cartesinus commited on
Commit
1cfbdd1
1 Parent(s): e8581a6

add custom loading script

Browse files
Files changed (2) hide show
  1. README.md +18 -0
  2. leyzer-fedcsis.py +229 -0
README.md ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - text-classification
5
+ language:
6
+ - en
7
+ - pl
8
+ - es
9
+ tags:
10
+ - natural-language-understanding
11
+ size_categories:
12
+ - 10K<n<100K
13
+ ---
14
+
15
+ # Leyzer: A Dataset for Multilingual Virtual Assistants
16
+ Leyzer is a multilingual text corpus designed to study multilingual and cross-lingual natural language understanding (NLU) models and the strategies of localization of
17
+ virtual assistants. It consists of 20 domains across three languages: English, Spanish and Polish, with 186 intents and a wide range of samples, ranging from 1 to 672
18
+ sentences per intent. For more stats please refer to wiki.
leyzer-fedcsis.py ADDED
@@ -0,0 +1,229 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+
3
+ """Leyzer"""
4
+
5
+ import json
6
+ import datasets
7
+ import pandas as pd
8
+
9
+
10
+ _DESCRIPTION = """\
11
+ Leyzer is a multilingual text corpus designed to study multilingual and cross-lingual natural language
12
+ understanding (NLU) models and the strategies of localization of virtual assistants. It consists of 20
13
+ domains across three languages: English, Spanish and Polish, with 186 intents and a wide range of
14
+ samples, ranging from 1 to 672 sentences per intent.
15
+ """
16
+
17
+ _CITATION = """\
18
+ @inproceedings{sowanski2020leyzer,
19
+ title={Leyzer: A Dataset for Multilingual Virtual Assistants},
20
+ author={Sowa{\'n}ski, Marcin and Janicki, Artur},
21
+ booktitle={International Conference on Text, Speech, and Dialogue},
22
+ pages={477--486},
23
+ year={2020},
24
+ organization={Springer}
25
+ }
26
+ """
27
+
28
+ _LICENSE = """\
29
+
30
+ """
31
+
32
+ _URL = "https://github.com/cartesinus/leyzer/raw/0.2.0/corpora/0.2.0/leyzer-0.2.0-fedcsis.tar.gz"
33
+
34
+ _LANGUAGES = ['en-US', 'es-ES', 'pl-PL']
35
+
36
+ _DOMAINS = ['Airconditioner', 'Calendar', 'Console', 'Contacts', 'Email', 'Facebook', 'Fitbit',
37
+ 'Gdrive', 'Instagram', 'News', 'Phone', 'Slack', 'Speaker', 'Spotify', 'Translate',
38
+ 'Twitter', 'Weather', 'Websearch', 'Wikipedia', 'Yelp', 'Youtube']
39
+
40
+ _INTENTS = ["AddAlbumToPlaylist", "AddEventOnDateWithName", "AddEventWithName",
41
+ "AddSongToPlaylist", "AddSongToPlaylistWithName", "AddSongWithNameToPlaylistWithName",
42
+ "AddWeight", "CallContact", "CallEmergency", "CallNumber", "ChangeLanguageToLanguage",
43
+ "ChangeTemperature", "CheckCalendarEventName", "CheckCalendarOnDate", "CheckChannelHistory",
44
+ "CheckLastMessages", "CheckMessagesFromUser", "CheckMessagesInChannel",
45
+ "CheckUserStatus", "ConsoleCD", "ConsoleCP", "ConsoleEdit", "ConsoleLS",
46
+ "ConsoleMV", "ConsoleRM", "CreateContactWithNameAndWithNumberAndWithType",
47
+ "CreateFile", "CreateFileWithName", "CreatePlaylist", "CreatePlaylistWithName",
48
+ "DecreaseVolume", "DecreaseVolumeByPercent", "DetectLanguage", "DownloadAsPdf",
49
+ "EditContactWithName", "EditContactWithNumber", "EditNumberInContactWithName",
50
+ "EditNumberInContactWithNameToNumberAndToType", "FindChannelWithQuery",
51
+ "FindQuery", "FindQueryOnChannel", "FollowUser", "GetHumidity", "GetTemperature",
52
+ "GetTemperatureFromDevice", "GoToElementNumber", "IncreaseVolume",
53
+ "IncreaseVolumeByPercent", "MoonphaseInLocation", "MoonphaseInLocationOnDate",
54
+ "MuteOff", "MuteOn", "NextSong", "NextVideo", "NotifyOnEventInLocation",
55
+ "NotifyOnEventStart", "NotifyOnNewFromFollowing", "NotifyOnWeight", "NotifyWhenEventNameStart",
56
+ "NotifyWhenPortalUpdates", "NotifyWhenPortalUpdatesInSection", "NotNotifyOnEventInLocation",
57
+ "NotNotifyOnEventStart", "OpenCalendar", "OpenContacts", "OpenEmail", "OpenEngine",
58
+ "OpenFacebook", "OpenFileWithName", "OpenGdrive", "OpenInstagram", "OpenMyContact",
59
+ "OpenRestaurants", "OpenRestaurantsInLocation", "OpenSlack", "OpenSpotify",
60
+ "OpenTwitter", "OpenWeather", "OpenWikipedia", "OpenYT", "Pause",
61
+ "PlayAlbumOfTypeByArtist", "PlayPlaylist", "PlaySong", "PlaySongByArtist",
62
+ "PostPicture", "PostPictureWithCaption", "PostPictureWithUrl",
63
+ "PostPictureWithUrlAndWithCaption", "PostStatus", "PostWithLastPictureWithLocationAndWithHashtag",
64
+ "PostWithPicturesFromDateAndWithLocation", "PreviousSong", "RandomPage",
65
+ "RepeatOnOff", "ReplyToEmailFromAddress", "ScrollDown", "ScrollUp",
66
+ "SearchByCategory", "SearchByCategoryInLocation", "SearchByQuery",
67
+ "SearchByQueryInLocation", "SearchByRating", "SearchByRatingInLocation",
68
+ "SearchByReviewCount", "SearchByReviewCountInLocation", "SearchImagesOnEngine",
69
+ "SearchImagesWithTextOnEngine", "SearchImagesWithTextOnEngineWithWidthAndWithHeight",
70
+ "SearchImagesWithTextOnEngineWithWidthAndWithHeightAndWithCondition", "SearchInLocation", "SearchQuery",
71
+ "SearchRestaurant", "SearchText", "SearchTextOnEngine", "SeekByTime", "SendEmail",
72
+ "SendEmailToAddress", "SendEmailToAddressWithSubject", "SendEmailToAddressWithSubjectAndWithMessage",
73
+ "SendMessageToChannel", "SendMessageToUser", "SendPictureToChannel", "SendPictureToChannelWithCaption",
74
+ "SendPictureWithUrlToChannel", "SendSlackMessage", "SendTwitterMessage", "SetDefaultLanguage",
75
+ "SetDefaultLanguageToLanguage", "SetPurposeOnChannel", "SetStatusAway", "SetTemperatureBetweenValues",
76
+ "SetTemperatureOnDevice", "SetTemperatureToValue", "SetTemperatureToValueOnDevice", "SetTopicInChannel",
77
+ "ShareFileWithNameToAddress", "ShowAlbumWithName", "ShowBMI", "ShowChannelWithCategory",
78
+ "ShowChannelWithName", "ShowContactWithEmail", "ShowContactWithName", "ShowContactWithNumber",
79
+ "ShowContactWithNumberAndWithEmail", "ShowDanceability", "ShowEmailFromSender", "ShowEmailFromTime",
80
+ "ShowEmailWithLabel", "ShowEmailWithPriority", "ShowEmailWithSubject", "ShowFilesFromTime",
81
+ "ShowFilesWithSize", "ShowFilesWithStar", "ShowFilesWithType", "ShowLastNumberPictures",
82
+ "ShowNewestFiles", "ShowNews", "ShowNewsFromSection", "ShowOldestFiles", "ShowPictures",
83
+ "ShowPicturesFromDate", "ShowPicturesWithFilter", "ShowPicturesWithHashtag",
84
+ "ShowPicturesWithLocation", "ShowSMS", "ShowSteps", "ShowStepsOnDate", "ShowSubscribedChannels",
85
+ "ShowTwittsWithHashtag", "ShuffleOnOff", "SMSToContact", "SMSToContactWithMessage",
86
+ "SortAlphabetical", "SortReverseAlphabetical", "SunriseAndSunsetInLocation",
87
+ "SunriseAndSunsetInLocationAndOnDate", "SunriseInLocation", "SunsetInLocation",
88
+ "TakePicture", "TranslateText", "TranslateTextFromLanguage", "TranslateTextFromLanguageToLanguage",
89
+ "TranslateTextFromLanguageToLanguageWithEngine", "TranslateTextToLanguage",
90
+ "TranslateToLanguage", "TurnOff", "TurnOn", "TurnOnOrOff", "UnfollowUser",
91
+ "WeatherInLocation", "WeatherTomorrow", "WeatherTomorrowInLocation",
92
+ "NotifyOnNewOnChannel"
93
+ ]
94
+
95
+ _LEVELS = ['L0TC', 'L1TC', 'L2TC', 'REPHRASE']
96
+
97
+ _VPS = ['verb_pattern_01', 'verb_pattern_02', 'verb_pattern_03', 'verb_pattern_04',
98
+ 'verb_pattern_05', 'verb_pattern_06', 'verb_pattern_07', 'verb_pattern_08',
99
+ 'verb_pattern_09', 'verb_pattern_10', 'verb_pattern_11', 'verb_pattern_12', 'verb_pattern_13']
100
+
101
+
102
+ class LeyzerConfig(datasets.BuilderConfig):
103
+ """BuilderConfig for Leyzer."""
104
+
105
+ def __init__(self, dataset_version=None, *args, **kwargs):
106
+ """BuilderConfig for Leyzer.
107
+ Args:
108
+ **kwargs: keyword arguments forwarded to super.
109
+ """
110
+ super(LeyzerConfig, self).__init__(*args, **kwargs)
111
+ self.dataset_version = dataset_version if dataset_version else "1.0"
112
+ self.data_url = _URL
113
+
114
+
115
+ class Leyzer(datasets.GeneratorBasedBuilder):
116
+ """Leyzer"""
117
+
118
+ # All individual locale datasets are served from the latest version.
119
+ BUILDER_CONFIGS = [
120
+ LeyzerConfig(
121
+ name = name,
122
+ dataset_version = '0.2.0',
123
+ version = datasets.Version("0.2.0"),
124
+ description = f"The Leyzer v0.2.0 corpora for {name}",
125
+ ) for name in _LANGUAGES
126
+ ]
127
+ # Version 1.0
128
+ BUILDER_CONFIGS.append(LeyzerConfig(
129
+ name = "all",
130
+ dataset_version = '0.2.0',
131
+ version = datasets.Version("0.2.0"),
132
+ description = f"The Leyzer v0.2.0 corpora for entire corpus",
133
+ )
134
+ )
135
+
136
+ DEFAULT_CONFIG_NAME = "all"
137
+
138
+ def _info(self):
139
+ return datasets.DatasetInfo(
140
+ description=_DESCRIPTION,
141
+ features=datasets.Features(
142
+ {
143
+ "id": datasets.Value("string"),
144
+ "locale": datasets.Value("string"),
145
+ "partition": datasets.Value("string"),
146
+ "domain": datasets.features.ClassLabel(names=_DOMAINS),
147
+ "intent": datasets.features.ClassLabel(names=_INTENTS),
148
+ "utterance": datasets.Value("string"),
149
+ "bio": datasets.Value("string"),
150
+ },
151
+ ),
152
+ supervised_keys=None,
153
+ homepage="https://github.com/cartesinus/leyzer/",
154
+ citation=_CITATION,
155
+ license=_LICENSE,
156
+ )
157
+
158
+ def _split_generators(self, dl_manager):
159
+
160
+ archive = dl_manager.download(self.config.data_url)
161
+ return [
162
+ datasets.SplitGenerator(
163
+ name=datasets.Split.TRAIN,
164
+ gen_kwargs={
165
+ "files": dl_manager.iter_archive(archive),
166
+ "split": "train",
167
+ "lang": self.config.name,
168
+ }
169
+ ),
170
+ datasets.SplitGenerator(
171
+ name=datasets.Split.VALIDATION,
172
+ gen_kwargs={
173
+ "files": dl_manager.iter_archive(archive),
174
+ "split": "valid",
175
+ "lang": self.config.name,
176
+ }
177
+ ),
178
+ datasets.SplitGenerator(
179
+ name=datasets.Split.TEST,
180
+ gen_kwargs={
181
+ "files": dl_manager.iter_archive(archive),
182
+ "split": "test",
183
+ "lang": self.config.name,
184
+ }
185
+ ),
186
+ ]
187
+
188
+ def _generate_examples(self, files, split, lang):
189
+
190
+ key_ = 0
191
+
192
+ if lang == "all":
193
+ lang = _LANGUAGES.copy()
194
+ else:
195
+ lang = [lang]
196
+
197
+ for path, f in files:
198
+ l = path.split("/")[-1].split(".")[0]
199
+
200
+ if not lang:
201
+ break
202
+ elif l in lang:
203
+ lang.remove(l)
204
+ else:
205
+ continue
206
+
207
+ # Read the file
208
+ lines = f.read().decode(encoding="utf-8").split("\n")
209
+
210
+ for line in lines:
211
+ if not line:
212
+ continue
213
+
214
+ data = json.loads(line)
215
+
216
+ if data["partition"] != split:
217
+ continue
218
+
219
+ yield key_, {
220
+ "id": data["id"],
221
+ "locale": data["locale"],
222
+ "partition": data["partition"],
223
+ "domain": data["domain"],
224
+ "intent": data["intent"],
225
+ "utterance": data["utterance"],
226
+ "bio": data["bio"],
227
+ }
228
+
229
+ key_ += 1