ingerid commited on
Commit
2ebb4dd
1 Parent(s): d6900c6

docs: update dataset card

Browse files
Files changed (1) hide show
  1. README.md +28 -18
README.md CHANGED
@@ -58,34 +58,40 @@ Northern Norway | Nord-Norge | Nordland, Troms og Finnmark
58
 
59
  ### Data Instances
60
 
61
- A data point is an audio segment, including a relative path to the `.wav`-file, and the transcription. Additional information is provided about the speaker, the orthographic standard for the transcriptuion, whether the segment overlaps with the previous or next, and the setting for the recording.
62
 
63
  ```
64
- {'source_file_id': 'nb-1',
65
- 'segment_id': '0008970-0013860',
66
- 'segment_order': 0,
67
- 'duration': 4.89,
68
- 'overlap_previous': False,
69
- 'overlap_next': False,
70
- 'speaker_id': 'P36',
71
- 'gender': 1,
72
- 'dialect': 0,
73
- 'orthography': 0,
74
- 'source_type': 0,
75
- 'file_name': 'nb-1_0008970-0013860.wav',
76
- 'transcription': 'hallo og velkommen hit til Nasjonalbiblioteket.',
77
- 'audio': {
78
- 'path': 'data/train/bm/nb-1_0008970-0013860.wav',
79
- 'array': array([-0.00033569, 0.00222778, -0.0005188 , ..., 0.00057983, 0.0005188 ]),
 
 
 
 
 
80
  'sampling_rate': 16000}
81
  }
 
82
  ```
83
 
84
  ### Data Fields
85
 
86
  data field | description | Value type / example
87
  --- | --- | ---
88
- `source_file_id` | original file the segment appears in. | (str) e.g. `50f-X`, `tr-X` or `nb-X`, where X is a number.
89
  `segment_id` | segment start and end timestamp. | `{starttime}-{endtime}` (str)
90
  `segment_order` | order of segment in the original file. | (int)
91
  `duration` | duration of segment in seconds. | (float)
@@ -98,7 +104,11 @@ data field | description | Value type / example
98
  `source_type` | type of recording of original file, either `live-event` or `podcast`, as a ClassLabel index number | `0`: live-event or `1`: podcast (int)
99
  `file_name` | file name of the audio segment, without the path | `{source_file_id}_{segment_id}.wav` (str)
100
  `transcription` | orthographic transcription text | (str)
 
 
 
101
  `audio` | the audio segment data, with the relative file `path`, the bytes `array`, and the `sampling_rate` | (dict)
 
102
 
103
  ### Data Splits
104
 
 
58
 
59
  ### Data Instances
60
 
61
+ A data point is an audio segment, including a relative path to the `.wav`-file, and the transcription. Additional information is provided about the speaker, the orthographic standard for the transcription, whether the segment overlaps with the previous or next, and the setting for the recording. The transcription also comes in 3 different normalized versions: "orthographic" (orthographically correct text, with punctuation, integer numbers, and standardized word forms), "verbatim" (with tokens marking hesitations, laughter, foreign phrases and unknown words, but no punctuation) and "annotations" (as is from the annotation process, with punctuation, tags, and alternate word forms).
62
 
63
  ```
64
+ {
65
+ 'source_file_id': 'nb-1',
66
+ 'segment_id': '0008970-0013860',
67
+ 'segment_order': 0,
68
+ 'duration': 4.89,
69
+ 'overlap_previous': False,
70
+ 'overlap_next': False,
71
+ 'speaker_id': 'P36',
72
+ 'gender': 1,
73
+ 'dialect': 0,
74
+ 'orthography': 0,
75
+ 'source_type': 0,
76
+ 'file_name': 'data/train/bm/nb-1_0008970-0013860.wav',
77
+ 'transcription': 'hallo og velkommen hit til Nasjonalbiblioteket.',
78
+ 'annotations': 'hallo og velkommen hit til Nasjonalbiblioteket.',
79
+ 'orthographic': 'hallo og velkommen hit til Nasjonalbiblioteket.',
80
+ 'verbatim': 'hallo og velkommen hit til Nasjonalbiblioteket',
81
+ 'audio': {
82
+ 'path': "data/train/bm/nb-1_0008970-0013860.wav",
83
+ 'array': array([-0.00033569, 0.00222778, -0.0005188 , ..., 0.00067139,
84
+ 0.00057983, 0.0005188 ]),
85
  'sampling_rate': 16000}
86
  }
87
+
88
  ```
89
 
90
  ### Data Fields
91
 
92
  data field | description | Value type / example
93
  --- | --- | ---
94
+ `source_file_id` | original file the segment appears in. | e.g. `50f-X`, `tr-X` or `nb-X`, where X is a number. (str)
95
  `segment_id` | segment start and end timestamp. | `{starttime}-{endtime}` (str)
96
  `segment_order` | order of segment in the original file. | (int)
97
  `duration` | duration of segment in seconds. | (float)
 
104
  `source_type` | type of recording of original file, either `live-event` or `podcast`, as a ClassLabel index number | `0`: live-event or `1`: podcast (int)
105
  `file_name` | file name of the audio segment, without the path | `{source_file_id}_{segment_id}.wav` (str)
106
  `transcription` | orthographic transcription text | (str)
107
+ `orthographic` | close to orthographically correct text transcription in the given `orthography` standard. Contains punctuation, numbers, and standard word forms. | (str)
108
+ `verbatim` | transcription text mapping to the uttered words as close as possible. Contains tokens marking hesitations, laughter, foreign phrases and unknown words, but no punctuation. | (str)
109
+ `annotations` | transcription text "as is" from the annotation process. Contains false starts, metatags for non-linguistic noises, punctuation, and alternate word forms (`<uttered word>\<orthographic standard word>`) | (str)
110
  `audio` | the audio segment data, with the relative file `path`, the bytes `array`, and the `sampling_rate` | (dict)
111
+ "orthographic" (orthographically correct text, with punctuation, integer numbers, and standardized word forms), "verbatim" (with tokens marking hesitations, laughter, foreign phrases and unknown words, but no punctuation) and "annotations" (as is from the annotation process, with punctuation, tags, and alternate word forms).
112
 
113
  ### Data Splits
114