Publications

2020

1. NLPCC

   A Large-Scale Chinese Short-Text Conversation Dataset

   Wang, Yida, Ke, Pei, Zheng, Yinhe,  Huang, Kaili, Jiang, Yong, Zhu, Xiaoyan, and Huang, Minlie

   In Natural Language Processing and Chinese Computing (NLPCC), 2020

   Abs Bib PDF Code

   The advancements of neural dialogue generation models show promising results on modeling short-text conversations. However, training such models usually needs a large-scale high-quality dialogue corpus, which is hard to access. In this paper, we present a large-scale cleaned Chinese conversation dataset LCCC, which contains a base version (6.8 million dialogues) and a large version (12.0 million dialogues). The quality of our dataset is ensured by a rigorous data cleaning pipeline, which is built based on a set of rules and a classifier that is trained on manually annotated 110K dialogue pairs. We also release pre-training dialogue models which are trained on LCCC-base and LCCC-large respectively. The cleaned dataset and the pre-training models will facilitate the research of short-text conversation modeling. All the models and datasets are available at https://github.com/thu-coai/CDial-GPT.

   @inproceedings{wang2020chinese,
     title = {A Large-Scale Chinese Short-Text Conversation Dataset},
     author = {Wang, Yida and Ke, Pei and Zheng, Yinhe and Huang, Kaili and Jiang, Yong and Zhu, Xiaoyan and Huang, Minlie},
     booktitle = {Natural Language Processing and Chinese Computing (NLPCC),},
     year = {2020},
   }

2. TACL

   CrossWOZ: A Large-Scale Chinese Cross-Domain Task-Oriented Dialogue Dataset

   Zhu, Qi,  Huang, Kaili, Zhang, Zheng, Zhu, Xiaoyan, and Huang, Minlie

   Transactions of the Association for Computational Linguistics (TACL), 2020

   Abs Bib PDF Code

   To advance multi-domain (cross-domain) dialogue modeling as well as alleviate the shortage of Chinese task-oriented datasets, we propose CrossWOZ, the first large-scale Chinese Cross-Domain Wizard-of-Oz task-oriented dataset. It contains 6K dialogue sessions and 102K utterances for 5 domains, including hotel, restaurant, attraction, metro, and taxi. Moreover, the corpus contains rich annotation of dialogue states and dialogue acts on both user and system sides. About 60% of the dialogues have cross-domain user goals that favor inter-domain dependency and encourage natural transition across domains in conversation. We also provide a user simulator and several benchmark models for pipelined task-oriented dialogue systems, which will facilitate researchers to compare and evaluate their models on this corpus. The large size and rich annotation of CrossWOZ make it suitable to investigate a variety of tasks in cross-domain dialogue modeling, such as dialogue state tracking, policy learning, user simulation, etc.

   @article{a-zhu-etal-2020-crosswoz,
     title = {{C}ross{WOZ}: A Large-Scale {C}hinese Cross-Domain Task-Oriented Dialogue Dataset},
     author = {Zhu, Qi and Huang, Kaili and Zhang, Zheng and Zhu, Xiaoyan and Huang, Minlie},
     journal = {Transactions of the Association for Computational Linguistics (TACL),},
     volume = {8},
     year = {2020},
     address = {Cambridge, MA},
     publisher = {MIT Press},
     url = {https://aclanthology.org/2020.tacl-1.19},
     doi = {10.1162/tacl_a_00314},
     pages = {281--295}
   }

3. ACL

   KdConv: A Chinese Multi-domain Dialogue Dataset Towards Multi-turn Knowledge-driven Conversation

   Zhou, Hao, Zheng, Chujie,  Huang, Kaili, Huang, Minlie, and Zhu, Xiaoyan

   In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL), Jul 2020

   Abs Bib PDF Code

   The research of knowledge-driven conversational systems is largely limited due to the lack of dialog data which consist of multi-turn conversations on multiple topics and with knowledge annotations. In this paper, we propose a Chinese multi-domain knowledge-driven conversation dataset, KdConv, which grounds the topics in multi-turn conversations to knowledge graphs. Our corpus contains 4.5K conversations from three domains (film, music, and travel), and 86K utterances with an average turn number of 19.0. These conversations contain in-depth discussions on related topics and natural transition between multiple topics. To facilitate the following research on this corpus, we provide several benchmark models. Comparative results show that the models can be enhanced by introducing background knowledge, yet there is still a large space for leveraging knowledge to model multi-turn conversations for further research. Results also show that there are obvious performance differences between different domains, indicating that it is worth to further explore transfer learning and domain adaptation. The corpus and benchmark models are publicly available.

   @inproceedings{zhou-etal-2020-kdconv,
     title = {{K}d{C}onv: A {C}hinese Multi-domain Dialogue Dataset Towards Multi-turn Knowledge-driven Conversation},
     author = {Zhou, Hao and Zheng, Chujie and Huang, Kaili and Huang, Minlie and Zhu, Xiaoyan},
     booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL),},
     month = jul,
     year = {2020},
     address = {Online},
     publisher = {Association for Computational Linguistics},
     url = {https://www.aclweb.org/anthology/2020.acl-main.635},
     doi = {10.18653/v1/2020.acl-main.635},
     pages = {7098--7108}
   }