Retrieval-Free Knowledge-Grounded Dialogue Response Generation with Adapters

By Yan Xu, Etsuko Ishii, Zihan Liu, Genta Indra Winata, Dan Su, Andrea Madotto, Pascale Fung. 2021

47 citations recorded by the source corpus.

To diversify and enrich generated dialogue responses, knowledge-grounded dialogue has been investigated in recent years. The existing methods tackle the knowledge grounding challenge by retrieving the relevant sentences over a large corpus and augmenting the dialogues with explicit extra information. Despite their success, however, the existing works have drawbacks on the inference efficiency. This paper proposes KnowExpert, an end-to-end framework to bypass the explicit retrieval process and inject knowledge into the pre-trained language models with lightweight adapters and adapt to the knowledge-grounded dialogue task. To the best of our knowledge, this is the first attempt to tackle this challenge without retrieval in this task under an open-domain chit-chat scenario. The experimental results show that KnowExpert performs comparably with some retrieval-based baselines while being time-efficient in inference, demonstrating the effectiveness of our proposed method.

Subfields: Artificial Intelligence, Retrieval-Augmented Generation, 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

1. Pascale Fung
2. Andrea Madotto
3. Xiang Ren
4. Wenliang Dai
5. Nayeon Lee
6. Yan Xu
7. Yang Liu
8. Etsuko Ishii

In Retrieval-Augmented Generation

1. Xiang Ren
2. Dilek Z. Hakkani-Tür

In Natural Language Processing

1. Pascale Fung
2. Andrea Madotto
3. Xiang Ren
4. Nayeon Lee
5. Yang Liu
6. Etsuko Ishii
7. Yejin Bang
8. Dilek Z. Hakkani-Tür

Citations in the corpus

• Survey of Hallucination in Natural Language Generation (2022, background): “Recently, there have been some attempts to generate informative responses without explicit knowledge inputs, but with the help of the implicit knowledge inside large pre-trained language models instead [217, 239] during the inference time.”
• Think Before You Speak: Using Self-talk to Generate Implicit Commonsense Knowledge for Response Generation (2021, background): “…knowledge retrieval upon a knowledge base (KB) (Zhou et al., 2018; Zhao et al., 2020) or cast knowledge as a latent factor in the generation (Tuan et al., 2020; Xu et al., 2021), posing limitation due to coverage in the KB or the lack of intepretability by not explicitly expressing knowledge in NL.”
• Think Before You Speak: Explicitly Generating Implicit Commonsense Knowledge for Response Generation (2021, background): “Some work also casts knowledge as a latent factor in generation (Tuan et al., 2020; Xu et al., 2021), which makes it hard to examine the quality of knowledge generation and how exactly RG uses the implicit knowledge, posing interpretability concerns.”