Leveraging Slot Descriptions for Zero-Shot Cross-Domain Dialogue StateTracking

By Zhaojiang Lin, Bing Liu, Seungwhan Moon, Paul A. Crook, Zhenpeng Zhou, Zhiguang Wang, Zhou Yu, Andrea Madotto, Eunjoon Cho, Rajen Subba. 2021

97 citations recorded by the source corpus.

Zero-shot cross-domain dialogue state tracking (DST) enables us to handle unseen domains without the expense of collecting in-domain data. In this paper, we propose a slot descriptions enhanced generative approach for zero-shot cross-domain DST. Specifically, our model first encodes a dialogue context and a slot with a pre-trained self-attentive encoder, and generates slot value in auto-regressive manner. In addition, we incorporate Slot Type Informed Descriptions that capture the shared information of different slots to facilitates the cross-domain knowledge transfer. Experimental results on MultiWOZ shows that our model significantly improve existing state-of-the-art results in zero-shot cross-domain setting.

Subfields: Artificial Intelligence, Dialogue & Conversational AI, Meta & Continual Learning

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. Minlie Huang
2. Xiaoyan Zhu
3. Xin Jiang
4. Qun Liu
5. Fei Mi
6. Yasheng Wang
7. Qi Zhu
8. Bing Li

In Dialogue & Conversational AI

1. Minlie Huang
2. Xiaoyan Zhu
3. Qi Zhu
4. Bing Li
5. Fei Mi
6. Jason Aaron Edward Weston

In Meta & Continual Learning

1. Qi Zhu
2. Xiaoyan Zhu
3. Minlie Huang
4. Bing Li
5. Fei Mi

Citations in the corpus

• Continual Prompt Tuning for Dialog State Tracking (2022, methodology): “…while the latter183 can incorporate more information about a slot as184 a query, such as a brief natural language descrip-185 tion (Rastogi et al., 2020), slot type information186 (Lin et al., 2021), possible values (Lee et al., 2021),187 and the task definition and constraint (Mi et al.,188 2021).”
• CINS: Comprehensive Instruction for Few-shot Learning in Task-oriented Dialog Systems (2021, methodology): “Following Peng et al. (2020b); Kale and Rastogi (2020b), the “Naive” canonical representation A = ∑A i=1 ai(si = vi) is the set of actions produced by the system, where A is the total number of actions for this turn.”