id
int64
783k
26.6M
langvar
int64
1
1
txt
stringlengths
1
29
txt_degr
stringlengths
1
20
meaning
int64
40.2k
37.6M
langvar_uid
stringclasses
1 value
26,111,254
1
qafar af
qafaraf
36,299,577
aar-000
26,111,255
1
qafar
qafar
36,299,577
aar-000
26,238,566
1
Afaraf
afaraf
36,867,197
aar-000
26,238,566
1
Afaraf
afaraf
36,867,198
aar-000
1,684,561
1
dʼa
da
19,080,613
aar-000
1,453,573
1
tah
tah
2,675,740
aar-000
26,069,335
1
Eroppah
eroppah
36,229,630
aar-000
1,452,951
1
abe
abe
2,675,921
aar-000
1,452,951
1
abe
abe
2,675,924
aar-000
1,453,204
1
ellel
ellel
2,675,879
aar-000
1,689,216
1
áf
af
19,080,854
aar-000
1,453,024
1
amoyta
amoyta
2,676,263
aar-000
1,453,024
1
amoyta
amoyta
2,676,284
aar-000
1,453,024
1
amoyta
amoyta
2,676,362
aar-000
18,715,653
1
3Ng~u
3ngu
20,033,955
aar-000
18,725,077
1
3Xub
3xub
20,033,958
aar-000
18,730,646
1
3bul
3bul
20,033,961
aar-000
18,599,285
1
3nu
3nu
20,031,129
aar-000
18,657,467
1
3r3b
3r3b
20,033,932
aar-000
18,699,874
1
3r3b3
3r3b3
20,033,948
aar-000
18,754,267
1
3ro
3ro
20,033,976
aar-000
18,604,033
1
3tu
3tu
20,031,130
aar-000
18,651,926
1
3yuft3
3yuft3
20,033,929
aar-000
23,303,833
1
Afrikah
afrikah
36,229,626
aar-000
23,303,833
1
Afrikah
afrikah
28,508,376
aar-000
4,540,577
1
Almazán
almazan
1,189,814
aar-000
3,723,643
1
Arabic
arabic
4,444,606
aar-000
3,727,298
1
English
english
4,444,708
aar-000
6,097,848
1
Enqlizxsh - English
enqlizxshenglish
1,673,301
aar-000
3,728,248
1
French
french
4,444,731
aar-000
3,726,575
1
German
german
4,444,687
aar-000
3,730,652
1
Italian
italian
4,444,794
aar-000
6,097,849
1
My Happy Ending
myhappyending
1,641,679
aar-000
5,895,368
1
Simon Doull
simondoull
1,542,123
aar-000
19,667,027
1
United States
unitedstates
31,737,076
aar-000
18,663,119
1
X3b3l
x3b3l
20,033,934
aar-000
18,786,519
1
X3lE
x3le
20,033,990
aar-000
1,453,056
1
a saku
asaku
2,675,661
aar-000
1,453,056
1
a saku
asaku
28,095,618
aar-000
1,621,690
1
ab
ab
19,081,215
aar-000
1,621,690
1
ab
ab
19,081,216
aar-000
1,621,690
1
ab
ab
19,081,586
aar-000
1,621,690
1
ab
ab
19,082,248
aar-000
18,663,121
1
ab3la
ab3la
20,033,934
aar-000
1,452,946
1
abatasa
abatasa
2,675,628
aar-000
1,452,946
1
abatasa
abatasa
28,082,280
aar-000
17,463,567
1
abaːl
abal
19,081,631
aar-000
1,452,948
1
abba-bada
abbabada
2,675,828
aar-000
1,452,948
1
abba-bada
abbabada
2,675,830
aar-000
1,452,948
1
abba-bada
abbabada
28,138,052
aar-000
1,452,948
1
abba-bada
abbabada
28,138,118
aar-000
1,452,949
1
abbah-ina
abbahina
2,675,979
aar-000
1,452,950
1
abbire
abbire
2,675,655
aar-000
23,065,322
1
abeesa
abeesa
28,030,564
aar-000
1,452,953
1
able
able
2,676,421
aar-000
1,452,954
1
aboyya
aboyya
2,675,977
aar-000
1,452,956
1
abu
abu
2,675,636
aar-000
1,452,956
1
abu
abu
2,676,154
aar-000
1,452,956
1
abu
abu
28,219,762
aar-000
1,452,957
1
abuke
abuke
2,676,120
aar-000
18,730,647
1
abul
abul
20,033,961
aar-000
1,452,958
1
abur
abur
2,676,358
aar-000
1,452,959
1
aburo
aburo
2,675,998
aar-000
1,452,959
1
aburo
aburo
2,676,409
aar-000
1,720,679
1
abäs
abas
19,081,273
aar-000
1,697,941
1
abäsáː
abasa
19,080,809
aar-000
2,292,288
1
abəːsə́ːmaː
abəsəma
19,080,684
aar-000
17,463,691
1
abəːyáː
abəya
19,082,154
aar-000
1,716,512
1
abə́la
abəla
19,080,826
aar-000
17,463,690
1
abə́ːyaː
abəya
19,082,153
aar-000
1,452,952
1
abʼha
abha
2,676,345
aar-000
17,463,448
1
adar
adar
19,081,450
aar-000
17,463,448
1
adar
adar
19,081,458
aar-000
17,463,448
1
adar
adar
19,081,460
aar-000
1,452,963
1
adarras
adarras
2,675,918
aar-000
1,452,966
1
addah
addah
2,675,889
aar-000
1,452,967
1
addal
addal
2,675,819
aar-000
1,452,970
1
adige
adige
2,676,309
aar-000
1,452,972
1
admo
admo
2,675,756
aar-000
1,452,973
1
admo abe
admoabe
2,675,757
aar-000
1,452,975
1
adure
adure
2,676,276
aar-000
1,652,805
1
af
af
19,080,854
aar-000
1,652,805
1
af
af
19,081,737
aar-000
1,652,805
1
af
af
19,081,929
aar-000
1,733,474
1
af maː-liː
afmali
19,080,963
aar-000
17,463,477
1
affaʼra
affara
19,081,487
aar-000
1,703,015
1
agaboːytá-t angəːg
agaboytatangəg
19,080,889
aar-000
1,715,806
1
agaboːytát ábbaː
agaboytatabba
19,080,697
aar-000
1,614,621
1
agaboːytát-iná
agaboytatina
19,080,699
aar-000
1,613,161
1
agaboːytáː
agaboyta
19,080,650
aar-000
1,613,161
1
agaboːytáː
agaboyta
19,080,660
aar-000
1,681,354
1
agaboːytáː ak raːbtä́ nəːm
agaboytaakrabtanəm
19,080,711
aar-000
1,452,980
1
agabu
agabu
2,675,931
aar-000
17,463,705
1
agam
agam
19,082,215
aar-000
1,452,981
1
agda
agda
2,676,246
aar-000
1,695,672
1
aggaʼf
aggaf
19,080,934
aar-000
1,452,982
1
aggile
aggile
2,675,736
aar-000
1,452,982
1
aggile
aggile
2,675,831
aar-000
1,452,982
1
aggile
aggile
2,675,839
aar-000
1,452,984
1
agime
agime
2,676,001
aar-000

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
464
Edit dataset card