Project Overview
A curated overview of verified Kurdish corpus-creation research, including general text collections, educational corpora, syntactically annotated resources, regional varieties and under-resourced Kurdish data.
Technologies & Methods
Applications
- Language Research
- NLP Model Training
- Linguistic Analysis
Related Publications
Sorani Kurdish versus Kurmanji Kurdish: An Empirical Comparison
Kyumars Sheykh Esmaili , Shahin Salavati (2013)
An empirical comparison of Sorani and Kurmanji based on the Pewan news corpus, examining linguistic, lexical and orthographic characteristics across the two Kurdish varieties.
A Dependency Treebank for Kurmanji Kurdish
Memduh Gökırmak , Francis M. Tyers (2017)
Development of a Universal Dependencies treebank for Kurmanji Kurdish using morphological analysis and manual annotation, providing syntactic training and evaluation data for Kurdish language …
Toward Kurdish language processing: Experiments in collecting and processing the AsoSoft text corpus
Hadi Veisi , Mohammad Mohammadamini , Hawre Hosseini (2020)
Development and analysis of a large Central Kurdish text corpus collected from websites, books and magazines, with Kurdish-specific normalization and TEI-based organization.
Developing a Fine-grained Corpus for a Less-resourced Language: the case of Kurdish
Roshna Abdulrahman , Hossein Hassani , Sina Ahmadi (2019)
Development of the Kurdish Textbooks Corpus, a normalized Sorani corpus built from K-12 textbooks and organized across twelve educational subject areas.
Approaches to Corpus Creation for Low-Resource Language Technology: the Case of Southern Kurdish and Laki
Sina Ahmadi , Zahra Azin , Sara Belelli , Antonios Anastasopoulos (2023)
A corpus-creation study for severely under-resourced Southern Kurdish varieties and Laki using web extraction, radio material and digitized fieldwork.
Language and Speech Technology for Central Kurdish Varieties
Sina Ahmadi , Daban Q. Jaff , Md Mahfuz Ibn Alam , Antonios Anastasopoulos (2024)
Development of CORDI, a variety-sensitive Central Kurdish corpus containing transcripts and regional metadata from movies and television across six urban varieties.
Related Datasets
Pewan Corpus and Retrieval Test Collection
Published: January 1, 2013 | Size: Sorani: 115,340 documents / 18,110,723 words; Kurmanji: 25,572 documents / 4,120,027 words
A comparable Sorani and Kurmanji news corpus collected from material published between 2003 and 2012, accompanied by information-retrieval resources.
UD Northern Kurdish-Kurmanji
Published: January 1, 2017 | Size: 754 sentences; 10,188 tokens; 10,259 syntactic words in the inspected current UD release
A Universal Dependencies treebank for Northern Kurdish (Kurmanji), containing manually reviewed linguistic annotation for syntax and morphology.
AsoSoft Kurdish Text Corpus
Published: January 1, 2018 | Size: Full research corpus: approximately 188M tokens and 458,000 documents; public releases: 75M and 5M tokens
A large Central Kurdish text corpus collected from websites, books and magazines. The research paper reports approximately 188 million tokens and 458,000 documents; public releases are smaller …
Kurdish Textbooks Corpus (KTC)
Published: January 1, 2019 | Size: 31 textbooks; 693,800 tokens; 110,297 types
A Sorani corpus created from Kurdish K-12 textbooks and organized by educational subject.
Southern Kurdish and Laki Corpus Collection
Published: January 1, 2023 | Size: Feyli: 15,985 documents / 2.182M tokens; Kermanshahi: 18 / 10,127 tokens; Laki: 7 / 6,340 tokens
A collection developed for severely under-resourced Southern Kurdish varieties and Laki using web material, radio scripts and digitized fieldwork.
CORDI
Published: January 1, 2024 | Size: 311 movies/episodes; 186,038 utterances; more than 100 hours of audiovisual material
A corpus covering six Central Kurdish urban varieties using volunteer transcriptions of movies and television, with dialect and speaker metadata.
Kurdish Sorani (Central Kurdish) Text Corpus - UKH/AIIC-Kurdish Academy
Published: July 11, 2026 | Size: Not verified
A Central Kurdish text corpus built from books in 15 academic subject categories through collaboration involving UKH/AIIC and the Kurdish Academy. No verified document, sentence or token count was …