Kazakh CV dictionary v2.0.0#

Maintainer: Vox Communis
Language: Kazakh
Dialect: N/A
Phone set: Epitran
Number of words: 2,104
Phones: a b c d e h i j k l m n o p q r s t u w z æ ŋ ɑ ɕː ə ɡ ɪ ʁ ʃ ʊ ʏ ʒ χ
License: CC-0
Compatible MFA version: v2.0.0
Citation:

@misc{Ahn_Chodroff_2022,
	author={Ahn, Emily and Chodroff, Eleanor},
	title={VoxCommunis Corpus},
	address={\url{https://osf.io/t957v}},
	publisher={OSF},
	year={2022},
	month={Jan}
}

If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.

Acoustic models

Kazakh CV acoustic model v2.0.0

Installation#

Install from the MFA command line:

mfa model download dictionary kazakh_cv

Or download from the release page.

Intended use#

This dictionary is intended for forced alignment of Kazakh transcripts.

This dictionary uses the Epitran phone set for Kazakh, and was used in training the Kazakh Epitran acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors#

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations#

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias#

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.

IPA Charts#

Consonants#

Obstruent symbols to the left of are unvoiced and those to the right are voiced.

Manner	Labial	Alveolar	Alveopalatal	Palatal	Velar	Uvular	Glottal
Nasal	m Occurrences: 343 Examples: * мырза: [m ə r z ɑ] * мықты: [m ə q t ə] * өлім: [ʏ l ɪ m] * жемес: [ʒ j e m j e s]	n Occurrences: 814 Examples: * еріне: [j e r ɪ n j e] * найза: [n ɑ j z ɑ] * орыны: [o r ə n ə] * туған: [t w ʁ ɑ n]			ŋ Occurrences: 308 Examples: * еңбек: [j e ŋ b j e k] * атың.: [ɑ t ə ŋ] * үйің: [ʏ j ɪ ŋ] * күйің: [k ʏ j ɪ ŋ]
Stop	p Occurrences: 179 Examples: * пышақ: [p ə ʃ ɑ q] * аспақ: [ɑ s p ɑ q] * жауап: [ʒ ɑ w ɑ p] * ішпе,: [ɪ ʃ p j e] b Occurrences: 334 Examples: * еңбек: [j e ŋ b j e k] * бар,: [b ɑ r] * бақыт: [b ɑ q ə t] * түбі: [t ʏ b ɪ]	t Occurrences: 673 Examples: * атың.: [ɑ t ə ŋ] * туған: [t w ʁ ɑ n] * тісін: [t ɪ s ɪ n] * тіле.: [t ɪ l j e] d Occurrences: 562 Examples: * досқа: [d o s q ɑ] * аздың: [ɑ z d ə ŋ] * қойды: [q o j d ə] * суда: [s w d ɑ]		c Occurrences: 1 Examples:	k Occurrences: 319 Examples: * еңбек: [j e ŋ b j e k] * күйің: [k ʏ j ɪ ŋ] * кең: [k j e ŋ] * кісің: [k ɪ s ɪ ŋ] ɡ Occurrences: 127 Examples: * ізгі: [ɪ z ɡ ɪ] * елігі: [j e l ɪ ɡ ɪ] * үзген: [ʏ z ɡ j e n] * үйге: [ʏ j ɡ j e]	q Occurrences: 539 Examples: * досқа: [d o s q ɑ] * қисық: [q i j s ə q] * пышақ: [p ə ʃ ɑ q] * аспақ: [ɑ s p ɑ q]
Sibilant		s Occurrences: 651 Examples: * досқа: [d o s q ɑ] * қисық: [q i j s ə q] * аспақ: [ɑ s p ɑ q] * тісін: [t ɪ s ɪ n] z Occurrences: 220 Examples: * найза: [n ɑ j z ɑ] * мырза: [m ə r z ɑ] * аздың: [ɑ z d ə ŋ] * сөз: [s ʏ z]	ʃ Occurrences: 179 Examples: * пышақ: [p ə ʃ ɑ q] * қылыш: [q ə l ə ʃ] * аштық: [ɑ ʃ t ə q] * ішпе,: [ɪ ʃ p j e] ʒ Occurrences: 288 Examples: * жүре: [ʒ ʏ r j e] * жылуы: [ʒ ə l w ə] * жауап: [ʒ ɑ w ɑ p] * жауға: [ʒ ɑ w ʁ ɑ]	ɕː Occurrences: 2 Examples: * ащы: [ɑ ɕː ə]
Fricative						χ Occurrences: 7 Examples: * хаты: [χ ɑ t ə] * халық: [χ ɑ l ə q] * хан: [χ ɑ n] * ханға: [χ ɑ n ʁ ɑ] ʁ Occurrences: 217 Examples: * туған: [t w ʁ ɑ n] * жауға: [ʒ ɑ w ʁ ɑ] * суға: [s w ʁ ɑ] * жонға: [ʒ o n ʁ ɑ]	h Occurrences: 1 Examples:
Approximant	w Occurrences: 166 Examples: * туған: [t w ʁ ɑ n] * жылуы: [ʒ ə l w ə] * жауап: [ʒ ɑ w ɑ p] * жауға: [ʒ ɑ w ʁ ɑ]			j Occurrences: 1,347 Examples: * еріне: [j e r ɪ n j e] * еңбек: [j e ŋ b j e k] * найза: [n ɑ j z ɑ] * жүре: [ʒ ʏ r j e]
Trill		r Occurrences: 632 Examples: * еріне: [j e r ɪ n j e] * орыны: [o r ə n ə] * жүре: [ʒ ʏ r j e] * бар,: [b ɑ r]
Lateral		l Occurrences: 519 Examples: * жылуы: [ʒ ə l w ə] * тіле.: [t ɪ l j e] * қылыш: [q ə l ə ʃ] * ала: [ɑ l ɑ]

Vowels#

Vowel symbols to the left of are unrounded and those to the right are rounded.

	Front	Near-Front	Central	Near-Back	Back
Close	i Occurrences: 45 Examples: * қисық: [q i j s ə q] * итке: [i j t k j e] * ит,: [i j t] * ине: [i j n j e]				u Occurrences: 1 Examples: * тыю: [t ə j u w]
		ɪ Occurrences: 744 Examples: * еріне: [j e r ɪ n j e] * үйің: [ʏ j ɪ ŋ] * күйің: [k ʏ j ɪ ŋ] * тісін: [t ɪ s ɪ n] ʏ Occurrences: 302 Examples: * жүре: [ʒ ʏ r j e] * үйің: [ʏ j ɪ ŋ] * күйің: [k ʏ j ɪ ŋ] * түссе: [t ʏ s s j e]		ʊ Occurrences: 110 Examples: * шұқыр: [ʃ ʊ q ə r] * ұрысы: [ʊ r ə s ə] * сұра.: [s ʊ r ɑ] * сұрап: [s ʊ r ɑ p]
Close-Mid	e Occurrences: 983 Examples: * еріне: [j e r ɪ n j e] * еңбек: [j e ŋ b j e k] * жүре: [ʒ ʏ r j e] * тіле.: [t ɪ l j e]				o Occurrences: 258 Examples: * досқа: [d o s q ɑ] * орыны: [o r ə n ə] * сосын: [s o s ə n] * қойды: [q o j d ə]
			ə Occurrences: 1,035 Examples: * орыны: [o r ə n ə] * атың.: [ɑ t ə ŋ] * қисық: [q i j s ə q] * пышақ: [p ə ʃ ɑ q]
Open-Mid
	æ Occurrences: 59 Examples: * әдет: [æ d j e t] * әйел: [æ j j e l] * әуесі: [æ w j e s ɪ] * мәлім: [m æ l ɪ m]
Open			a Occurrences: 1 Examples:		ɑ Occurrences: 1,758 Examples: * досқа: [d o s q ɑ] * найза: [n ɑ j z ɑ] * атың.: [ɑ t ə ŋ] * туған: [t w ʁ ɑ n]