EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

CAFE: Spontaneous code-switching speech dataset in Algerian dialect, French and English

Houssam Eddine-Othman Lachemat, Abbas Akli, Nourredine Oukas, Yassine El Kheir, Samia Haboussi, Shammur Absar Chowdhury

Data in Brief · 2025-12 · 63 · 112150

DOI: 10.1016/j.dib.2025.112150 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Data in Brief, 2025-12. Verfasst von Houssam Eddine-Othman Lachemat und 5 weiteren Autor:innen. Beteiligte Einrichtung: LIM Laboratory, Computer Sciences Department, Faculty of Exact Sciences, University of Bouira, 10000 Bouira, Algeria. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 22 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Publicly available datasets capturing spontaneous multilingual speech-especially those involving code-switching between Algerian Arabic, French, and English-are critically scarce. This lack of resources hinders the development of automatic speech recognition (ASR) and multilingual NLP systems for low-resource languages and under-represented Arabic dialects. We introduce CAFE, a novel dataset comprising approximately 37 h of spontaneous, in vivo human-human conversations among 100+ speakers across Algeria. The dialogues cover diverse everyday topics such as sports, science, and technology, and exhibit a rich range of natural conversational phenomena, including explicit code-switching, overlapping speech, non-lexical vocalizations (e.g., laughter, fillers, ambient noise), and dialectal variation reflecting Algeria's sociolinguistic landscape. CAFE is released in two tiers: CAFE-small (2 h 36 m): A fully human-annotated subset with high-quality transcriptions, vocal event labels, and linguistic annotations, supporting ASR evaluation, NLP tasks, and code-switching analysis. CAFE-large (∼34 h 35 m): The remainder of the corpus, automatically labeled, suitable for pretraining and semi-supervised learning. To support controlled experiments, CAFE-small includes two curated subsets: (i) CAFE-small-clean (2 h 18 m): Contains utterances with no overlapping speech. (ii) CAFE-small-overlap (17 m): Contains 23 files with overlap segments and timestamps. The dataset also provides rich metadata, including audio chunk IDs, dialect labels, and both raw and linguistically processed transcripts. CAFE offers a valuable resource for advancing ASR, dialect identification, and sociolinguistic analysis in multilingual and low-resource settings.

Schlagwörter

Stichwörter der Autor:innen

Automatic speech recognitionDialect identificationLinguistic annotationLow-resource languagesOverlapping speechPseudo-labelingSpeech segmentation

Autor:innen und Institutionen

  • LIM Laboratory, Computer Sciences Department, Faculty of Exact Sciences, University of Bouira, 10000 Bouira, Algeria.
  • Qatar Computing and Research Institute (QCRI), Hamad Bin Khalifa University, Doha, Qatar.
  • Speech and Language Technology, The German Research Center for Artificial Intelligence, 10559 Berlin, Germany.

Zitieren

APA

Lachemat, H. E., Akli, A., Oukas, N., El Kheir, Y., Haboussi, S., & Chowdhury, S. A. (2025). CAFE: Spontaneous code-switching speech dataset in Algerian dialect, French and English. Data in Brief, 63, 112150. https://doi.org/10.1016/j.dib.2025.112150

Vancouver

Lachemat HE, Akli A, Oukas N, El Kheir Y, Haboussi S, Chowdhury SA. CAFE: Spontaneous code-switching speech dataset in Algerian dialect, French and English. Data in Brief. 2025;63:112150. doi:10.1016/j.dib.2025.112150

BibTeX
@article{Lachemat2025,
  author = {Lachemat, Houssam Eddine-Othman and Akli, Abbas and Oukas, Nourredine and El Kheir, Yassine and Haboussi, Samia and Chowdhury, Shammur Absar},
  title = {CAFE: Spontaneous code-switching speech dataset in Algerian dialect, French and English},
  journal = {Data in Brief},
  year = {2025},
  volume = {63},
  pages = {112150},
  doi = {10.1016/j.dib.2025.112150}
}
Kostenlos starten →