 2019), SST (Chen et al., 2020), TODBERT (Wu et al., 2020), and DST-Picklist (Zhang et al., 2019a); the latter includes Neural Reading (Gao et al., 2019), TRADE (Wu et al., 2019a), COMER (Ren et al., 2019), SOM-DST (Kim et al.,…”
Learning Knowledge Bases with Parameters for Task-Oriented Dialogue Systems (2020, methodology): “We evaluate DAMD (Zhang et al., 2019a) with our plain text scorer.”
Paraphrase Augmented Task-Oriented Dialog Generation (2020, methodology): “For DAMD the process is similar since the only difference between these two models is that DAMD has an additional action span decoder between the belief span decoder and the response decoder.”
ProtoQA: A Question Answering Dataset for Prototypical Common-Sense Reasoning (2020, background): “Such sets of diverse answers represent the nature of common sense knowledge and may be useful in applications such as dialogue systems, where multiple responses are appropriate for a given context (Zhang et al., 2019b).”