Datasets:
GEM
/

Modalities:
Text
Languages:
Chinese
Libraries:
Datasets
License:
magnetic commited on
Commit
f7549bd
1 Parent(s): 77023ed

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +68 -25
README.md CHANGED
@@ -20,7 +20,9 @@ task_ids:
20
  - dialogue-modeling
21
  ---
22
  # Dataset Card for RiSAWOZ
 
23
  ## Table of Contents
 
24
  - [Dataset Card for RiSAWOZ](#dataset-card-for-risawoz)
25
  - [Table of Contents](#table-of-contents)
26
  - [Dataset Description](#dataset-description)
@@ -45,25 +47,37 @@ task_ids:
45
  - [Licensing Information](#licensing-information)
46
  - [Citation Information](#citation-information)
47
  - [Contributions](#contributions)
 
48
  ## Dataset Description
49
- - **Homepage:** https://terryqj0107.github.io/RiSAWOZ_webpage
50
- - **Repository:** https://github.com/terryqj0107/RiSAWOZ
51
- - **Paper:** https://aclanthology.org/2020.emnlp-main.67.pdf
 
52
  - **Point of Contact:** Deyi Xiong (dyxiong@tju.edu.cn)
 
53
  ### Dataset Summary
 
54
  RiSAWOZ contains 11.2K human-to-human (H2H) multiturn semantically annotated dialogues, with more than 150K utterances spanning over 12 domains, which is larger than all previous annotated H2H conversational datasets. Both single- and multi-domain dialogues are constructed, accounting for 65% and 35%, respectively.
 
55
  ### Supported Tasks and Leaderboards
 
56
  - Natural Language Understanding
57
  - Dialogue State Tracking
58
  - Dialogue Context-to-Text Generation
59
  - Coreference Resolution
60
  - Unified Generative Ellipsis and Coreference Resolution
 
61
  ### Languages
62
- Mandarin Chinese
 
 
63
  ## Dataset Structure
 
64
  ### Data Instances
 
65
  A JSON formatted example of a typical instance in RiSAWOZ dataset:
66
- ```
 
67
  {
68
  "dialogue_id": "attraction_goal_4-63###6177",
69
  "goal": "attraction_goal_4-63: 你是苏州人,但不怎么出去玩。你朋友来苏州找你,你准备带他逛逛“水乡古镇”,你希望客服给你推荐个消费水平“中等”的地方。然后你要问清楚这地方“是否地铁直达”、“特点”、“门票价格”这些信息。最后,你要感谢客服的帮助,然后说再见。",
@@ -577,66 +591,93 @@ A JSON formatted example of a typical instance in RiSAWOZ dataset:
577
  ]
578
  }
579
  ```
 
580
  ### Data Fields
 
581
  The data fields are the same among all splits.
 
582
  - dialogue_id (string): dialogue ID
583
  - goal (string): natural language descriptions of the user goal
584
  - domains (list of strings): domains mentioned in current dialogue session
585
  - dialogue (list of dicts): dialog turns and corresponding annotations. Each turn includes:
586
- - turn_id (int): turn ID
587
- - turn_domain (list of strings): domain mentioned in current turn
588
- - user_utterance (string): user utterance
589
- - system_utterance (string): system utterance
590
- - belief_state (dict): dialogue state, including:
591
- - inform slot-values (dict): the slots and corresponding values informed until current turn
592
- - turn_inform (dict): the slots and corresponding values informed in current turn
593
- - turn request (dict): the slots requested in current turn
594
- - user_actions (list of lists): user dialogue acts in current turn
595
- - user_actions (list of lists): system dialogue acts in current turn
596
- - db_results (list of strings): database search results
597
- - segmented_user_utterance (string): word segmentation result of user utterance
598
- - segmented_system_utterance (string): word segmentation result of system utterance
 
599
  ### Data Splits
 
600
  | | train | validation | test |
601
  | --------- | -----: | ---------: | ----: |
602
  | #dialogues | 10000 | 600 | 600 |
603
  | #turns | 134580 | 8116 | 9286 |
604
  | Avg. turns | 13.5 | 13.5 | 15.5 |
 
605
  ## Dataset Creation
 
606
  ### Curation Rationale
 
607
  Gather human-to-human dialog in Chinese.
 
608
  ### Communicative Goal
 
609
  Generate system response given dialogue context across multiple domains.
 
610
  ### Annotations
 
611
  #### Annotation process
 
612
  - Does the dataset have additional annotations for each instance?
613
- - crowd-sourced
614
  - What is the number of raters
615
- - 51<n<100
616
  - Describe the qualifications required of an annotator.
617
- - Chinese native speaker
618
  - How many annotators saw each training example?
619
- - 3
620
  - How many annotators saw each test example?
621
- - 3
 
622
  ### Personal and Sensitive Information
 
623
  The slots and values as well as utterances do not contain any personal information.
 
624
  ## Considerations for Using the Data
 
625
  ### Social Impact of Dataset
 
626
  RiSAWOZ is the first large-scale multi-domain Chinese Wizard-of-Oz dataset with rich semantic annotations.
 
627
  ## Additional Information
 
628
  ### Dataset Curators
 
629
  - Jun Quan (Soochow University, Suzhou, China)
630
  - Shian Zhang (Soochow University, Suzhou, China)
631
  - Qian Cao (Soochow University, Suzhou, China)
632
  - Zizhong Li (Tianjin University, Tianjin, China)
633
  - Deyi Xiong (Tianjin University, Tianjin, China)
 
634
  ### Funding Information
 
635
  the National Key Research and Development Project
 
636
  ### Licensing Information
 
637
  cc-by-4.0: Creative Commons Attribution 4.0 International
 
638
  ### Citation Information
639
- ```
 
640
  @inproceedings{quan-etal-2020-risawoz,
641
  title = "{R}i{SAWOZ}: A Large-Scale Multi-Domain {W}izard-of-{O}z Dataset with Rich Semantic Annotations for Task-Oriented Dialogue Modeling",
642
  author = "Quan, Jun and
@@ -653,5 +694,7 @@ cc-by-4.0: Creative Commons Attribution 4.0 International
653
  pages = "930--940",
654
  }
655
  ```
 
656
  ### Contributions
657
- Thanks to Tianhao Shen, Chaobin You and Deyi Xiong for adding this dataset to GEM.
 
 
20
  - dialogue-modeling
21
  ---
22
  # Dataset Card for RiSAWOZ
23
+
24
  ## Table of Contents
25
+
26
  - [Dataset Card for RiSAWOZ](#dataset-card-for-risawoz)
27
  - [Table of Contents](#table-of-contents)
28
  - [Dataset Description](#dataset-description)
 
47
  - [Licensing Information](#licensing-information)
48
  - [Citation Information](#citation-information)
49
  - [Contributions](#contributions)
50
+
51
  ## Dataset Description
52
+
53
+ - **Homepage:** <https://terryqj0107.github.io/RiSAWOZ_webpage>
54
+ - **Repository:** <https://github.com/terryqj0107/RiSAWOZ>
55
+ - **Paper:** <https://aclanthology.org/2020.emnlp-main.67.pdf>
56
  - **Point of Contact:** Deyi Xiong (dyxiong@tju.edu.cn)
57
+
58
  ### Dataset Summary
59
+
60
  RiSAWOZ contains 11.2K human-to-human (H2H) multiturn semantically annotated dialogues, with more than 150K utterances spanning over 12 domains, which is larger than all previous annotated H2H conversational datasets. Both single- and multi-domain dialogues are constructed, accounting for 65% and 35%, respectively.
61
+
62
  ### Supported Tasks and Leaderboards
63
+
64
  - Natural Language Understanding
65
  - Dialogue State Tracking
66
  - Dialogue Context-to-Text Generation
67
  - Coreference Resolution
68
  - Unified Generative Ellipsis and Coreference Resolution
69
+
70
  ### Languages
71
+
72
+ Mandarin Chinese (zh-CN)
73
+
74
  ## Dataset Structure
75
+
76
  ### Data Instances
77
+
78
  A JSON formatted example of a typical instance in RiSAWOZ dataset:
79
+
80
+ ```JSON
81
  {
82
  "dialogue_id": "attraction_goal_4-63###6177",
83
  "goal": "attraction_goal_4-63: 你是苏州人,但不怎么出去玩。你朋友来苏州找你,你准备带他逛逛“水乡古镇”,你希望客服给你推荐个消费水平“中等”的地方。然后你要问清楚这地方“是否地铁直达”、“特点”、“门票价格”这些信息。最后,你要感谢客服的帮助,然后说再见。",
 
591
  ]
592
  }
593
  ```
594
+
595
  ### Data Fields
596
+
597
  The data fields are the same among all splits.
598
+
599
  - dialogue_id (string): dialogue ID
600
  - goal (string): natural language descriptions of the user goal
601
  - domains (list of strings): domains mentioned in current dialogue session
602
  - dialogue (list of dicts): dialog turns and corresponding annotations. Each turn includes:
603
+ - turn_id (int): turn ID
604
+ - turn_domain (list of strings): domain mentioned in current turn
605
+ - user_utterance (string): user utterance
606
+ - system_utterance (string): system utterance
607
+ - belief_state (dict): dialogue state, including:
608
+ - inform slot-values (dict): the slots and corresponding values informed until current turn
609
+ - turn_inform (dict): the slots and corresponding values informed in current turn
610
+ - turn request (dict): the slots requested in current turn
611
+ - user_actions (list of lists): user dialogue acts in current turn
612
+ - user_actions (list of lists): system dialogue acts in current turn
613
+ - db_results (list of strings): database search results
614
+ - segmented_user_utterance (string): word segmentation result of user utterance
615
+ - segmented_system_utterance (string): word segmentation result of system utterance
616
+
617
  ### Data Splits
618
+
619
  | | train | validation | test |
620
  | --------- | -----: | ---------: | ----: |
621
  | #dialogues | 10000 | 600 | 600 |
622
  | #turns | 134580 | 8116 | 9286 |
623
  | Avg. turns | 13.5 | 13.5 | 15.5 |
624
+
625
  ## Dataset Creation
626
+
627
  ### Curation Rationale
628
+
629
  Gather human-to-human dialog in Chinese.
630
+
631
  ### Communicative Goal
632
+
633
  Generate system response given dialogue context across multiple domains.
634
+
635
  ### Annotations
636
+
637
  #### Annotation process
638
+
639
  - Does the dataset have additional annotations for each instance?
640
+ - crowd-sourced
641
  - What is the number of raters
642
+ - 51<n<100
643
  - Describe the qualifications required of an annotator.
644
+ - Chinese native speaker
645
  - How many annotators saw each training example?
646
+ - 3
647
  - How many annotators saw each test example?
648
+ - 3
649
+
650
  ### Personal and Sensitive Information
651
+
652
  The slots and values as well as utterances do not contain any personal information.
653
+
654
  ## Considerations for Using the Data
655
+
656
  ### Social Impact of Dataset
657
+
658
  RiSAWOZ is the first large-scale multi-domain Chinese Wizard-of-Oz dataset with rich semantic annotations.
659
+
660
  ## Additional Information
661
+
662
  ### Dataset Curators
663
+
664
  - Jun Quan (Soochow University, Suzhou, China)
665
  - Shian Zhang (Soochow University, Suzhou, China)
666
  - Qian Cao (Soochow University, Suzhou, China)
667
  - Zizhong Li (Tianjin University, Tianjin, China)
668
  - Deyi Xiong (Tianjin University, Tianjin, China)
669
+
670
  ### Funding Information
671
+
672
  the National Key Research and Development Project
673
+
674
  ### Licensing Information
675
+
676
  cc-by-4.0: Creative Commons Attribution 4.0 International
677
+
678
  ### Citation Information
679
+
680
+ ```BibTeX
681
  @inproceedings{quan-etal-2020-risawoz,
682
  title = "{R}i{SAWOZ}: A Large-Scale Multi-Domain {W}izard-of-{O}z Dataset with Rich Semantic Annotations for Task-Oriented Dialogue Modeling",
683
  author = "Quan, Jun and
 
694
  pages = "930--940",
695
  }
696
  ```
697
+
698
  ### Contributions
699
+
700
+ Thanks to Tianhao Shen, Chaobin You and Deyi Xiong for adding this dataset to GEM.