README

PHP library to identify human languages from text samples. Returns confidence scores for each.

Installation

PEAR

$ pear install Text_LanguageDetect

Composer

$ composer require pear/text_languagedetect

Usage

Also see the examples in the docs/ directory and the official documentation.

Language detection

Simple language detection:

<?php
require_once 'Text/LanguageDetect.php';

$text = 'Was wäre, wenn ich Ihnen das jetzt sagen würde?';

$ld = new Text_LanguageDetect();
$language = $ld->detectSimple($text);

echo $language;
//output: german

Show the three most probable languages with their confidence score:

<?php
require_once 'Text/LanguageDetect.php';

$text = 'Was wäre, wenn ich Ihnen das jetzt sagen würde?';

$ld = new Text_LanguageDetect();
//3 most probable languages
$results = $ld->detect($text, 3);

foreach ($results as $language => $confidence) {
    echo $language . ': ' . number_format($confidence, 2) . "\n";
}

//output:
//german: 0.35
//dutch: 0.25
//swedish: 0.20
?>

Language code

Instead of returning the full language name, ISO 639-2 two and three letter codes can be returned:

<?php
require_once 'Text/LanguageDetect.php';
$ld = new Text_LanguageDetect();

//will output the ISO 639-1 two-letter language code
// "de"
$ld->setNameMode(2);
echo $ld->detectSimple('Das ist ein kleiner Text') . "\n";

//will output the ISO 639-2 three-letter language code
// "deu"
$ld->setNameMode(3);
echo $ld->detectSimple('Das ist ein kleiner Text') . "\n";
?>

Supported languages

albanian
arabic
azeri
bengali
bulgarian
cebuano
croatian
czech
danish
dutch
english
estonian
farsi
finnish
french
german
hausa
hawaiian
hindi
hungarian
icelandic
indonesian
italian
kazakh
kyrgyz
latin
latvian
lithuanian
macedonian
mongolian
nepali
norwegian
pashto
pidgin
polish
portuguese
romanian
russian
serbian
slovak
slovene
somali
spanish
swahili
swedish
tagalog
turkish
ukrainian
urdu
uzbek
vietnamese
welsh

Notes

Where are the data from?

I don't recall where I got the original data set. It's just the frequencies of 3-letter combinations in each supported language. It could be generated from a few random wikipedia pages from each language.

pear / text_languagedetect

Maintainers

Details