michal-tichy/canine-c-multilabel-li
Text Classification • 0.1B • Updated • 9
Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
json
Size:
< 1K
ArXiv:
Tags:
language_identification
License:
non-language list | single list | joint list | cyrilic_to_czech list | cyrilic_to_english list | antspeak list | arabic list | diacritics list | leet list | homoglyphs list |
|---|---|---|---|---|---|---|---|---|---|
[{"text":"marek.sykora.krizova.cz","label":[]},{"text":"kratochvil.cz","label":[]},{"text":"sedlacek(...TRUNCATED) | [{"text":"Ale nie je to tak.","label":["slk_Latn"]},{"text":"V tejto chvíli nie je.","label":["slk_(...TRUNCATED) | [{"text":"Neznamená to ale, že je to nemožné.","label":["ces_Latn","slk_Latn"]},{"text":"A to an(...TRUNCATED) | [{"text":"№ 10; Lubimov V.P. O Laptevskom spiske Pravdy Russkoj // Pravda Russkaja.","label":["rus(...TRUNCATED) | [{"text":"№ 10; Lyubimov V.P. O Laptevskom spiske Pravdy' Russkoj // Pravda Russkaya.","label":["r(...TRUNCATED) | [{"text":"' 1 9 7 9 s t a n d u p t o u r . c i t a t i o n I n 1 9 9 8 , h e r e l e a s e d B r a (...TRUNCATED) | [{"text":" ۱۲۸ شەشىمى) بەرىلگەن.","label":["kaz_Cyrl"],"original":"№ 128 шеш(...TRUNCATED) | [{"text":"%20ligy%20v%20h%C3%A1dzanej%20mu%C5%BEov/ 1. liga hadzanarov 2007/08 na stranke zivotpo.",(...TRUNCATED) | [{"text":"' 1979 standup to_r. c1tati0n 1n 1998, he rel3@s3d Brand0n Project, 4 bl_3s 4lbum.","label(...TRUNCATED) | [{"text":"%20ligy%20v%20һ%С3%A1dzanеј%20mu%С5%BEov/ 1. ligа hádzаnárov 2007/08 nа ѕtrán(...TRUNCATED) |
CHALIS - Challenging Language Identification Samples
This dataset was presented in the paper CHALIS: A Challenge Dataset for Language Identification in Difficult Scenarios.
CHALIS is a multilabel dataset composed of several sections aimed at testing language identification abilities in difficult scenarios.
Main contribution comes in the form of gathering and classifying by a human expert of sentences belonging to four language pairs (Spanish - Catalan, Portuguese - Galician, Danish - Norwegian, Czech - Slovak)
These were sorted into two categories: joint for sentences which belong to both languages and single which belong to only one of the languages.
Sections
The dataset has the four following sections (with further subsections in case of transiteration):
- non-language - The non-language data such as emails, ISBNs, etc.
- single - The sentences which were assigned only to one language during dataset gathering. This creates a fairly challenging dataset for recognizing the less prevalent language of the pair.
- joint - The sentences which were assigned to both cousin languages during dataset gathering.
- transliteration
- czech - Transliteration of Russian into the Czech alphabet.
- cyrilic - Transliteration of Cyrilic written sentences into latin characters.
- arabic - Transliteration of Kazakh language written in Cyrilic into Arabic.
- latin - Transliteration of diacritic heavy language sentences into purely latin characters.
- antspeak - Transliteration of English sentences into antspeak.
- leet - Transliteration of English sentences into leetspeak.
- random - Substitution of random letters in English, Czech, and Slovak sentences for ones from Russian and Greek alphabets which preserve the human readable renderings of the sentences.
- Downloads last month
- 122
Models trained or fine-tuned on michal-tichy/CHALIS