Datasets:
GEM
/

Modalities:
Text
Languages:
English
ArXiv:
Libraries:
Datasets
License:
parquet-converter commited on
Commit
1283d6b
1 Parent(s): f272b92

Update parquet files

Browse files
.gitattributes DELETED
@@ -1,27 +0,0 @@
1
- *.7z filter=lfs diff=lfs merge=lfs -text
2
- *.arrow filter=lfs diff=lfs merge=lfs -text
3
- *.bin filter=lfs diff=lfs merge=lfs -text
4
- *.bin.* filter=lfs diff=lfs merge=lfs -text
5
- *.bz2 filter=lfs diff=lfs merge=lfs -text
6
- *.ftz filter=lfs diff=lfs merge=lfs -text
7
- *.gz filter=lfs diff=lfs merge=lfs -text
8
- *.h5 filter=lfs diff=lfs merge=lfs -text
9
- *.joblib filter=lfs diff=lfs merge=lfs -text
10
- *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
- *.model filter=lfs diff=lfs merge=lfs -text
12
- *.msgpack filter=lfs diff=lfs merge=lfs -text
13
- *.onnx filter=lfs diff=lfs merge=lfs -text
14
- *.ot filter=lfs diff=lfs merge=lfs -text
15
- *.parquet filter=lfs diff=lfs merge=lfs -text
16
- *.pb filter=lfs diff=lfs merge=lfs -text
17
- *.pt filter=lfs diff=lfs merge=lfs -text
18
- *.pth filter=lfs diff=lfs merge=lfs -text
19
- *.rar filter=lfs diff=lfs merge=lfs -text
20
- saved_model/**/* filter=lfs diff=lfs merge=lfs -text
21
- *.tar.* filter=lfs diff=lfs merge=lfs -text
22
- *.tflite filter=lfs diff=lfs merge=lfs -text
23
- *.tgz filter=lfs diff=lfs merge=lfs -text
24
- *.xz filter=lfs diff=lfs merge=lfs -text
25
- *.zip filter=lfs diff=lfs merge=lfs -text
26
- *.zstandard filter=lfs diff=lfs merge=lfs -text
27
- *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
README.md DELETED
@@ -1,702 +0,0 @@
1
- ---
2
- annotations_creators:
3
- - crowd-sourced
4
- language_creators:
5
- - unknown
6
- language:
7
- - en
8
- license:
9
- - cc-by-sa-4.0
10
- multilinguality:
11
- - unknown
12
- size_categories:
13
- - unknown
14
- source_datasets:
15
- - original
16
- task_categories:
17
- - conversational
18
- task_ids: []
19
- pretty_name: schema_guided_dialog
20
- tags:
21
- - dialog-response-generation
22
- ---
23
-
24
- # Dataset Card for GEM/schema_guided_dialog
25
-
26
- ## Dataset Description
27
-
28
- - **Homepage:** n/a
29
- - **Repository:** [Github[(https://github.com/google-research-datasets/dstc8-schema-guided-dialogue)
30
- - **Paper:** https://arxiv.org/abs/1909.05855
31
- - **Leaderboard:** N/A
32
- - **Point of Contact:** Abhinav Rastogi
33
-
34
- ### Link to Main Data Card
35
-
36
- You can find the main data card on the [GEM Website](https://gem-benchmark.com/data_cards/schema_guided_dialog).
37
-
38
- ### Dataset Summary
39
-
40
- The GEM version of this dataset functions as a response generation dataset. The input specifies dialog acts that a model needs to verbalize. The Schema-Guided Dialog dataset is challenging since it comprises multiple domains from hotel and travel to restaurants, and a wide range of dialog acts. The context of each conversation is provided as well.
41
-
42
- You can load the dataset via:
43
- ```
44
- import datasets
45
- data = datasets.load_dataset('GEM/schema_guided_dialog')
46
- ```
47
- The data loader can be found [here](https://huggingface.co/datasets/GEM/schema_guided_dialog).
48
-
49
- #### website
50
- n/a
51
-
52
- #### paper
53
- [Arxiv](https://arxiv.org/abs/1909.05855)
54
-
55
- #### authors
56
- Abhinav Rastogi, Xiaoxue Zang, Srinivas Sunkara, Raghav Gupta, Pranav Khaitan, Amir Fayazi, Maria Wang, and Guan-Lin Chao
57
-
58
- ## Dataset Overview
59
-
60
- ### Where to find the Data and its Documentation
61
-
62
- #### Download
63
-
64
- <!-- info: What is the link to where the original dataset is hosted? -->
65
- <!-- scope: telescope -->
66
- [Github[(https://github.com/google-research-datasets/dstc8-schema-guided-dialogue)
67
-
68
- #### Paper
69
-
70
- <!-- info: What is the link to the paper describing the dataset (open access preferred)? -->
71
- <!-- scope: telescope -->
72
- [Arxiv](https://arxiv.org/abs/1909.05855)
73
-
74
- #### BibTex
75
-
76
- <!-- info: Provide the BibTex-formatted reference for the dataset. Please use the correct published version (ACL anthology, etc.) instead of google scholar created Bibtex. -->
77
- <!-- scope: microscope -->
78
- ```
79
- {
80
- @inproceedings{rastogi2020towards,
81
- title={Towards scalable multi-domain conversational agents: The schema-guided dialogue dataset},
82
- author={Rastogi, Abhinav and Zang, Xiaoxue and Sunkara, Srinivas and Gupta, Raghav and Khaitan, Pranav},
83
- booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
84
- volume={34},
85
- number={05},
86
- pages={8689--8696},
87
- year={2020}
88
- }
89
- ```
90
-
91
- #### Contact Name
92
-
93
- <!-- quick -->
94
- <!-- info: If known, provide the name of at least one person the reader can contact for questions about the dataset. -->
95
- <!-- scope: periscope -->
96
- Abhinav Rastogi
97
-
98
- #### Contact Email
99
-
100
- <!-- info: If known, provide the email of at least one person the reader can contact for questions about the dataset. -->
101
- <!-- scope: periscope -->
102
- schema-guided-dst@google.com
103
-
104
- #### Has a Leaderboard?
105
-
106
- <!-- info: Does the dataset have an active leaderboard? -->
107
- <!-- scope: telescope -->
108
- no
109
-
110
-
111
- ### Languages and Intended Use
112
-
113
- #### Multilingual?
114
-
115
- <!-- quick -->
116
- <!-- info: Is the dataset multilingual? -->
117
- <!-- scope: telescope -->
118
- no
119
-
120
- #### Covered Languages
121
-
122
- <!-- quick -->
123
- <!-- info: What languages/dialects are covered in the dataset? -->
124
- <!-- scope: telescope -->
125
- `English`
126
-
127
- #### Whose Language?
128
-
129
- <!-- info: Whose language is in the dataset? -->
130
- <!-- scope: periscope -->
131
- The language structure is machine-generated, and the language realizations are produced by crowd workers.
132
- The dataset paper does not provide demographic information for the crowd workers.
133
-
134
-
135
- #### License
136
-
137
- <!-- quick -->
138
- <!-- info: What is the license of the dataset? -->
139
- <!-- scope: telescope -->
140
- cc-by-sa-4.0: Creative Commons Attribution Share Alike 4.0 International
141
-
142
- #### Intended Use
143
-
144
- <!-- info: What is the intended use of the dataset? -->
145
- <!-- scope: microscope -->
146
- The Schema-Guided Dialogue (SGD) dataset contains 18K multi-domain task-oriented dialogues between a human and a virtual assistant, which covers 17 domains ranging from banks and events to media, calendar, travel, and weather.
147
- The language presents in the datset is only English.
148
- The SGD dataset provides a challenging testbed for a number of tasks in task-oriented dialogue, including language understanding, slot filling, dialogue state tracking and response generation.
149
- For the creation of the SGD dataset, they developed a multi-domain dialogue simulator that generates dialogue outlines over an arbitrary combination of APIs, dialogue states and system actions. Then, they used a crowd-sourcing procedure to paraphrase these outlines to natural language utterances.
150
- This novel crowd-sourcing procedure preserves all annotations obtained from the simulator and does not require any extra annotations after dialogue collection.
151
-
152
-
153
- #### Primary Task
154
-
155
- <!-- info: What primary task does the dataset support? -->
156
- <!-- scope: telescope -->
157
- Dialog Response Generation
158
-
159
- #### Communicative Goal
160
-
161
- <!-- quick -->
162
- <!-- info: Provide a short description of the communicative goal of a model trained for this task on this dataset. -->
163
- <!-- scope: periscope -->
164
- The goal of a speaker who generates the target utterance is to help users accomplish tasks including but not limited to finding flights, booking restaurants, searching for nearby events and movies.
165
-
166
-
167
-
168
- ### Credit
169
-
170
- #### Curation Organization Type(s)
171
-
172
- <!-- info: In what kind of organization did the dataset curation happen? -->
173
- <!-- scope: telescope -->
174
- `industry`
175
-
176
- #### Curation Organization(s)
177
-
178
- <!-- info: Name the organization(s). -->
179
- <!-- scope: periscope -->
180
- Google
181
-
182
- #### Dataset Creators
183
-
184
- <!-- info: Who created the original dataset? List the people involved in collecting the dataset and their affiliation(s). -->
185
- <!-- scope: microscope -->
186
- Abhinav Rastogi, Xiaoxue Zang, Srinivas Sunkara, Raghav Gupta, Pranav Khaitan, Amir Fayazi, Maria Wang, and Guan-Lin Chao
187
-
188
- #### Funding
189
-
190
- <!-- info: Who funded the data creation? -->
191
- <!-- scope: microscope -->
192
- Google
193
-
194
- #### Who added the Dataset to GEM?
195
-
196
- <!-- info: Who contributed to the data card and adding the dataset to GEM? List the people+affiliations involved in creating this data card and who helped integrate this dataset into GEM. -->
197
- <!-- scope: microscope -->
198
- Wanyu Du wrote the initial data card and Yacine Jernite the data loader. Simon Mille updated the data card with the additional splits. Sebastian Gehrmann migrated the data card and loader to the v2 version and extended the missing information.
199
-
200
-
201
- ### Dataset Structure
202
-
203
- #### Data Fields
204
-
205
- <!-- info: List and describe the fields present in the dataset. -->
206
- <!-- scope: telescope -->
207
- Each dialog instance has the following fields:
208
-
209
- * `dialogue_id`: A unique identifier for a dialogue.
210
- * `services`: A list of services present in the dialogue.
211
- * `turns`: A list of annotated system or user utterances. Each turn consists of the following fields:
212
- * `speaker`: The speaker for the turn, either `USER` or `SYSTEM`.
213
- * `utterance`: A string containing the natural language utterance.
214
- * `frames`: A list of frames, each frame containing annotations for a single service and consists of the following fields:
215
- * `service`: The name of the service corresponding to the frame. The slots and intents used in the following fields are taken from the schema of this service.
216
- * `slots`: A list of slot spans in the utterance, only provided for non-categorical slots. Each slot span contains the following fields:
217
- * `slot`: The name of the slot.
218
- * `start`: The index of the starting character in the utterance corresponding to the slot value.
219
- * `exclusive_end`: The index of the character just after the last character corresponding to the slot value in the utterance.
220
- * `actions`: A list of actions corresponding to the system. Each action has the following fields:
221
- * `act`: The type of action.
222
- * `slot`: (optional) A slot argument for some of the actions.
223
- * `values`: (optional) A list of values assigned to the slot. If the values list is non-empty, then the slot must be present.
224
- * `canonical_values`: (optional) The values in their canonicalized form as used by the service. It is a list of strings of the same length as values.
225
- * `service_call`: (system turns only, optional) The request sent to the service. It consists of the following fields:
226
- * `method`: The name of the intent or function of the service or API being executed.
227
- * `parameters`: A pair of lists of the same lengths: `parameter_slot_name` contains slot names and `parameter_canonical_value` contains the corresponding values in their canonicalized form.
228
- * `service_results`: (system turns only, optional) A list of entities containing the results obtained from the service. It is only available for turns in which a service call is made. Each entity is represented as a pair of lists of the same length: `service_slot_name` contains slot names and `service_canonical_value` contains the corresponding canonical values.
229
- * `state`: (user turns only) The dialogue state corresponding to the service. It consists of the following fields:
230
- * `active_intent`: The intent corresponding to the service of the frame which is currently being fulfilled by the system. It takes the value "NONE" if none of the intents are active.
231
- * `requested_slots`: A list of slots requested by the user in the current turn.
232
- * `slot_values`: A pair of lists of the same lengths: `slot_name` contains slot names and `slot_value_list` contains the corresponding lists of strings. For categorical slots, this list contains a single value assigned to the slot. For non-categorical slots, all the values in this list are spoken variations of each other and are equivalent (e.g, "6 pm", "six in the evening", "evening at 6" etc.).
233
-
234
-
235
-
236
- #### Example Instance
237
-
238
- <!-- info: Provide a JSON formatted example of a typical instance in the dataset. -->
239
- <!-- scope: periscope -->
240
- ```
241
- {'dialogue_id': '1_00000',
242
- 'services': ['Restaurants_1'],
243
- 'turns':
244
- {'frames':
245
- [{'actions': [{'act': [6],
246
- 'canonical_values': [['FindRestaurants']],
247
- 'slot': ['intent'],
248
- 'values': [['FindRestaurants']]}],
249
- 'service': ['Restaurants_1'],
250
- 'service_call': [{'method': '',
251
- 'parameters': {'parameter_canonical_value': [],
252
- 'parameter_slot_name': []}}],
253
- 'service_results': [{'service_results_list': []}],
254
- 'slots': [{'exclusive_end': [], 'slot': [], 'start': []}],
255
- 'state': [{'active_intent': 'FindRestaurants',
256
- 'requested_slots': [],
257
- 'slot_values': {'slot_name': [], 'slot_value_list': []}}]},
258
- {'actions': [{'act': [13],
259
- 'canonical_values': [[]],
260
- 'slot': ['city'],
261
- 'values': [[]]}],
262
- 'service': ['Restaurants_1'],
263
- 'service_call': [{'method': '',
264
- 'parameters': {'parameter_canonical_value': [],
265
- 'parameter_slot_name': []}}],
266
- 'service_results': [{'service_results_list': []}],
267
- 'slots': [{'exclusive_end': [], 'slot': [], 'start': []}],
268
- 'state': [{'active_intent': '',
269
- 'requested_slots': [],
270
- 'slot_values': {'slot_name': [], 'slot_value_list': []}}]},
271
- ...,]}
272
- 'speaker': [0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1],
273
- 'utterance': [
274
- 'I am feeling hungry so I would like to find a place to eat.',
275
- 'Do you have a specific which you want the eating place to be located at?',
276
- 'I would like for it to be in San Jose.',
277
- 'Is there a specific cuisine type you enjoy, such as Mexican, Italian or something else?',
278
- 'I usually like eating the American type of food.',
279
- 'I see that at 71 Saint Peter there is a good restaurant which is in San Jose.',
280
- 'Can you give me the address of this restaurant.',
281
- 'If you want to go to this restaurant you can find it at 71 North San Pedro Street.',
282
- 'Can you give me the phone number that I can contact them with?',
283
- 'If you want to phone them you can at 408-971-8523.',
284
- 'Is there some other restaurant which you can suggest?',
285
- 'How would you like Bazille restaurant which is situated in San Jose.',
286
- 'Do you have another restaurant matching my needs? For example a restaurant which is economical and is located in Palo Alto.',
287
- 'I see that 7 restaurants suit to what you requested. Bird Dog seems as a good restaurant and is located in Palo Alto.',
288
- 'Alright, that seems good. I would like to make a booking at this restaurant.',
289
- 'For which time do you want the booking to be?',
290
- 'I will be eating there at 11:30 am so make it for then.',
291
- 'Can you please confirm that you want to book a table for 2 at 11:30 am at the Bird Dog restaurant in Palo Alto for today.',
292
- 'That suits me well. Can you tell me if they feature live music?',
293
- 'Your booking has been made without errors, but unfortunately they do not have live music.',
294
- 'Will I be able to find liquor there? Can you give me the address of their location?',
295
- 'The restaurant is located at 420 Ramona Street. Unfortunately they do not serve alcohol at the restaurant.',
296
- 'I appreciate it very much. That would be all.',
297
- 'Have a good time!'
298
- ]}
299
- ```
300
-
301
- #### Data Splits
302
-
303
- <!-- info: Describe and name the splits in the dataset if there are more than one. -->
304
- <!-- scope: periscope -->
305
- The dataset is split into a train, validation, and test set with the following sizes:
306
-
307
- | | Train | Validation | Test |
308
- | --- | --- | --- | --- |
309
- | \# of dialogues | 16142 | 2482 | 4201 |
310
- | \# of turns | 48426 | 7446 | 12603 |
311
-
312
- #### Splitting Criteria
313
-
314
- <!-- info: Describe any criteria for splitting the data, if used. If there are differences between the splits (e.g., if the training annotations are machine-generated and the dev and test ones are created by humans, or if different numbers of annotators contributed to each example), describe them here. -->
315
- <!-- scope: microscope -->
316
- The data is generally split i.i.d, but some topics only appear in training and some only for testing. For example, the domains Messaging, Payment, and Train are test-only.
317
-
318
-
319
-
320
- ## Dataset in GEM
321
-
322
- ### Rationale for Inclusion in GEM
323
-
324
- #### Why is the Dataset in GEM?
325
-
326
- <!-- info: What does this dataset contribute toward better generation evaluation and why is it part of GEM? -->
327
- <!-- scope: microscope -->
328
- This dataset comprises a wide range of dialog capabilities and thus enables the evaluation of many more generation capabilities of comparable datasets. Its collection methodology ensures a high diversity but also high quality of the data.
329
-
330
- #### Similar Datasets
331
-
332
- <!-- info: Do other datasets for the high level task exist? -->
333
- <!-- scope: telescope -->
334
- yes
335
-
336
- #### Unique Language Coverage
337
-
338
- <!-- info: Does this dataset cover other languages than other datasets for the same task? -->
339
- <!-- scope: periscope -->
340
- no
341
-
342
- #### Difference from other GEM datasets
343
-
344
- <!-- info: What else sets this dataset apart from other similar datasets in GEM? -->
345
- <!-- scope: microscope -->
346
- The domains a lot more diverse than other datasets.
347
-
348
- #### Ability that the Dataset measures
349
-
350
- <!-- info: What aspect of model ability can be measured with this dataset? -->
351
- <!-- scope: periscope -->
352
- surface realization, compositionality.
353
-
354
-
355
- ### GEM-Specific Curation
356
-
357
- #### Modificatied for GEM?
358
-
359
- <!-- info: Has the GEM version of the dataset been modified in any way (data, processing, splits) from the original curated data? -->
360
- <!-- scope: telescope -->
361
- yes
362
-
363
- #### GEM Modifications
364
-
365
- <!-- info: What changes have been made to he original dataset? -->
366
- <!-- scope: periscope -->
367
- `data points modified`
368
-
369
- #### Modification Details
370
-
371
- <!-- info: For each of these changes, described them in more details and provided the intended purpose of the modification -->
372
- <!-- scope: microscope -->
373
- We are focusing on the response-generation part of the dataset and thus reformatted the dataset to treat the service agent utterances as the targets to be generated and the previous customer utterance and the agent's dialog act as the input. We additionally reformat the dialog acts to directly conform to the format described in this [paper](https://arxiv.org/abs/2004.15006).
374
-
375
-
376
- #### Additional Splits?
377
-
378
- <!-- info: Does GEM provide additional splits to the dataset? -->
379
- <!-- scope: telescope -->
380
- yes
381
-
382
- #### Split Information
383
-
384
- <!-- info: Describe how the new splits were created -->
385
- <!-- scope: periscope -->
386
- 9 challenge sets for Schema-Guided Dialog were added to the GEM evaluation suite.
387
-
388
- 1. We created subsets of the training and development sets of 500 randomly selected inputs each.
389
- 2. We applied 5 transformations to respectively 5 sets of 500 randomly selected inputs: (i) back-translation, (ii)-(iii) introduction of typographical errors, using Butterfingers with two thresholds (0.02 and 0.05), resulting in two sets with different amounts of typos introduced (there are more typos with the 0.05 threshold than with the 0.02 one), (iv) removal of final punctuations (when any), and (v) input scrambling, for which the order of the dialogue acts was randomly reassigned.
390
- 3. For the input size, we created subpopulations based on the number of dialogue acts in the input.
391
-
392
- | DA number | Frequency English |
393
- |---------------|-------------------|
394
- | 1 | 5049 |
395
- | 2 | 2517 |
396
- | 3 | 1328 |
397
- | 4 | 469 |
398
- | 5 | 335 |
399
- | 6 | 256 |
400
- | 7 | 46 |
401
-
402
- We also split the test data according to the type of dialogue act, represented by cardinal numbers in the dataset.
403
-
404
- | DA type | Frequency English |
405
- |--------------|-------------------|
406
- | 2 | 1397 |
407
- | 3 | 983 |
408
- | 4 | 1027 |
409
- | 5 | 958 |
410
- | 9 | 72 |
411
- | 10 | 1024 |
412
- | 11 | 1246 |
413
- | 12 | 500 |
414
- | 13 | 2078 |
415
- | 15 | 715 |
416
-
417
-
418
- #### Split Motivation
419
-
420
- <!-- info: What aspects of the model's generation capacities were the splits created to test? -->
421
- <!-- scope: periscope -->
422
- Generalization and Robustness.
423
-
424
-
425
- ### Getting Started with the Task
426
-
427
- #### Pointers to Resources
428
-
429
- <!-- info: Getting started with in-depth research on the task. Add relevant pointers to resources that researchers can consult when they want to get started digging deeper into the task. -->
430
- <!-- scope: microscope -->
431
- * [Paper for dataset and DST baseline](https://arxiv.org/pdf/1909.05855.pdf)
432
- * [DSTC8 overview paper](https://arxiv.org/pdf/2002.01359.pdf)
433
- * [Code for DST baseline](https://github.com/google-research/google-research/tree/master/schema_guided_dst)
434
- * [Natural language generation baseline paper](https://arxiv.org/pdf/2004.15006.pdf)
435
- * [Blog post announcing the dataset](https://ai.googleblog.com/2019/10/introducing-schema-guided-dialogue.html)
436
-
437
-
438
-
439
-
440
- ## Previous Results
441
-
442
- ### Previous Results
443
-
444
- #### Measured Model Abilities
445
-
446
- <!-- info: What aspect of model ability can be measured with this dataset? -->
447
- <!-- scope: telescope -->
448
- Surface realization and compositionally.
449
-
450
- #### Metrics
451
-
452
- <!-- info: What metrics are typically used for this task? -->
453
- <!-- scope: periscope -->
454
- `BLEURT`, `BLEU`, `ROUGE`
455
-
456
- #### Proposed Evaluation
457
-
458
- <!-- info: List and describe the purpose of the metrics and evaluation methodology (including human evaluation) that the dataset creators used when introducing this task. -->
459
- <!-- scope: microscope -->
460
- The original paper focused on the task of dialog state prediction instead of response generation and thus did not suggest any set of metrics.
461
-
462
- #### Previous results available?
463
-
464
- <!-- info: Are previous results available? -->
465
- <!-- scope: telescope -->
466
- no
467
-
468
-
469
-
470
- ## Dataset Curation
471
-
472
- ### Original Curation
473
-
474
- #### Original Curation Rationale
475
-
476
- <!-- info: Original curation rationale -->
477
- <!-- scope: telescope -->
478
- Previous multi-domain task-oriented dialogue datsets do not sufficiently capture the real-world challenges in virtual assistants, since they cover few domains and assume a single static ontology per domain.
479
- The SGD datset is created to cover 17 domains with over 16K dialogues, and contain multiple different APIs in most domains, many of which have overlapping functionalities but different interfaces, which reflects common real-world scenarios.
480
- The wide range of available annotations can be used for intent prediction, slot filling, dialogue state tracking, policy imitation learning, language generation, user simulation learning, among other tasks in large-scale virtual assistants.
481
-
482
-
483
- #### Communicative Goal
484
-
485
- <!-- info: What was the communicative goal? -->
486
- <!-- scope: periscope -->
487
- The goal of a speaker who generates the target utterance is to help users accomplish tasks including but not limited to finding flights, booking restaurants, searching for nearby events and movies.
488
-
489
-
490
- #### Sourced from Different Sources
491
-
492
- <!-- info: Is the dataset aggregated from different data sources? -->
493
- <!-- scope: telescope -->
494
- no
495
-
496
-
497
- ### Language Data
498
-
499
- #### How was Language Data Obtained?
500
-
501
- <!-- info: How was the language data obtained? -->
502
- <!-- scope: telescope -->
503
- `Machine-generated`
504
-
505
- #### Generation Method Link
506
-
507
- <!-- info: If text was machine-generated for the dataset, provide a link to the generation method if available (N/A otherwise). -->
508
- <!-- scope: periscope -->
509
- [Github](https://github.com/google-research-datasets/dstc8-schema-guided-dialogue)
510
-
511
- #### Language Producers
512
-
513
- <!-- info: What further information do we have on the language producers? -->
514
- <!-- scope: microscope -->
515
- The dialogue outlines are first generated by a simulator. The dialogue simulator interacts with the services to generate dialogue outlines. It consists of two agents playing the roles of the user and the system, interacting with each other using a finite set of actions specified through dialogue acts over a probabilistic automaton designed to capture varied dialogue trajectories. It is worth noting that the simulation automaton does not include any domain-specific constraints: all domain-specific constraints are encoded in the schema and scenario.
516
-
517
- The dialogue paraphrasing framework then converts the outlines generated by the simulator into a natural conversation. Users may refer to the slot values in the dialogue acts in various different ways during the conversation, e.g., “los angeles” may be referred to as “LA” or “LAX”. To introduce these natural variations in the slot values, different slot values are replaced with a randomly selected variation while being kept consistent across user turns in a dialogue. The actions are then converted to pseudo-natural language utterances using a set of manually defined action-to-text templates, and the resulting utterances for the different actions in a turn are concatenated together.
518
-
519
-
520
- #### Topics Covered
521
-
522
- <!-- info: Does the language in the dataset focus on specific topics? How would you describe them? -->
523
- <!-- scope: periscope -->
524
- The dataset covers the following domains: Alarm, Banks, Buses, Calendar, Events, Flights, Homes, Hotels, Media, Messaging, Movies, Music, Payment, RentalCars, Restaurants, RideSharing, Services, Train, Travel, and Weather. The domain ‘Service’ includes salons, dentists, doctors etc. The ‘Alarm’, ‘Messaging’, ‘Payment’ and ‘Train’ domains are only present in the dev or test sets.
525
- to test generalization to new domains.
526
-
527
- #### Data Validation
528
-
529
- <!-- info: Was the text validated by a different worker or a data curator? -->
530
- <!-- scope: telescope -->
531
- not validated
532
-
533
- #### Was Data Filtered?
534
-
535
- <!-- info: Were text instances selected or filtered? -->
536
- <!-- scope: telescope -->
537
- not filtered
538
-
539
-
540
- ### Structured Annotations
541
-
542
- #### Additional Annotations?
543
-
544
- <!-- quick -->
545
- <!-- info: Does the dataset have additional annotations for each instance? -->
546
- <!-- scope: telescope -->
547
- crowd-sourced
548
-
549
- #### Number of Raters
550
-
551
- <!-- info: What is the number of raters -->
552
- <!-- scope: telescope -->
553
- unknown
554
-
555
- #### Raters per Training Example
556
-
557
- <!-- info: How many annotators saw each training example? -->
558
- <!-- scope: periscope -->
559
- 0
560
-
561
- #### Raters per Test Example
562
-
563
- <!-- info: How many annotators saw each test example? -->
564
- <!-- scope: periscope -->
565
- 0
566
-
567
- #### Annotation Service?
568
-
569
- <!-- info: Was an annotation service used? -->
570
- <!-- scope: telescope -->
571
- unknown
572
-
573
- #### Annotation Values
574
-
575
- <!-- info: Purpose and values for each annotation -->
576
- <!-- scope: microscope -->
577
- The dialogue transformed by these steps is sent to the crowd workers to be reformulated into more natural language. One crowd worker is tasked with paraphrasing all utterances of a dialogue to ensure naturalness and coherence. The crowd workers are asked to exactly repeat the slot values in their paraphrases so that the span indices for the slots can be recovered via string matching.
578
-
579
-
580
- #### Any Quality Control?
581
-
582
- <!-- info: Quality control measures? -->
583
- <!-- scope: telescope -->
584
- none
585
-
586
-
587
- ### Consent
588
-
589
- #### Any Consent Policy?
590
-
591
- <!-- info: Was there a consent policy involved when gathering the data? -->
592
- <!-- scope: telescope -->
593
- no
594
-
595
- #### Justification for Using the Data
596
-
597
- <!-- info: If not, what is the justification for reusing the data? -->
598
- <!-- scope: microscope -->
599
- While no policy is reported, we assume that one was in place for the collection.
600
-
601
-
602
- ### Private Identifying Information (PII)
603
-
604
- #### Contains PII?
605
-
606
- <!-- quick -->
607
- <!-- info: Does the source language data likely contain Personal Identifying Information about the data creators or subjects? -->
608
- <!-- scope: telescope -->
609
- no PII
610
-
611
- #### Justification for no PII
612
-
613
- <!-- info: Provide a justification for selecting `no PII` above. -->
614
- <!-- scope: periscope -->
615
- The SGD dataset does not use identity categories and does not contain sensitive data.
616
-
617
-
618
-
619
- ### Maintenance
620
-
621
- #### Any Maintenance Plan?
622
-
623
- <!-- info: Does the original dataset have a maintenance plan? -->
624
- <!-- scope: telescope -->
625
- no
626
-
627
-
628
-
629
- ## Broader Social Context
630
-
631
- ### Previous Work on the Social Impact of the Dataset
632
-
633
- #### Usage of Models based on the Data
634
-
635
- <!-- info: Are you aware of cases where models trained on the task featured in this dataset ore related tasks have been used in automated systems? -->
636
- <!-- scope: telescope -->
637
- no
638
-
639
-
640
- ### Impact on Under-Served Communities
641
-
642
- #### Addresses needs of underserved Communities?
643
-
644
- <!-- info: Does this dataset address the needs of communities that are traditionally underserved in language technology, and particularly language generation technology? Communities may be underserved for exemple because their language, language variety, or social or geographical context is underepresented in NLP and NLG resources (datasets and models). -->
645
- <!-- scope: telescope -->
646
- no
647
-
648
-
649
- ### Discussion of Biases
650
-
651
- #### Any Documented Social Biases?
652
-
653
- <!-- info: Are there documented social biases in the dataset? Biases in this context are variations in the ways members of different social categories are represented that can have harmful downstream consequences for members of the more disadvantaged group. -->
654
- <!-- scope: telescope -->
655
- no
656
-
657
- #### Are the Language Producers Representative of the Language?
658
-
659
- <!-- info: Does the distribution of language producers in the dataset accurately represent the full distribution of speakers of the language world-wide? If not, how does it differ? -->
660
- <!-- scope: periscope -->
661
- Due to the combination of the automatic generation and crowd rater paraphasing, the language can be very formulaic. While this may be acceptable for the model part (i.e., we may actually desire an automated agent to form formulaic responses), the input utterances of the simulated customers likely do not cover the entire spectrum of the English language.
662
-
663
-
664
-
665
- ## Considerations for Using the Data
666
-
667
- ### PII Risks and Liability
668
-
669
-
670
-
671
- ### Licenses
672
-
673
- #### Copyright Restrictions on the Dataset
674
-
675
- <!-- info: Based on your answers in the Intended Use part of the Data Overview Section, which of the following best describe the copyright and licensing status of the dataset? -->
676
- <!-- scope: periscope -->
677
- `open license - commercial use allowed`
678
-
679
- #### Copyright Restrictions on the Language Data
680
-
681
- <!-- info: Based on your answers in the Language part of the Data Curation Section, which of the following best describe the copyright and licensing status of the underlying language data? -->
682
- <!-- scope: periscope -->
683
- `open license - commercial use allowed`
684
-
685
-
686
- ### Known Technical Limitations
687
-
688
- #### Technical Limitations
689
-
690
- <!-- info: Describe any known technical limitations, such as spurrious correlations, train/test overlap, annotation biases, or mis-annotations, and cite the works that first identified these limitations when possible. -->
691
- <!-- scope: microscope -->
692
- The dialogues under each domain distributed unevenly, where the flights domain has 3644 dialogues while the payment domain only contains 222 dialogues.
693
- Besides, all dialogues are paraphrased by crowd-workers, and it is possible that crow-workers with different culture backgrounds will exhibit biased opinions.
694
-
695
-
696
- #### Unsuited Applications
697
-
698
- <!-- info: When using a model trained on this dataset in a setting where users or the public may interact with its predictions, what are some pitfalls to look out for? In particular, describe some applications of the general task featured in this dataset that its curation or properties make it less suitable for. -->
699
- <!-- scope: microscope -->
700
- Since the initial data was automatically generated, the coverage of entity names is necessarily biased. An agent thus needs to be evaluated in a more realistic environment.
701
-
702
-
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
dataset_infos.json DELETED
@@ -1 +0,0 @@
1
- {"schema_guided_dialog": {"description": "GEM is a benchmark environment for Natural Language Generation with a focus on its Evaluation,\nboth through human annotations and automated Metrics.\n\nGEM aims to:\n- measure NLG progress across 13 datasets spanning many NLG tasks and languages.\n- provide an in-depth analysis of data and models presented via data statements and challenge sets.\n- develop standards for evaluation of generated text using both automated and human metrics.\n\nIt is our goal to regularly update GEM and to encourage toward more inclusive practices in dataset development\nby extending existing data or developing datasets for additional languages.\n", "citation": "@article{gem_benchmark,\n author = {Sebastian Gehrmann and\n Tosin P. Adewumi and\n Karmanya Aggarwal and\n Pawan Sasanka Ammanamanchi and\n Aremu Anuoluwapo and\n Antoine Bosselut and\n Khyathi Raghavi Chandu and\n Miruna{-}Adriana Clinciu and\n Dipanjan Das and\n Kaustubh D. Dhole and\n Wanyu Du and\n Esin Durmus and\n Ondrej Dusek and\n Chris Emezue and\n Varun Gangal and\n Cristina Garbacea and\n Tatsunori Hashimoto and\n Yufang Hou and\n Yacine Jernite and\n Harsh Jhamtani and\n Yangfeng Ji and\n Shailza Jolly and\n Dhruv Kumar and\n Faisal Ladhak and\n Aman Madaan and\n Mounica Maddela and\n Khyati Mahajan and\n Saad Mahamood and\n Bodhisattwa Prasad Majumder and\n Pedro Henrique Martins and\n Angelina McMillan{-}Major and\n Simon Mille and\n Emiel van Miltenburg and\n Moin Nadeem and\n Shashi Narayan and\n Vitaly Nikolaev and\n Rubungo Andre Niyongabo and\n Salomey Osei and\n Ankur P. Parikh and\n Laura Perez{-}Beltrachini and\n Niranjan Ramesh Rao and\n Vikas Raunak and\n Juan Diego Rodriguez and\n Sashank Santhanam and\n Joao Sedoc and\n Thibault Sellam and\n Samira Shaikh and\n Anastasia Shimorina and\n Marco Antonio Sobrevilla Cabezudo and\n Hendrik Strobelt and\n Nishant Subramani and\n Wei Xu and\n Diyi Yang and\n Akhila Yerukola and\n Jiawei Zhou},\n title = {The {GEM} Benchmark: Natural Language Generation, its Evaluation and\n Metrics},\n journal = {CoRR},\n volume = {abs/2102.01672},\n year = {2021},\n url = {https://arxiv.org/abs/2102.01672},\n archivePrefix = {arXiv},\n eprint = {2102.01672}\n}\n", "homepage": "https://gem-benchmark.github.io/", "license": "CC-BY-SA-4.0", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "gem_parent_id": {"dtype": "string", "id": null, "_type": "Value"}, "dialog_acts": [{"act": {"num_classes": 18, "names": ["AFFIRM", "AFFIRM_INTENT", "CONFIRM", "GOODBYE", "INFORM", "INFORM_COUNT", "INFORM_INTENT", "NEGATE", "NEGATE_INTENT", "NOTIFY_FAILURE", "NOTIFY_SUCCESS", "OFFER", "OFFER_INTENT", "REQUEST", "REQUEST_ALTS", "REQ_MORE", "SELECT", "THANK_YOU"], "id": null, "_type": "ClassLabel"}, "slot": {"dtype": "string", "id": null, "_type": "Value"}, "values": [{"dtype": "string", "id": null, "_type": "Value"}]}], "context": [{"dtype": "string", "id": null, "_type": "Value"}], "dialog_id": {"dtype": "string", "id": null, "_type": "Value"}, "service": {"dtype": "string", "id": null, "_type": "Value"}, "turn_id": {"dtype": "int32", "id": null, "_type": "Value"}, "prompt": {"dtype": "string", "id": null, "_type": "Value"}, "target": {"dtype": "string", "id": null, "_type": "Value"}, "references": [{"dtype": "string", "id": null, "_type": "Value"}]}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "gem", "config_name": "schema_guided_dialog", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 146648117, "num_examples": 164982, "dataset_name": "gem"}, "validation": {"name": "validation", "num_bytes": 9376504, "num_examples": 10000, "dataset_name": "gem"}, "test": {"name": "test", "num_bytes": 10160596, "num_examples": 10000, "dataset_name": "gem"}, "challenge_train_sample": {"name": "challenge_train_sample", "num_bytes": 441326, "num_examples": 500, "dataset_name": "gem"}, "challenge_validation_sample": {"name": "challenge_validation_sample", "num_bytes": 491492, "num_examples": 500, "dataset_name": "gem"}, "challenge_test_backtranslation": {"name": "challenge_test_backtranslation", "num_bytes": 512834, "num_examples": 500, "dataset_name": "gem"}, "challenge_test_bfp02": {"name": "challenge_test_bfp02", "num_bytes": 529404, "num_examples": 500, "dataset_name": "gem"}, "challenge_test_bfp05": {"name": "challenge_test_bfp05", "num_bytes": 515151, "num_examples": 500, "dataset_name": "gem"}, "challenge_test_nopunc": {"name": "challenge_test_nopunc", "num_bytes": 509332, "num_examples": 500, "dataset_name": "gem"}, "challenge_test_scramble": {"name": "challenge_test_scramble", "num_bytes": 514644, "num_examples": 500, "dataset_name": "gem"}}, "download_checksums": {"https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_sgd_context.zip": {"num_bytes": 16544230, "checksum": "abb2af00031152dbead4a75275dc195a576005529cc19b7f942669f5d257ef30"}, "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/schema_guided_dialog.zip": {"num_bytes": 1282238, "checksum": "79231851df998a9dc2a1298f8061cf7e9e9ad0b1ea34f7e5124eb31960a4b842"}}, "download_size": 17826468, "post_processing_size": null, "dataset_size": 169699400, "size_in_bytes": 187525868}, "default": {"description": "The Schema-Guided Dialogue (SGD) dataset contains 18K multi-domain task-oriented\ndialogues between a human and a virtual assistant, which covers 17 domains\nranging from banks and events to media, calendar, travel, and weather. The\nlanguage presents in the datset is only English. The SGD dataset provides a\nchallenging testbed for a number of tasks in task-oriented dialogue, including\nlanguage understanding, slot filling, dialogue state tracking and response\ngeneration. For the creation of the SGD dataset, they developed a multi-domain\ndialogue simulator that generates dialogue outlines over an arbitrary combination\nof APIs, dialogue states and system actions. Then, they used a crowd-sourcing\nprocedure to paraphrase these outlines to natural language utterances. This novel\ncrowd-sourcing procedure preserves all annotations obtained from the simulator and\ndoes not require any extra annotations after dialogue collection.\n\n", "citation": "@inproceedings{rastogi2020towards,\n title={Towards scalable multi-domain conversational agents: The schema-guided dialogue dataset},\n author={Rastogi, Abhinav and Zang, Xiaoxue and Sunkara, Srinivas and Gupta, Raghav and Khaitan, Pranav},\n booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},\n volume={34},\n number={05},\n pages={8689--8696},\n year={2020}\n}\n", "homepage": "", "license": "", "features": {"gem_id": {"dtype": "string", "id": null, "_type": "Value"}, "gem_parent_id": {"dtype": "string", "id": null, "_type": "Value"}, "dialog_acts": [{"act": {"num_classes": 18, "names": ["AFFIRM", "AFFIRM_INTENT", "CONFIRM", "GOODBYE", "INFORM", "INFORM_COUNT", "INFORM_INTENT", "NEGATE", "NEGATE_INTENT", "NOTIFY_FAILURE", "NOTIFY_SUCCESS", "OFFER", "OFFER_INTENT", "REQUEST", "REQUEST_ALTS", "REQ_MORE", "SELECT", "THANK_YOU"], "id": null, "_type": "ClassLabel"}, "slot": {"dtype": "string", "id": null, "_type": "Value"}, "values": [{"dtype": "string", "id": null, "_type": "Value"}]}], "context": [{"dtype": "string", "id": null, "_type": "Value"}], "dialog_id": {"dtype": "string", "id": null, "_type": "Value"}, "service": {"dtype": "string", "id": null, "_type": "Value"}, "turn_id": {"dtype": "int32", "id": null, "_type": "Value"}, "prompt": {"dtype": "string", "id": null, "_type": "Value"}, "target": {"dtype": "string", "id": null, "_type": "Value"}, "references": [{"dtype": "string", "id": null, "_type": "Value"}], "linearized_input": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "schema_guided_dialog", "config_name": "default", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 189823851, "num_examples": 164982, "dataset_name": "schema_guided_dialog"}, "validation": {"name": "validation", "num_bytes": 11342250, "num_examples": 10000, "dataset_name": "schema_guided_dialog"}, "test": {"name": "test", "num_bytes": 12125104, "num_examples": 10000, "dataset_name": "schema_guided_dialog"}, "challenge_train_sample": {"name": "challenge_train_sample", "num_bytes": 540672, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_validation_sample": {"name": "challenge_validation_sample", "num_bytes": 591290, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_test_backtranslation": {"name": "challenge_test_backtranslation", "num_bytes": 609480, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_test_bfp02": {"name": "challenge_test_bfp02", "num_bytes": 630597, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_test_bfp05": {"name": "challenge_test_bfp05", "num_bytes": 611891, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_test_nopunc": {"name": "challenge_test_nopunc", "num_bytes": 613410, "num_examples": 500, "dataset_name": "schema_guided_dialog"}, "challenge_test_scramble": {"name": "challenge_test_scramble", "num_bytes": 615233, "num_examples": 500, "dataset_name": "schema_guided_dialog"}}, "download_checksums": {"https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_sgd_context.zip": {"num_bytes": 16544230, "checksum": "abb2af00031152dbead4a75275dc195a576005529cc19b7f942669f5d257ef30"}, "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/schema_guided_dialog.zip": {"num_bytes": 1282238, "checksum": "79231851df998a9dc2a1298f8061cf7e9e9ad0b1ea34f7e5124eb31960a4b842"}}, "download_size": 17826468, "post_processing_size": null, "dataset_size": 217503778, "size_in_bytes": 235330246}}
 
 
default/schema_guided_dialog-challenge_test_backtranslation.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ca7a48328d37272ce437e47349170d29927213141d1a5571f309929a9d2ee9b
3
+ size 280568
default/schema_guided_dialog-challenge_test_bfp02.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63cd75e05f4069cf20b18461d89ec1627d231ab0c18d053d75bc5c8aba43886a
3
+ size 291654
default/schema_guided_dialog-challenge_test_bfp05.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a411822466cef14ced4316165d01517a68b6c6db22960676c48b916be4b78221
3
+ size 291054
default/schema_guided_dialog-challenge_test_nopunc.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37144c3f376d5d974d559a6967cb79b20ef8e25494ceb514148bc1662307ffd4
3
+ size 282578
default/schema_guided_dialog-challenge_test_scramble.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d8e6297cb32df26da948356dd62e150eaf71b39c4e1807e4396d9e66329dc79
3
+ size 288852
default/schema_guided_dialog-challenge_train_sample.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34c40baf8b0cc7fb0768ec7779c6c3aec365d7fe0570c26ddd744d41f736210e
3
+ size 246487
default/schema_guided_dialog-challenge_validation_sample.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4cb00ace54c2b8361c7bb0c8bf15e8a6b6d99cffc5420370ea97da303329571
3
+ size 260331
default/schema_guided_dialog-test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86452e89ae684aa898243c45e81989ec9c171c738ccc14dca294c3751eb461c7
3
+ size 5342805
default/schema_guided_dialog-train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f361a2809e1dd7bc197d2eb4877470ffeb9dfa53bc1bfa0ce2618f5362075887
3
+ size 34921910
default/schema_guided_dialog-validation.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b66274fa8b56d80709e4d6684246ff7df99e4215037049c86227a226918d4af
3
+ size 4701583
schema_guided_dialog.json DELETED
@@ -1,180 +0,0 @@
1
- {
2
- "overview": {
3
- "where": {
4
- "has-leaderboard": "no",
5
- "leaderboard-url": "N/A",
6
- "leaderboard-description": "N/A",
7
- "website": "n/a",
8
- "data-url": "[Github[(https://github.com/google-research-datasets/dstc8-schema-guided-dialogue)",
9
- "paper-url": "[Arxiv](https://arxiv.org/abs/1909.05855)",
10
- "paper-bibtext": "```\n{\n@inproceedings{rastogi2020towards,\n title={Towards scalable multi-domain conversational agents: The schema-guided dialogue dataset},\n author={Rastogi, Abhinav and Zang, Xiaoxue and Sunkara, Srinivas and Gupta, Raghav and Khaitan, Pranav},\n booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},\n volume={34},\n number={05},\n pages={8689--8696},\n year={2020}\n}\n```",
11
- "contact-email": "schema-guided-dst@google.com",
12
- "contact-name": "Abhinav Rastogi"
13
- },
14
- "languages": {
15
- "is-multilingual": "no",
16
- "license": "cc-by-sa-4.0: Creative Commons Attribution Share Alike 4.0 International",
17
- "task-other": "N/A",
18
- "language-names": [
19
- "English"
20
- ],
21
- "communicative": "The goal of a speaker who generates the target utterance is to help users accomplish tasks including but not limited to finding flights, booking restaurants, searching for nearby events and movies.\n",
22
- "language-speakers": "The language structure is machine-generated, and the language realizations are produced by crowd workers.\nThe dataset paper does not provide demographic information for the crowd workers.\n",
23
- "language-dialects": "n/a",
24
- "intended-use": "The Schema-Guided Dialogue (SGD) dataset contains 18K multi-domain task-oriented dialogues between a human and a virtual assistant, which covers 17 domains ranging from banks and events to media, calendar, travel, and weather.\nThe language presents in the datset is only English.\nThe SGD dataset provides a challenging testbed for a number of tasks in task-oriented dialogue, including language understanding, slot filling, dialogue state tracking and response generation.\nFor the creation of the SGD dataset, they developed a multi-domain dialogue simulator that generates dialogue outlines over an arbitrary combination of APIs, dialogue states and system actions. Then, they used a crowd-sourcing procedure to paraphrase these outlines to natural language utterances.\nThis novel crowd-sourcing procedure preserves all annotations obtained from the simulator and does not require any extra annotations after dialogue collection.\n",
25
- "license-other": "N/A",
26
- "task": "Dialog Response Generation"
27
- },
28
- "credit": {
29
- "organization-type": [
30
- "industry"
31
- ],
32
- "organization-names": "Google",
33
- "creators": "Abhinav Rastogi, Xiaoxue Zang, Srinivas Sunkara, Raghav Gupta, Pranav Khaitan, Amir Fayazi, Maria Wang, and Guan-Lin Chao",
34
- "funding": "Google",
35
- "gem-added-by": "Wanyu Du wrote the initial data card and Yacine Jernite the data loader. Simon Mille updated the data card with the additional splits. Sebastian Gehrmann migrated the data card and loader to the v2 version and extended the missing information."
36
- },
37
- "structure": {
38
- "data-fields": "Each dialog instance has the following fields:\n\n* `dialogue_id`: A unique identifier for a dialogue.\n* `services`: A list of services present in the dialogue.\n* `turns`: A list of annotated system or user utterances. Each turn consists of the following fields:\n\t* `speaker`: The speaker for the turn, either `USER` or `SYSTEM`.\n\t* `utterance`: A string containing the natural language utterance.\n\t* `frames`: A list of frames, each frame containing annotations for a single service and consists of the following fields:\n\t\t* `service`: The name of the service corresponding to the frame. The slots and intents used in the following fields are taken from the schema of this service.\n\t\t* `slots`: A list of slot spans in the utterance, only provided for non-categorical slots. Each slot span contains the following fields:\n\t\t\t* `slot`: The name of the slot.\n\t\t\t* `start`: The index of the starting character in the utterance corresponding to the slot value.\n\t\t\t* `exclusive_end`: The index of the character just after the last character corresponding to the slot value in the utterance.\n\t\t* `actions`: A list of actions corresponding to the system. Each action has the following fields:\n\t\t\t* `act`: The type of action.\n\t\t\t* `slot`: (optional) A slot argument for some of the actions.\n\t\t\t* `values`: (optional) A list of values assigned to the slot. If the values list is non-empty, then the slot must be present.\n\t\t\t* `canonical_values`: (optional) The values in their canonicalized form as used by the service. It is a list of strings of the same length as values.\n\t\t* `service_call`: (system turns only, optional) The request sent to the service. It consists of the following fields:\n\t\t\t* `method`: The name of the intent or function of the service or API being executed.\n\t\t\t* `parameters`: A pair of lists of the same lengths: `parameter_slot_name` contains slot names and `parameter_canonical_value` contains the corresponding values in their canonicalized form.\n\t\t* `service_results`: (system turns only, optional) A list of entities containing the results obtained from the service. It is only available for turns in which a service call is made. Each entity is represented as a pair of lists of the same length: `service_slot_name` contains slot names and `service_canonical_value` contains the corresponding canonical values.\n\t\t* `state`: (user turns only) The dialogue state corresponding to the service. It consists of the following fields:\n\t\t\t* `active_intent`: The intent corresponding to the service of the frame which is currently being fulfilled by the system. It takes the value \"NONE\" if none of the intents are active.\n\t\t\t* `requested_slots`: A list of slots requested by the user in the current turn.\n\t\t\t* `slot_values`: A pair of lists of the same lengths: `slot_name` contains slot names and `slot_value_list` contains the corresponding lists of strings. For categorical slots, this list contains a single value assigned to the slot. For non-categorical slots, all the values in this list are spoken variations of each other and are equivalent (e.g, \"6 pm\", \"six in the evening\", \"evening at 6\" etc.).\n\n",
39
- "structure-description": "n/a",
40
- "structure-labels": "n/a",
41
- "structure-example": "```\n{'dialogue_id': '1_00000',\n 'services': ['Restaurants_1'],\n 'turns':\n {'frames':\n \t[{'actions': [{'act': [6],\n 'canonical_values': [['FindRestaurants']],\n 'slot': ['intent'],\n 'values': [['FindRestaurants']]}],\n 'service': ['Restaurants_1'],\n 'service_call': [{'method': '',\n 'parameters': {'parameter_canonical_value': [],\n 'parameter_slot_name': []}}],\n 'service_results': [{'service_results_list': []}],\n 'slots': [{'exclusive_end': [], 'slot': [], 'start': []}],\n 'state': [{'active_intent': 'FindRestaurants',\n \t\t\t 'requested_slots': [],\n \t\t\t 'slot_values': {'slot_name': [], 'slot_value_list': []}}]},\n {'actions': [{'act': [13],\n 'canonical_values': [[]],\n 'slot': ['city'],\n 'values': [[]]}],\n 'service': ['Restaurants_1'],\n 'service_call': [{'method': '',\n 'parameters': {'parameter_canonical_value': [],\n 'parameter_slot_name': []}}],\n 'service_results': [{'service_results_list': []}],\n 'slots': [{'exclusive_end': [], 'slot': [], 'start': []}],\n 'state': [{'active_intent': '',\n \t\t 'requested_slots': [],\n \t\t 'slot_values': {'slot_name': [], 'slot_value_list': []}}]},\n ...,]}\n 'speaker': [0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1],\n 'utterance': [\n 'I am feeling hungry so I would like to find a place to eat.',\n 'Do you have a specific which you want the eating place to be located at?',\n 'I would like for it to be in San Jose.',\n 'Is there a specific cuisine type you enjoy, such as Mexican, Italian or something else?',\n 'I usually like eating the American type of food.',\n 'I see that at 71 Saint Peter there is a good restaurant which is in San Jose.',\n 'Can you give me the address of this restaurant.',\n 'If you want to go to this restaurant you can find it at 71 North San Pedro Street.',\n 'Can you give me the phone number that I can contact them with?',\n 'If you want to phone them you can at 408-971-8523.',\n 'Is there some other restaurant which you can suggest?',\n 'How would you like Bazille restaurant which is situated in San Jose.',\n 'Do you have another restaurant matching my needs? For example a restaurant which is economical and is located in Palo Alto.',\n 'I see that 7 restaurants suit to what you requested. Bird Dog seems as a good restaurant and is located in Palo Alto.',\n 'Alright, that seems good. I would like to make a booking at this restaurant.',\n 'For which time do you want the booking to be?',\n 'I will be eating there at 11:30 am so make it for then.',\n 'Can you please confirm that you want to book a table for 2 at 11:30 am at the Bird Dog restaurant in Palo Alto for today.',\n 'That suits me well. Can you tell me if they feature live music?',\n 'Your booking has been made without errors, but unfortunately they do not have live music.',\n 'Will I be able to find liquor there? Can you give me the address of their location?',\n 'The restaurant is located at 420 Ramona Street. Unfortunately they do not serve alcohol at the restaurant.',\n 'I appreciate it very much. That would be all.',\n 'Have a good time!'\n ]}\n```",
42
- "structure-splits": "The dataset is split into a train, validation, and test set with the following sizes:\n\n| | Train | Validation | Test |\n| --- | --- | --- | --- |\n| \\# of dialogues | 16142 | 2482 | 4201 |\n| \\# of turns | 48426 | 7446 | 12603 |",
43
- "structure-splits-criteria": "The data is generally split i.i.d, but some topics only appear in training and some only for testing. For example, the domains Messaging, Payment, and Train are test-only. ",
44
- "structure-outlier": "n/a"
45
- },
46
- "what": {
47
- "dataset": "The GEM version of this dataset functions as a response generation dataset. The input specifies dialog acts that a model needs to verbalize. The Schema-Guided Dialog dataset is challenging since it comprises multiple domains from hotel and travel to restaurants, and a wide range of dialog acts. The context of each conversation is provided as well. "
48
- }
49
- },
50
- "curation": {
51
- "original": {
52
- "is-aggregated": "no",
53
- "aggregated-sources": "N/A",
54
- "rationale": "Previous multi-domain task-oriented dialogue datsets do not sufficiently capture the real-world challenges in virtual assistants, since they cover few domains and assume a single static ontology per domain.\nThe SGD datset is created to cover 17 domains with over 16K dialogues, and contain multiple different APIs in most domains, many of which have overlapping functionalities but different interfaces, which reflects common real-world scenarios.\nThe wide range of available annotations can be used for intent prediction, slot filling, dialogue state tracking, policy imitation learning, language generation, user simulation learning, among other tasks in large-scale virtual assistants.\n",
55
- "communicative": "The goal of a speaker who generates the target utterance is to help users accomplish tasks including but not limited to finding flights, booking restaurants, searching for nearby events and movies.\n"
56
- },
57
- "language": {
58
- "found": [],
59
- "crowdsourced": [],
60
- "created": "N/A",
61
- "machine-generated": "[Github](https://github.com/google-research-datasets/dstc8-schema-guided-dialogue)",
62
- "validated": "not validated",
63
- "is-filtered": "not filtered",
64
- "filtered-criteria": "N/A",
65
- "obtained": [
66
- "Machine-generated"
67
- ],
68
- "producers-description": "The dialogue outlines are first generated by a simulator. The dialogue simulator interacts with the services to generate dialogue outlines. It consists of two agents playing the roles of the user and the system, interacting with each other using a finite set of actions specified through dialogue acts over a probabilistic automaton designed to capture varied dialogue trajectories. It is worth noting that the simulation automaton does not include any domain-specific constraints: all domain-specific constraints are encoded in the schema and scenario.\n\nThe dialogue paraphrasing framework then converts the outlines generated by the simulator into a natural conversation. Users may refer to the slot values in the dialogue acts in various different ways during the conversation, e.g., \u201clos angeles\u201d may be referred to as \u201cLA\u201d or \u201cLAX\u201d. To introduce these natural variations in the slot values, different slot values are replaced with a randomly selected variation while being kept consistent across user turns in a dialogue. The actions are then converted to pseudo-natural language utterances using a set of manually defined action-to-text templates, and the resulting utterances for the different actions in a turn are concatenated together.\n",
69
- "topics": "The dataset covers the following domains: Alarm, Banks, Buses, Calendar, Events, Flights, Homes, Hotels, Media, Messaging, Movies, Music, Payment, RentalCars, Restaurants, RideSharing, Services, Train, Travel, and Weather. The domain \u2018Service\u2019 includes salons, dentists, doctors etc. The \u2018Alarm\u2019, \u2018Messaging\u2019, \u2018Payment\u2019 and \u2018Train\u2019 domains are only present in the dev or test sets.\nto test generalization to new domains."
70
- },
71
- "annotations": {
72
- "origin": "crowd-sourced",
73
- "rater-number": "unknown",
74
- "rater-qualifications": "n/a",
75
- "rater-training-num": "0",
76
- "rater-test-num": "0",
77
- "rater-annotation-service-bool": "unknown",
78
- "rater-annotation-service": [],
79
- "values": "The dialogue transformed by these steps is sent to the crowd workers to be reformulated into more natural language. One crowd worker is tasked with paraphrasing all utterances of a dialogue to ensure naturalness and coherence. The crowd workers are asked to exactly repeat the slot values in their paraphrases so that the span indices for the slots can be recovered via string matching.\n",
80
- "quality-control": "none",
81
- "quality-control-details": "N/A"
82
- },
83
- "consent": {
84
- "has-consent": "no",
85
- "consent-policy": "N/A",
86
- "consent-other": "N/A",
87
- "no-consent-justification": "While no policy is reported, we assume that one was in place for the collection."
88
- },
89
- "pii": {
90
- "has-pii": "no PII",
91
- "no-pii-justification": "The SGD dataset does not use identity categories and does not contain sensitive data.\n",
92
- "is-pii-identified": "N/A",
93
- "pii-identified-method": "N/A",
94
- "is-pii-replaced": "N/A",
95
- "pii-replaced-method": "N/A",
96
- "pii-categories": []
97
- },
98
- "maintenance": {
99
- "has-maintenance": "no",
100
- "description": "N/A",
101
- "contact": "N/A",
102
- "contestation-mechanism": "N/A",
103
- "contestation-link": "N/A",
104
- "contestation-description": "N/A"
105
- }
106
- },
107
- "gem": {
108
- "rationale": {
109
- "sole-task-dataset": "yes",
110
- "sole-language-task-dataset": "no",
111
- "distinction-description": "The domains a lot more diverse than other datasets. ",
112
- "contribution": "This dataset comprises a wide range of dialog capabilities and thus enables the evaluation of many more generation capabilities of comparable datasets. Its collection methodology ensures a high diversity but also high quality of the data. ",
113
- "model-ability": "surface realization, compositionality."
114
- },
115
- "curation": {
116
- "has-additional-curation": "yes",
117
- "modification-types": [
118
- "data points modified"
119
- ],
120
- "modification-description": "We are focusing on the response-generation part of the dataset and thus reformatted the dataset to treat the service agent utterances as the targets to be generated and the previous customer utterance and the agent's dialog act as the input. We additionally reformat the dialog acts to directly conform to the format described in this [paper](https://arxiv.org/abs/2004.15006).\n",
121
- "has-additional-splits": "yes",
122
- "additional-splits-description": "9 challenge sets for Schema-Guided Dialog were added to the GEM evaluation suite.\n\n1. We created subsets of the training and development sets of 500 randomly selected inputs each.\n2. We applied 5 transformations to respectively 5 sets of 500 randomly selected inputs: (i) back-translation, (ii)-(iii) introduction of typographical errors, using Butterfingers with two thresholds (0.02 and 0.05), resulting in two sets with different amounts of typos introduced (there are more typos with the 0.05 threshold than with the 0.02 one), (iv) removal of final punctuations (when any), and (v) input scrambling, for which the order of the dialogue acts was randomly reassigned.\n3. For the input size, we created subpopulations based on the number of dialogue acts in the input.\n\n| DA number | Frequency English |\n|---------------|-------------------|\n| 1 | 5049 |\n| 2 | 2517 |\n| 3 | 1328 |\n| 4 | 469 |\n| 5 | 335 |\n| 6 | 256 |\n| 7 | 46 |\n\nWe also split the test data according to the type of dialogue act, represented by cardinal numbers in the dataset.\n\n| DA type | Frequency English |\n|--------------|-------------------|\n| 2 | 1397 |\n| 3 | 983 |\n| 4 | 1027 |\n| 5 | 958 |\n| 9 | 72 |\n| 10 | 1024 |\n| 11 | 1246 |\n| 12 | 500 |\n| 13 | 2078 |\n| 15 | 715 |\n",
123
- "additional-splits-capacicites": "Generalization and Robustness."
124
- },
125
- "starting": {
126
- "research-pointers": "* [Paper for dataset and DST baseline](https://arxiv.org/pdf/1909.05855.pdf)\n* [DSTC8 overview paper](https://arxiv.org/pdf/2002.01359.pdf)\n* [Code for DST baseline](https://github.com/google-research/google-research/tree/master/schema_guided_dst)\n* [Natural language generation baseline paper](https://arxiv.org/pdf/2004.15006.pdf)\n* [Blog post announcing the dataset](https://ai.googleblog.com/2019/10/introducing-schema-guided-dialogue.html)\n",
127
- "technical-terms": "n/a"
128
- }
129
- },
130
- "results": {
131
- "results": {
132
- "other-metrics-definitions": "N/A",
133
- "has-previous-results": "no",
134
- "current-evaluation": "N/A",
135
- "previous-results": "N/A",
136
- "model-abilities": "Surface realization and compositionally. ",
137
- "metrics": [
138
- "BLEURT",
139
- "BLEU",
140
- "ROUGE"
141
- ],
142
- "original-evaluation": "The original paper focused on the task of dialog state prediction instead of response generation and thus did not suggest any set of metrics. "
143
- }
144
- },
145
- "considerations": {
146
- "pii": {
147
- "risks-description": "n/a"
148
- },
149
- "licenses": {
150
- "dataset-restrictions-other": "N/A",
151
- "data-copyright-other": "N/A",
152
- "dataset-restrictions": [
153
- "open license - commercial use allowed"
154
- ],
155
- "data-copyright": [
156
- "open license - commercial use allowed"
157
- ]
158
- },
159
- "limitations": {
160
- "data-technical-limitations": "The dialogues under each domain distributed unevenly, where the flights domain has 3644 dialogues while the payment domain only contains 222 dialogues.\nBesides, all dialogues are paraphrased by crowd-workers, and it is possible that crow-workers with different culture backgrounds will exhibit biased opinions.\n",
161
- "data-unsuited-applications": "Since the initial data was automatically generated, the coverage of entity names is necessarily biased. An agent thus needs to be evaluated in a more realistic environment. "
162
- }
163
- },
164
- "context": {
165
- "previous": {
166
- "is-deployed": "no",
167
- "described-risks": "N/A",
168
- "changes-from-observation": "N/A"
169
- },
170
- "underserved": {
171
- "helps-underserved": "no",
172
- "underserved-description": "N/A"
173
- },
174
- "biases": {
175
- "has-biases": "no",
176
- "bias-analyses": "N/A",
177
- "speaker-distibution": "Due to the combination of the automatic generation and crowd rater paraphasing, the language can be very formulaic. While this may be acceptable for the model part (i.e., we may actually desire an automated agent to form formulaic responses), the input utterances of the simulated customers likely do not cover the entire spectrum of the English language. "
178
- }
179
- }
180
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
schema_guided_dialog.py DELETED
@@ -1,209 +0,0 @@
1
- import json
2
- import os
3
- import datasets
4
-
5
- _CITATION = """\
6
- @inproceedings{rastogi2020towards,
7
- title={Towards scalable multi-domain conversational agents: The schema-guided dialogue dataset},
8
- author={Rastogi, Abhinav and Zang, Xiaoxue and Sunkara, Srinivas and Gupta, Raghav and Khaitan, Pranav},
9
- booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
10
- volume={34},
11
- number={05},
12
- pages={8689--8696},
13
- year={2020}
14
- }
15
- """
16
-
17
- _DESCRIPTION = """\
18
- The Schema-Guided Dialogue (SGD) dataset contains 18K multi-domain task-oriented
19
- dialogues between a human and a virtual assistant, which covers 17 domains
20
- ranging from banks and events to media, calendar, travel, and weather. The
21
- language presents in the datset is only English. The SGD dataset provides a
22
- challenging testbed for a number of tasks in task-oriented dialogue, including
23
- language understanding, slot filling, dialogue state tracking and response
24
- generation. For the creation of the SGD dataset, they developed a multi-domain
25
- dialogue simulator that generates dialogue outlines over an arbitrary combination
26
- of APIs, dialogue states and system actions. Then, they used a crowd-sourcing
27
- procedure to paraphrase these outlines to natural language utterances. This novel
28
- crowd-sourcing procedure preserves all annotations obtained from the simulator and
29
- does not require any extra annotations after dialogue collection.
30
-
31
- """
32
-
33
- _URLs = {
34
- "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_sgd_context.zip",
35
- "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/schema_guided_dialog.zip",
36
- }
37
-
38
- _SGD_ACTS = [
39
- "AFFIRM",
40
- "AFFIRM_INTENT",
41
- "CONFIRM",
42
- "GOODBYE",
43
- "INFORM",
44
- "INFORM_COUNT",
45
- "INFORM_INTENT",
46
- "NEGATE",
47
- "NEGATE_INTENT",
48
- "NOTIFY_FAILURE",
49
- "NOTIFY_SUCCESS",
50
- "OFFER",
51
- "OFFER_INTENT",
52
- "REQUEST",
53
- "REQUEST_ALTS",
54
- "REQ_MORE",
55
- "SELECT",
56
- "THANK_YOU",
57
- ]
58
-
59
-
60
- def process_sgd(example):
61
- prompt = example["prompt"]
62
- inp = f'Prompt: "{prompt}", '
63
- for da in example["dialog_acts"]:
64
- act = _SGD_ACTS[da["act"]].lower()
65
- slot = da["slot"]
66
- values = " or ".join(da["values"])
67
- inp += f"Response Type: {act}"
68
- if slot:
69
- inp += f", Type of Slot: {slot}"
70
- if values:
71
- inp += f", Values: {values}"
72
- inp += ", "
73
- inp += f'Agent: {example["service"]}'
74
-
75
- return inp
76
-
77
-
78
- class SchemaGuidedDialog(datasets.GeneratorBasedBuilder):
79
- VERSION = datasets.Version("1.0.0")
80
- DEFAULT_CONFIG_NAME = "schema_guided_dialog"
81
-
82
- def _info(self):
83
- features = datasets.Features(
84
- {
85
- "gem_id": datasets.Value("string"),
86
- "gem_parent_id": datasets.Value("string"),
87
- "dialog_acts": [
88
- {
89
- "act": datasets.ClassLabel(names=_SGD_ACTS),
90
- "slot": datasets.Value("string"),
91
- "values": [datasets.Value("string")],
92
- }
93
- ],
94
- "context": [datasets.Value("string")],
95
- "dialog_id": datasets.Value("string"),
96
- "service": datasets.Value("string"),
97
- "turn_id": datasets.Value("int32"),
98
- "prompt": datasets.Value("string"),
99
- "target": datasets.Value("string"),
100
- "references": [datasets.Value("string")],
101
- "linearized_input": datasets.Value("string"),
102
- }
103
- )
104
- return datasets.DatasetInfo(
105
- description=_DESCRIPTION,
106
- features=features,
107
- supervised_keys=None,
108
- homepage="",
109
- citation=_CITATION,
110
- )
111
-
112
- def _split_generators(self, dl_manager):
113
- """Returns SplitGenerators."""
114
- dl_dir = dl_manager.download_and_extract(_URLs)
115
- challenge_sets = [
116
- (
117
- "challenge_train_sample",
118
- "train_schema_guided_dialog_RandomSample500_reformatted.json",
119
- ),
120
- (
121
- "challenge_validation_sample",
122
- "validation_schema_guided_dialog_RandomSample500_reformatted.json",
123
- ),
124
- (
125
- "challenge_test_backtranslation",
126
- "test_schema_guided_dialog_BackTranslation500_reformatted.json",
127
- ),
128
- (
129
- "challenge_test_bfp02",
130
- "test_schema_guided_dialog_ButterFingersPerturbation_p=0.02_500_reformatted.json",
131
- ),
132
- (
133
- "challenge_test_bfp05",
134
- "test_schema_guided_dialog_ButterFingersPerturbation_p=0.05_500_reformatted.json",
135
- ),
136
- (
137
- "challenge_test_nopunc",
138
- "test_schema_guided_dialog_WithoutPunctuation500_reformatted.json",
139
- ),
140
- (
141
- "challenge_test_scramble",
142
- "test_schema_guided_dialog_ScrambleInputStructure500_reformatted.json",
143
- ),
144
- ]
145
- return [
146
- datasets.SplitGenerator(
147
- name=spl,
148
- gen_kwargs={
149
- "filepath": os.path.join(dl_dir["data"], "gem_sgd.json"),
150
- "split": spl,
151
- },
152
- )
153
- for spl in ["train", "validation", "test"]
154
- ] + [
155
- datasets.SplitGenerator(
156
- name=challenge_split,
157
- gen_kwargs={
158
- "filepath": os.path.join(
159
- dl_dir["challenge_set"], "schema_guided_dialog", filename
160
- ),
161
- "split": challenge_split,
162
- },
163
- )
164
- for challenge_split, filename in challenge_sets
165
- ]
166
-
167
- def _generate_examples(self, filepath, split, filepaths=None, lang=None):
168
- """Yields examples."""
169
- if "challenge" in split:
170
- exples = json.load(open(filepath, encoding="utf-8"))
171
- if isinstance(exples, dict):
172
- assert len(exples) == 1, "multiple entries found"
173
- exples = list(exples.values())[0]
174
- for id_, exple in enumerate(exples):
175
- if len(exple) == 0:
176
- continue
177
- exple["gem_parent_id"] = exple["gem_id"]
178
- exple["gem_id"] = f"schema_guided_dialog-{split}-{id_}"
179
- exple["linearized_input"] = process_sgd(exple)
180
- yield id_, exple
181
- else:
182
- examples = json.load(open(filepath, encoding="utf-8"))[split]
183
- for id_, example in enumerate(examples):
184
- # Fix the one example that has an empty target.
185
- if not example["target"]:
186
- example["target"] = "Thank you, goodbye."
187
-
188
- exple = {
189
- "gem_id": f"schema_guided_dialog-{split}-{id_}",
190
- "gem_parent_id": f"schema_guided_dialog-{split}-{id_}",
191
- "dialog_acts": [
192
- {
193
- "act": act_id,
194
- "slot": slot,
195
- "values": values,
196
- }
197
- for act_id, slot, values in example["da"]
198
- ],
199
- "context": example["context"],
200
- "dialog_id": example["dialog_id"],
201
- "service": example["service"],
202
- "turn_id": example["turn_ix"],
203
- "prompt": example["prompt"],
204
- "target": example["target"],
205
- "references": [example["target"]],
206
- }
207
- exple["linearized_input"] = process_sgd(exple)
208
-
209
- yield id_, exple