Detect the Language of Text
Franc has no external dependencies and supports 310 languages; all languages spoken by more than one million speakers. Franc is a port of the JavaScript project of the same name, see https://github.com/wooorm/franc.
install.packages("franc")
library(franc)
Simply supply the text, and franc detects its language:
franc("Alle menslike wesens word vry")
#> [1] "afr"
franc("এটি একটি ভাষা একক IBM স্ক্রিপ্ট")
#> [1] "ben"
franc("Alle mennesker er født frie og")
#> [1] "nno"
head(franc_all("O Brasil caiu 26 posições"))
#> language score
#> 1 por 1.0000000
#> 2 src 0.8800937
#> 3 glg 0.8702576
#> 4 snn 0.8637002
#> 5 bos 0.8168618
#> 6 hrv 0.8103044
und
is the undefined
language, this is
returned if the input is too short (shorter than 10 characters by
default).
franc("the")
#> [1] "und"
franc("the", min_length = 3)
#> [1] "sco"
You can provide a whitelist or a blacklist:
franc_all("O Brasil caiu 26 posições",
whitelist = c("por", "src", "glg", "spa"))
#> language score
#> 1 por 1.0000000
#> 2 src 0.8800937
#> 3 glg 0.8702576
#> 4 spa 0.7995316
head(franc_all("O Brasil caiu 26 posições",
blacklist = c("src", "glg", "lav")))
#> language score
#> 1 por 1.0000000
#> 2 snn 0.8637002
#> 3 bos 0.8168618
#> 4 hrv 0.8103044
#> 5 cat 0.8065574
#> 6 spa 0.7995316
The R version of franc supports 310 languages. By default only the
languages with more than 1 million speakers are used, this is 175
languages. The min_speakers
argument can relax this, and
allows using more languages:
head(franc_all("O Brasil caiu 26 posições"))
#> language score
#> 1 por 1.0000000
#> 2 src 0.8800937
#> 3 glg 0.8702576
#> 4 snn 0.8637002
#> 5 bos 0.8168618
#> 6 hrv 0.8103044
head(franc_all("O Brasil caiu 26 posições", min_speakers = 0))
#> language score
#> 1 lad 1.0000000
#> 2 por 0.9442724
#> 3 pov 0.8788147
#> 4 ast 0.8677576
#> 5 roh 0.8363556
#> 6 src 0.8310482
MIT © Mango Solutions, Titus Wormer, Maciej Ceglowski, Jacob R. Rideout, Kent S. Johnson, Gábor Csárdi