Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    CastError
Message:      Couldn't cast
lang: string
type: string
bigram: string
count: int64
pmi: double
npmi: double
score: double
variants: int64
to
{'lang': Value('string'), 'type': Value('string'), 'bigram': Value('string'), 'count': Value('int64'), 'pmi': Value('float64'), 'npmi': Value('float64'), 'score': Value('float64')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 265, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 120, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              lang: string
              type: string
              bigram: string
              count: int64
              pmi: double
              npmi: double
              score: double
              variants: int64
              to
              {'lang': Value('string'), 'type': Value('string'), 'bigram': Value('string'), 'count': Value('int64'), 'pmi': Value('float64'), 'npmi': Value('float64'), 'score': Value('float64')}
              because column names don't match
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1347, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 980, in convert_to_parquet
                  builder.download_and_prepare(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 884, in download_and_prepare
                  self._download_and_prepare(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 947, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1739, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1922, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

lang
string
type
string
bigram
string
count
int64
pmi
float64
npmi
float64
score
float64
en
VERB+ADP
come on
408,628
5.51
0.582
10.85
en
VERB+ADP
talk about
115,459
6.58
0.583
9.81
en
VERB+ADP
sit down
43,156
7.9
0.622
9.58
en
VERB+ADP
shut up
51,375
7.42
0.596
9.32
en
VERB+ADP
look at
145,762
5.89
0.538
9.23
en
VERB+ADP
wake up
30,205
7.62
0.576
8.58
en
VERB+ADP
calm down
14,157
8.18
0.571
7.88
en
VERB+ADP
pick up
28,096
6.72
0.504
7.45
en
VERB+ADP
worry about
29,781
6.6
0.499
7.41
en
VERB+ADP
hurry up
21,796
6.6
0.482
6.95
en
VERB+ADP
find out
51,391
5.32
0.427
6.68
en
VERB+ADP
grow up
15,345
6.54
0.461
6.41
en
VERB+ADP
slow down
6,480
7.49
0.485
6.15
en
VERB+ADP
deal with
21,026
5.79
0.422
6.06
en
VERB+ADP
wait for
54,130
4.72
0.381
6
en
VERB+ADP
depend on
10,180
6.66
0.45
6
en
VERB+ADP
talk to
124,621
3.81
0.341
5.77
en
VERB+ADP
figure out
12,342
6.15
0.424
5.76
en
VERB+ADP
piss off
4,248
7.56
0.471
5.68
en
VERB+ADP
hold on
32,618
4.91
0.375
5.62
en
VERB+ADP
get out
107,801
3.74
0.329
5.5
en
VERB+ADP
stand by
10,442
6.08
0.412
5.5
en
VERB+ADP
look like
57,433
4.26
0.347
5.48
en
VERB+ADP
sound like
15,561
5.58
0.394
5.48
en
VERB+ADP
think about
59,654
4.21
0.345
5.47
en
VERB+ADP
surround by
2,620
7.9
0.472
5.36
en
VERB+ADP
bump into
1,817
8.55
0.495
5.36
en
VERB+ADP
end up
15,779
5.36
0.379
5.28
en
VERB+ADP
rid of
13,590
5.47
0.381
5.23
en
VERB+ADP
settle down
3,950
7.07
0.437
5.23
en
VERB+ADP
listen to
66,283
3.94
0.326
5.22
en
VERB+ADP
care about
16,866
5.13
0.365
5.12
en
VERB+ADP
hang on
15,253
5.23
0.368
5.11
en
VERB+ADP
accord to
13,018
5.31
0.368
5.03
en
VERB+ADP
cut off
6,378
6.14
0.397
5.02
en
VERB+ADP
look for
69,884
3.73
0.311
5
en
VERB+ADP
forget about
14,137
5.18
0.362
4.99
en
VERB+ADP
come along
10,775
5.42
0.368
4.94
en
VERB+ADP
set up
10,987
5.36
0.366
4.91
en
VERB+ADP
turn into
7,126
5.85
0.383
4.9
en
VERB+ADP
go on
140,982
3.11
0.283
4.84
en
VERB+ADP
testify against
771
9.35
0.505
4.84
en
VERB+ADP
pay for
19,546
4.62
0.334
4.76
en
VERB+ADP
rely on
2,774
6.87
0.413
4.72
en
VERB+ADP
run into
7,137
5.63
0.368
4.71
en
VERB+ADP
screw up
4,952
6.06
0.383
4.7
en
VERB+ADP
feel like
20,455
4.46
0.324
4.64
en
VERB+ADP
belong to
16,101
4.69
0.332
4.64
en
VERB+ADP
wipe out
3,155
6.53
0.397
4.61
en
VERB+ADP
fall into
4,549
6.02
0.377
4.58
en
VERB+ADP
live in
29,369
4.05
0.306
4.54
en
VERB+ADP
stand up
11,243
4.92
0.336
4.53
en
VERB+ADP
tag along
587
9.31
0.493
4.53
en
VERB+ADP
bicker amongst
95
14.8
0.687
4.52
en
VERB+ADP
show up
14,324
4.61
0.323
4.46
en
VERB+ADP
hang out
7,830
5.22
0.344
4.45
en
VERB+ADP
mix up
3,053
6.36
0.385
4.45
en
VERB+ADP
watch out
12,129
4.74
0.326
4.43
en
VERB+ADP
sleep with
12,696
4.7
0.325
4.42
en
VERB+ADP
turn around
5,085
5.66
0.359
4.42
en
VERB+ADP
stumble onto
218
11.55
0.568
4.41
en
VERB+ADP
pass through
2,771
6.38
0.383
4.38
en
VERB+ADP
open up
11,193
4.76
0.325
4.37
en
VERB+ADP
interfere with
2,478
6.51
0.387
4.36
en
VERB+ADP
clean up
6,135
5.36
0.346
4.35
en
VERB+ADP
turn out
12,660
4.61
0.318
4.34
en
VERB+ADP
hide behind
1,414
7.31
0.414
4.34
en
VERB+ADP
revolve around
415
9.69
0.5
4.34
en
VERB+ADP
laugh at
6,900
5.2
0.339
4.32
en
VERB+ADP
blow up
6,931
5.18
0.338
4.31
en
VERB+ADP
suffer from
3,015
6.16
0.372
4.3
en
VERB+ADP
base on
5,974
5.33
0.343
4.3
en
VERB+ADP
doze off
468
9.33
0.485
4.3
en
VERB+ADP
come from
36,225
3.65
0.282
4.27
en
VERB+ADP
hold onto
1,114
7.54
0.42
4.25
en
VERB+ADP
lighten up
1,125
7.52
0.419
4.24
en
VERB+ADP
agree with
6,280
5.17
0.334
4.22
en
VERB+ADP
check out
8,955
4.79
0.32
4.2
en
VERB+ADP
fill with
5,043
5.37
0.34
4.18
en
VERB+ADP
™ ª
113
13.05
0.613
4.18
en
VERB+ADP
obsess with
1,526
6.87
0.392
4.15
en
VERB+ADP
caption by
567
8.6
0.454
4.15
en
VERB+ADP
rise above
488
8.86
0.462
4.13
en
VERB+ADP
yell at
3,025
5.89
0.356
4.12
en
VERB+ADP
stay with
16,609
4.13
0.293
4.11
en
VERB+ADP
count on
7,459
4.83
0.317
4.08
en
VERB+ADP
stare at
1,262
7.04
0.395
4.07
en
VERB+ADP
go through
19,836
3.93
0.284
4.06
en
VERB+ADP
take off
15,882
4.11
0.291
4.06
en
VERB+ADP
walk into
3,724
5.55
0.342
4.05
en
VERB+ADP
split up
2,751
5.89
0.353
4.03
en
VERB+ADP
fool around
2,077
6.22
0.364
4.01
en
VERB+ADP
fall in
14,529
4.12
0.289
4
en
VERB+ADP
act like
6,031
4.95
0.318
4
en
VERB+ADP
break up
9,278
4.52
0.303
3.99
en
VERB+ADP
abide by
602
8.1
0.429
3.96
en
VERB+ADP
transform into
566
8.2
0.433
3.96
en
VERB+ADP
lie down
4,689
5.16
0.324
3.95
en
VERB+ADP
refer to
4,739
5.15
0.324
3.95
en
VERB+ADP
hook up
2,530
5.85
0.349
3.94
End of preview.

OpenSubtitles Collocations

NPMI-scored bigram collocations extracted from the OpenSubtitles parallel corpus. Three languages, three relation types, ~43K bigrams total.

Languages & Corpus Size

Language Code Corpus lines Bigrams
English en ~100M 15,000
Dutch nl ~105M 15,000
Serbian sr ~50M 13,586

Relation Types

  • ADJ+NOUN — adjective-noun pairs: "slim contract", "kreditan kartica"
  • VERB+ADP — phrasal verbs / verb-preposition: "come on", "houden van"
  • VERB+NOUN — verb-object pairs: "earn money", "verdienen geld"

Fields

Field Type Description
lang str Language code (en/nl/sr)
type str Relation type (ADJ+NOUN, VERB+ADP, VERB+NOUN)
bigram str Lemmatized bigram
count int Co-occurrence count in corpus
pmi float Pointwise mutual information
npmi float Normalized PMI (0–1 scale)
score float Composite ranking score (PMI + log frequency)
variants int Number of surface form variants (Serbian only)

Usage

from datasets import load_dataset

# Load one language
ds = load_dataset("vladvlasov256/opensubs-collocations", "en", split="train")

# Load all languages
ds = load_dataset("vladvlasov256/opensubs-collocations", "all", split="train")

# Filter
ds.filter(lambda x: x["type"] == "VERB+ADP" and x["npmi"] > 0.5)

Extraction Method

  • Source: OPUS OpenSubtitles v2018 (Lison & Tiedemann, 2016)
  • NLP: Stanza (tokenize, POS, lemma, depparse)
  • Patterns: ADJ+NOUN, VERB+NOUN, VERB+ADP extracted via dependency relations
  • Filtering: MIN_COUNT=3, TOP_N=5000 per pattern per language
  • Scoring: PMI and NPMI, ranked by composite score (PMI + log frequency)

Demo

See these collocations used in a live vocabulary extraction pipeline: vocab-nlp Space

Use Cases

  • Collocation whitelists for NLP pipelines
  • Language learning applications (phrase extraction, vocabulary selection)
  • Linguistic research on multi-word expressions

Citation

If you use this data, please cite the underlying corpus:

@inproceedings{lison2016opensubtitles,
  title={OpenSubtitles2016: Extracting Large Parallel Corpora from Movie and TV Subtitles},
  author={Lison, Pierre and Tiedemann, J{\"o}rg},
  booktitle={Proceedings of the 10th LREC},
  year={2016}
}

License

CC-BY 4.0 (following OpenSubtitles licensing).

Downloads last month
26