Common Voice Scripted Speech 25.0 - Adja

ajg — Adja (ajg)

This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Adja [ajg - ajg]. The dataset contains 19398 clips representing 12.87 hours of recorded speech (12.71 hours validated) from 23 speakers, recorded from a text corpus of 3,213 sentences.

Language

Adja (Aja) is a Gbe language spoken in Benin, Gabon, Ghana, Nigeria and Togo by 1M+ people. It is part of the Kwa branch of the Niger-Congo family. The language is used in daily communication, oral literature, and traditional expressions.

Demographic information

The dataset includes the following self-declared age and gender distributions. A coverage summary is shown below each table.

Gender

Self-declared gender information. The table shows clip and speaker counts with percentages. Speakers who did not declare a gender are listed as Unspecified. A dash (-) indicates zero.

Gender declared: 0 of 19,398 clips (0.0%), 0 of 23 speakers (0.0%)

Age

Self-declared age information. The table shows clip and speaker counts with percentages. Speakers who did not declare an age are listed as Unspecified. A dash (-) indicates zero.

Age declared: 3 of 19,398 clips (0.0%), 1 of 23 speakers (4.3%)

Data splits for modelling

Clip buckets

Training splits

Training split coverage: 3,207 of 19,158 validated clips (16.7%)

The dataset contains 19158 validated, 218 invalidated, and 22 unresolved clips. The average clip duration is 2.389 seconds.

Text corpus

Validated sentences: 3,213

The corpus contains 3,213 sentences: 3,213 validated and 0 unvalidated (0 pending review, 0 rejected), with 0 reported for review.

Writing system

Aja is a tonal language, typically with three level tones:

• High (´)

• Mid (unmarked)

• Low (`)

The orthography used in this dataset aligns with conventions adopted by Aja linguists and educators, particularly those documented by AGOSSOU Kogbédji Émile, PhD, and Kansiwer, PhD. Minor regional variations are harmonized through community consensus.

Symbol table

a an b c d d e s en f g gb y h hw i in j k kp I m n ny n o 5 on p r s sh t u un v w x xw y z 3

Sample

There follows a randomly selected sample of five sentences from the corpus.

1. Toxudekanono.

2. E ji vi.

3. Jeshi.

4. Wo nyomesoso ce yi ado wo ji.

5. Yovogbu amewo don.

Sources

• AGOSSOU K., Le grand livre des proverbes aja, Azové, Éphphata, 2011 (primary source).

• AGOSSOU K., Proverbes aja et hébreux : une étude littéraire, in Le Sycomore, 2014.

• AGOSSOU K., L’emphase dans l’hébreu biblique : identification et méthodes de traduction en langue aja, in Mélanges en hommage au professeur BADA, 2022.

• AGOSSOU K., La situation de la femme aja dans ses relations personnelles avec l’homme dans le mariage : l’égalité en question, in Mélanges en hommage au professeur HUANNOU, éd. Plumes Soleil, 2024.

In addition, the dataset includes sentences from:

• Common perceptions and everyday life expressions collected during fieldwork.

• Crowdsourced contributions supervised by the curators.

Fields

Clips

Each row of a tsv file represents a single audio clip, and contains the following information:

• client_id - hashed UUID of a given user

• path - relative path of the audio file

• text - supposed transcription of the audio

• up_votes - number of people who said audio matches the text

• down_votes - number of people who said audio does not match text

• age - age of the speaker1

• gender - gender of the speaker1

• accents - accents of the speaker1

• variant - variant of the language1

• segment - if sentence belongs to a custom dataset segment, it will be listed here

• prompt_upvotes - number of upvotes the sentence prompt received

• prompt_reports - number of reports the sentence prompt received

• is_edited - whether the clip's transcription has been edited

validated_sentences.tsv

The validated_sentences.tsv file contains one row per validated sentence in the text corpus:

• sentence_id - unique identifier for the sentence

• sentence - the sentence text

• variant - the variant of the language

• sentence_domain - the domain(s) the sentence belongs to

• source - the source the sentence was collected from

• is_used - whether the sentence is still in circulation for recording

• clips_count - number of clips recorded for this sentence

unvalidated_sentences.tsv

The unvalidated_sentences.tsv file contains one row per unvalidated sentence in the text corpus:

• sentence_id - unique identifier for the sentence

• sentence - the sentence text

• variant - the variant of the language

• sentence_domain - the domain(s) the sentence belongs to

• source - the source the sentence was collected from

• up_votes - number of upvotes the sentence received

• down_votes - number of downvotes the sentence received

• status - current status of the sentence (pending or rejected)

Get involved

Community links

• Common Voice Aja (ajg) page (link to be activated when available)

• Common Voice translators on Pontoon

• Common Voice Communities

Discussions

• Common Voice on Matrix

• Common Voice on Discourse

• Common Voice on Discord

• Common Voice on Telegram

Contribute

• Speak

• Write

• Listen

• Review

Acknowledgements

Datasheet authors

• Dr. Kogbédji Émile AGOSSOU (Université, PhD)

• Dr. Kansiwer (PhD)

• Justin Bakoubolo (PhD)

• Justin Bakoubolo — supervised the compilation of the dataset.

• Justin Bakoubolo

Funding

This dataset was partially funded by the Open Multilingual Speech Fund managed by Mozilla Common Voice.

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.

Footnotes

1. For a full list of age, gender, and accent options, see the demographics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2 ↩3 ↩4