Dataset Viewer
Auto-converted to Parquet
id
int64
2
26.7M
langvar
int64
1
13.4k
txt
stringlengths
1
703
txt_degr
stringlengths
1
670
meaning
int64
6
37.8M
langvar_uid
stringlengths
7
7
25,858,519
2,696
è-ṇīē
enie
37,566,322
aaa-000
18,580,658
2,696
Ghotuo
ghotuo
35,678,636
aaa-000
18,699,868
2,696
LELE
lele
20,033,948
aaa-000
18,725,074
2,696
Nw~o
nwo
20,033,958
aaa-000
18,631,881
2,696
Ze
ze
20,033,923
aaa-000
18,735,966
2,696
Zo
zo
20,033,962
aaa-000
18,651,919
2,696
be
be
20,033,929
aaa-000
18,710,058
2,696
bo
bo
20,033,952
aaa-000
18,663,113
2,696
dE
de
20,033,934
aaa-000
18,769,596
2,696
do
do
20,033,981
aaa-000
18,642,544
2,696
du
du
20,033,926
aaa-000
18,642,544
2,696
du
du
20,033,957
aaa-000
18,657,462
2,696
fia
fia
20,033,932
aaa-000
18,792,167
2,696
ho
ho
20,033,996
aaa-000
18,678,412
2,696
hogi
hogi
20,033,943
aaa-000
18,694,635
2,696
ko
ko
20,033,947
aaa-000
18,730,639
2,696
mE
me
20,033,961
aaa-000
18,730,639
2,696
mE
me
20,033,979
aaa-000
18,608,901
2,696
ma
ma
20,033,907
aaa-000
18,608,902
2,696
mama
mama
20,033,907
aaa-000
18,599,277
2,696
meme
meme
20,031,129
aaa-000
18,683,762
2,696
o
o
20,033,944
aaa-000
18,689,072
2,696
rue
rue
20,033,945
aaa-000
18,646,854
2,696
ta
ta
20,033,927
aaa-000
18,741,542
2,696
u*
u
20,033,965
aaa-000
18,620,968
2,696
va
va
20,033,916
aaa-000
18,620,968
2,696
va
va
20,034,004
aaa-000
18,798,050
2,696
vo
vo
20,033,999
aaa-000
18,637,228
2,696
wa
wa
20,033,925
aaa-000
18,673,960
2,696
xua
xua
20,033,938
aaa-000
17,468,110
2,696
è-vā
eva
37,566,316
aaa-000
17,477,035
2,696
èē-nè
eene
37,566,318
aaa-000
17,472,597
2,696
èē-sà
eesa
37,566,317
aaa-000
17,500,049
2,696
ì-gbēè
igbee
37,566,324
aaa-000
17,489,031
2,696
ì-hī
ihi
37,566,321
aaa-000
17,496,405
2,696
ì-sī
isi
37,566,323
aaa-000
17,481,208
2,696
ìí-žè
iize
37,566,319
aaa-000
17,485,214
2,696
ìē-hà
ieha
37,566,320
aaa-000
23,277,415
2,696
oghumhi
oghumhi
28,450,468
aaa-000
18,068,704
1,604
19,793,666
aab-000
18,068,702
1,604
gùnùvɔ
gunuvɔ
19,793,664
aab-000
18,068,753
1,604
yeráma
yerama
19,793,717
aab-000
18,068,743
1,604
ə̀va
əva
19,793,707
aab-000
18,651,920
1,604
3Su
3su
20,033,929
aab-000
18,668,358
1,604
3kuph~u
3kuphu
20,033,935
aab-000
18,764,402
1,604
3mu
3mu
20,033,979
aab-000
18,694,636
1,604
3ny~ine
3nyine
20,033,947
aab-000
18,637,229
1,604
3va
3va
20,033,925
aab-000
18,792,168
1,604
Sini
sini
20,033,996
aab-000
18,775,834
1,604
Sira
sira
20,033,986
aab-000
17,145,780
1,604
Təsu
təsu
35,677,607
aab-000
18,678,413
1,604
aSoto
asoto
20,033,943
aab-000
18,068,925
1,604
aanε̃
aanε
19,793,889
aab-000
18,069,136
1,604
agbaraʤa
agbaraʤa
19,794,115
aab-000
18,646,855
1,604
agb~e
agbe
20,033,927
aab-000
18,786,514
1,604
agb~ogb~o
agbogbo
20,033,990
aab-000
18,068,696
1,604
agbà
agba
19,793,658
aab-000
18,068,909
1,604
agrá
agra
19,793,873
aab-000
18,781,502
1,604
aki akere
akiakere
20,033,989
aab-000
18,720,890
1,604
ahro
ahro
20,033,957
aab-000
18,620,969
1,604
ahurwi
ahurwi
20,033,916
aab-000
18,068,551
1,604
ahóró bini
ahorobini
19,793,511
aab-000
18,759,293
1,604
akp~okp~a
akpokpa
20,033,978
aab-000
18,699,869
1,604
alimi
alimi
20,033,948
aab-000
18,068,864
1,604
ambídákù
ambidaku
19,793,828
aab-000
18,068,863
1,604
ambíhùwí
ambihuwi
19,793,827
aab-000
18,068,679
1,604
amo húhwá
amohuhwa
19,793,639
aab-000
18,068,783
1,604
aniŋgá
aniŋga
19,793,747
aab-000
18,615,806
1,604
any~imb~ere
anyimbere
20,033,915
aab-000
18,801,429
1,604
aph~e
aphe
20,034,000
aab-000
18,068,633
1,604
asubé
asube
19,793,593
aab-000
18,069,151
1,604
ataba
ataba
19,794,130
aab-000
18,673,961
1,604
atama
atama
20,033,938
aab-000
18,069,110
1,604
atambo
atambo
19,794,089
aab-000
18,068,941
1,604
atíʃa
atiʃa
19,793,905
aab-000
18,068,926
1,604
atúŋgú
atuŋgu
19,793,890
aab-000
18,068,855
1,604
atɔ́
atɔ
19,793,819
aab-000
18,068,824
1,604
avɔ
avɔ
19,793,788
aab-000
18,068,552
1,604
ayé
aye
19,793,512
aab-000
18,069,106
1,604
aŋbasa kpànà
aŋbasakpana
19,794,085
aab-000
18,069,105
1,604
aŋbasa kòbǒ
aŋbasakobo
19,794,084
aab-000
18,069,103
1,604
aŋbasa lomo
aŋbasalomo
19,794,082
aab-000
18,069,104
1,604
aŋbasa ŋgàtɔ́
aŋbasaŋgatɔ
19,794,083
aab-000
18,068,954
1,604
aɥé
aɥe
19,793,918
aab-000
18,069,185
1,604
aʃúnyìŋgwe
aʃunyiŋgwe
19,794,166
aab-000
18,068,953
1,604
aʧɔ́
aʧɔ
19,793,917
aab-000
18,715,646
1,604
bEnE
bene
20,033,955
aab-000
18,747,604
1,604
ba
ba
20,033,970
aab-000
18,068,768
1,604
bibiya
bibiya
19,793,732
aab-000
18,657,463
1,604
binato
binato
20,033,932
aab-000
18,068,718
1,604
binatɔ
binatɔ
19,793,680
aab-000
16,482,679
1,604
bini
bini
19,793,463
aab-000
18,069,184
1,604
binyíŋgwe
binyiŋgwe
19,794,165
aab-000
18,642,545
1,604
birokawri
birokawri
20,033,926
aab-000
18,068,948
1,604
birírí
biriri
19,793,912
aab-000
16,804,728
1,604
birɔ
birɔ
19,793,468
aab-000
18,068,882
1,604
biʃí
biʃi
19,793,846
aab-000
18,068,683
1,604
blà
bla
19,793,644
aab-000
18,068,517
1,604
blãblãsa
blablasa
19,793,476
aab-000
18,068,569
1,604
blê
ble
19,793,529
aab-000
End of preview. Expand in Data Studio

Dataset Card for panlex-meanings

This is a dataset of words in several thousand languages, extracted from https://panlex.org.

Dataset Details

Dataset Description

This dataset has been extracted from https://panlex.org (the 20240301 database dump) and rearranged on the per-language basis.

Each language subset consists of expressions (words and phrases). Each expression is associated with some meanings (if there is more than one meaning, they are in separate rows).

Thus, by joining per-language datasets by meaning ids, one can obtain a bilingual dictionary for the chosen language pair.

  • Curated by: David Dale (@cointegrated), based on a snapshot of the Panlex database (https://panlex.org/snapshot).
  • Language(s) (NLP): The Panlex database mentions 7558 languages, but only 6241 of them have at least one entry (where entry is a combination of expression and meaning), and only 1012 have at least 1000 entries. These 1012 languages are tagged in the current dataset.
  • License: CC0 1.0 Universal License, as explained in https://panlex.org/license.

Dataset Sources [optional]

Uses

The intended use of the dataset is to extract bilingual dictionaries for the purposes of language learning by machines or humans.

The code below illustrates how the dataset could be used to extract a bilingual Avar-English dictionary.

from datasets import load_dataset
ds_ava = load_dataset('cointegrated/panlex-meanings', name='ava', split='train')
ds_eng = load_dataset('cointegrated/panlex-meanings', name='eng', split='train')
df_ava = ds_ava.to_pandas()
df_eng = ds_eng.to_pandas()

df_ava_eng = df_ava.merge(df_eng, on='meaning', suffixes=['_ava', '_eng']).drop_duplicates(subset=['txt_ava', 'txt_eng'])
print(df_ava_eng.shape)
# (10565, 11)

print(df_ava_eng.sample(5)[['txt_ava', 'txt_eng', 'langvar_uid_ava']])
#           txt_ava txt_eng langvar_uid_ava
# 7921        калим     rug         ava-002
# 3279       хІераб     old         ava-001
# 41     бакьулълъи  middle         ava-000
# 9542        шумаш    nose         ava-006
# 15030     гӏащтӏи     axe         ava-000

Apart from these direct translations, one could also try extracting multi-hop translations (e.g. enrich the direct Avar-English word pairs with the word pairs that share a common Russian translation). However, given that many words have multiple meaning, this approach usually generates some false translations, so it should be used with caution.

Dataset Structure

The dataset is split by languages, denoted by their ISO 639 codes. Each language might contain multiple varieties; they are annotated within each per-language split.

To determine a code for your language, please consult the https://panlex.org webside. For additional information about a language, you may also want to consult https://glottolog.org/.

Each split contains the following fields:

  • id (int): id of the expression
  • langvar (int): id of the language variaety
  • txt (str): the full text of the expression
  • txt_degr (str): degraded (i.e. simplified to facilitate lookup) text
  • meaning (int): id of the meaning. This is the column to join for obtaining synonyms (within a language) or translations (across languages)
  • langvar_uid (str): more human-readable id of the language (e.g. eng-000 stands for generic English, eng-001 for simple English, eng-004 for American English). These ids could be looked up in the language dropdown at https://vocab.panlex.org/.

Dataset Creation

This dataset has been extracted from https://panlex.org (the 20240301 database dump) and automatically rearranged on the per-language basis.

The rearrangement consisted of the following steps:

  1. Grouping together the language varieties from the langvar table with the same lang_code.
  2. For each language, selecting the corresponding subset from the expr table.
  3. Joining the selected set with the denotation table, to get the meaning id. This increases the number of rows (for some languages, x5), because multiple meannings may be attached to the same expression.

Bias, Risks, and Limitations

As with any multilingual dataset, Panlex data may exhbit the problem of under- and mis-representation of some languages.

The dataset consists primarily of the standard written forms ("lemmas") of the expressions, so it may not well represent their usage within a language.

Citation

Kamholz, David, Jonathan Pool, and Susan M. Colowick. 2014. PanLex: Building a Resource for Panlingual Lexical Translation. Proceedings of the 9th International Conference on Language Resources and Evaluation (LREC 2014).

BibTeX:

@inproceedings{kamholz-etal-2014-panlex,
    title = "{P}an{L}ex: Building a Resource for Panlingual Lexical Translation",
    author = "Kamholz, David  and
      Pool, Jonathan  and
      Colowick, Susan",
    editor = "Calzolari, Nicoletta  and
      Choukri, Khalid  and
      Declerck, Thierry  and
      Loftsson, Hrafn  and
      Maegaard, Bente  and
      Mariani, Joseph  and
      Moreno, Asuncion  and
      Odijk, Jan  and
      Piperidis, Stelios",
    booktitle = "Proceedings of the Ninth International Conference on Language Resources and Evaluation ({LREC}'14)",
    month = may,
    year = "2014",
    address = "Reykjavik, Iceland",
    publisher = "European Language Resources Association (ELRA)",
    url = "http://www.lrec-conf.org/proceedings/lrec2014/pdf/1029_Paper.pdf",
    pages = "3145--3150",
    abstract = "PanLex, a project of The Long Now Foundation, aims to enable the translation of lexemes among all human languages in the world. By focusing on lexemic translations, rather than grammatical or corpus data, it achieves broader lexical and language coverage than related projects. The PanLex database currently documents 20 million lexemes in about 9,000 language varieties, with 1.1 billion pairwise translations. The project primarily engages in content procurement, while encouraging outside use of its data for research and development. Its data acquisition strategy emphasizes broad, high-quality lexical and language coverage. The project plans to add data derived from 4,000 new sources to the database by the end of 2016. The dataset is publicly accessible via an HTTP API and monthly snapshots in CSV, JSON, and XML formats. Several online applications have been developed that query PanLex data. More broadly, the project aims to make a contribution to the preservation of global linguistic diversity.",
}

Glossary

To understand the terms like "language", "language variety", "expression" and "meaning" more precisely, please read the Panlex documentation on their data model and database design.

Downloads last month
216