Common Voice Scripted Speech 23.0 - Hazaragi

[Hazargi] — Hazargi (haz)

This datasheet is for version 23.0 of the the Mozilla Common Voice Scripted Speech dataset for Hazargi (haz). The dataset contains 11 hours of recorded speech (11 hours validated) from 7 speakers.

Language

Hazargi is the language of Hazara people who live in Pakistan, Afghanistan, Iran, Europe, Australia and America. It contains 32 alphabets and the script is in Arabic with some additional characters which are Hazargi based.

Variants

The dataset includes literature, history, folk stories and mostly poetry

Demographic information

The dataset includes the following distribution of age and gender.

Gender

Self-declared gender information, frequency refers to the number of clips annotated with this gender.

Age

Self-declared age information, frequency refers to the number of clips annotated with this age band.

Text corpus

As mentioned before that there has not been enough work on Hazargi so I tried to gather different books from the people around, arranged them and made the dataset. I gathered some 19 Hazargi books which are in different contexts like poems, folk stories, history and literature, the word count is around 893,112.

Writing system

The writing system is in Hazargi with an Arabic script that includes some additional Hazargi based characters.

Symbol table

ا٬ ب٬ پ٬ ت٬ ݖ٬ ج٬ چ٬ خ٬ د٬ ۮ٬ ر٬ ز٬ ژ٬ س٬ ش٬ غ٬ ف٬ ق٬ ک٬ گ٬ ل٬ م٬ ن٬ و٬ۉ٬ ۆ٬ ۂ٬ ی٬ ې٬ ݷ٬ ئ

Sample

There follows a randomly selected sample of five sentences from the corpus. بود نئبود بودیار بود ار چاری مئمئد یار بود دۂ کار کیدۉ بېمار بود دۂ خۉردۉ تئیار بود بود نئبود یئگ آجݷ بود

Text domains

General

Datasheet authors

Mushtaq Mughul Ali Toorani Jawad Khawari Raziq Kohzad Mustafa Elkhani Shawkat shaoor Hussain Ali Yosufi Aziz Fayaz Qadir Nayil Yaseen Zameer Loyaqath Ajiz Manzoor poya T. Malistani Amir Shah Haidri Farhad Zahidi Doc Zaibul Nisa

Licence

This dataset is released under the Creative Commons Zero (CC-0) licence. By downloading this data you agree to not determine the identity of speakers in the dataset.