Exploring Prompt-based Few-shot Learning for Grounded Dialog Generation

By Chujie Zheng, Minlie Huang. 2021

47 citations recorded by the source corpus.

Dialog models can be greatly strengthened through grounding on various external information, but grounded dialog corpora are usually not naturally accessible. In this work, we focus on the few-shot learning for grounded dialog generation (GDG). We first propose a simple prompting method for GDG tasks, where different constructs of model input, such as the grounding source and the conversation context, are distinguished through continuous or discrete prompts. On three typical GDG tasks, we empirically demonstrate and analyze in-depth the effectiveness of our method. We then conduct extensive experiments to thoroughly investigate how our prompting method works with different pre-trained models. We show that prompted language models perform superiorly to conversational models, and further analyze various factors that influence the effects of prompting. Overall, our work introduces a prompt-based perspective to the few-shot learning for GDG tasks, and provides valuable findings and insights for future research.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Prompting & Reasoning, Dialogue & Conversational AI

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

1. Xiang Ren
2. Yang Liu
3. Dilek Z. Hakkani-Tür
4. Jay Pujara
5. Pei Zhou
6. Karthik Gopalakrishnan
7. Seokhwan Kim
8. Behnam Hedayatnia

In Computer Vision and Pattern Recognition

1. Xiang Ren
2. Dilek Z. Hakkani-Tür
3. Karthik Gopalakrishnan
4. Pei Zhou
5. Seokhwan Kim
6. Jay Pujara
7. Yang Liu
8. Behnam Hedayatnia

In Prompting & Reasoning

1. Xiang Ren
2. Pei Zhou
3. Yang Liu

In Dialogue & Conversational AI

1. Dilek Z. Hakkani-Tür
2. Behnam Hedayatnia

Citations in the corpus

• Think Before You Speak: Using Self-talk to Generate Implicit Commonsense Knowledge for Response Generation (2021, background): “Specifically, (Zheng and Huang, 2021) found that using prompts helps with few-shot knowledge-grounded RG where knowledge is provided to the model.”
• Think Before You Speak: Explicitly Generating Implicit Commonsense Knowledge for Response Generation (2021, background): “More recent work has found that NL prompts help LMs to perform better on various downstream tasks, including natural language generation (NLG) (Brown et al., 2020; Liu et al., 2021; Zheng and Huang, 2021).”