Common Voice Scripted Speech 25.0 - Bankon

Ɓàŋkón — Bankon (abb)

This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Bankon [Ɓàŋkón - abb]. The dataset contains 7922 clips representing 11.24 hours of recorded speech (11.19 hours validated) from 22 speakers, recorded from a text corpus of 996 sentences.

Language

Bankon, also known as Abo, is a Bantu language spoken in the Moungou Division of Cameroon's Littoral Region. Bakon is closely related to Barombi, also known as Lombi, Balombi, or Barumbi, which is spoken in the Meme Division of the Southwest Region of Cameroon. Therefore, a striking trait that distinguishes the two speech groups is not their intrinsic typological distance, but the colonial demarcation between the French-speaking and English-speaking parts of Cameroon. This is a direct consequence of the joint trusteeship that France and Great Britain held over Cameroon.

Variants

The language representatives who contributed to this dataset described Bankon and Barombi (also known as Lombi, Balombi, or Barumbi) as two closely related varieties. However, each is recognized as a distinct language in scientific and administrative nomenclatures. For example, the Administrative Atlas of Cameroon's Languages (Breton and Bikia Fohtung, 1991) recognizes them as such. These representatives added that there are two varieties of Bankon: South Bankon and North Bankon. According to the representatives, however, there are only minor lexical and phonetic differences between these two varieties. Speakers of each variety designate the other variety as Mwi Nwan, wich translates as "the other bank of thhe river". Additionnaly, speakers of South Bakon refer to North Bankon as Bankon Ba Ibom, or "Upper Bankon". Conversely, the speakers of North Bankon refer to South Bankon as Bankon Ba Nlon*, which translates as Lower Bankon. Other less common designations of North Bankon and South Bankon are Banaka and Bayeme, respectively.

Both the Southern and Northern dialects of Bankon are represented in the collection of sentence prompts, with a slight bias towards Southern Bankon.

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 7,922 clips (0.0%), 0 of 22 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: 20 of 7,922 clips (0.3%), 0 of 22 speakers (0.0%)

Data splits for modelling

Clip buckets

Training splits

Training split coverage: 996 of 7,887 validated clips (12.6%)

The dataset contains 7887 validated, 35 invalidated, and 0 unresolved clips. The average clip duration is 5.109 seconds.

Text corpus

Validated sentences: 996

The corpus contains 996 sentences: 996 validated and 0 unvalidated (0 pending review, 0 rejected), with 0 reported for review.

Writing system

The collection of sentence prompts provided by the language representatives aligns with the General Alphabet of Cameroonian Languages

Sample

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

1. Imbūs wɔ̌n ŋgandâk, mùleèdì a těm lɔ̀.

2. Ǹsɔn mwaŋ i kisùkulù kì mu sòs.

3. Ìwûmsɛ̀ di ŋgandòmbu nik, Mìsɛ ǎ kɛ̀ imbɔ̄k nyǎŋ nu mboŋ ì ìwùùsàgà i Kribi.

4. I pōnda yi maŋgòlò mbyaagɛ yi pǎn iɓaawàk di ŋga᷇n.

5. Yɛ̌ɛ̀ ǎ yɛ̄w mijɛ, ɓɔ̀ŋ saŋmbɔ̄k ndī nu pèla sɛ̄lɛ̄ nyu.

Sources

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 translators on Pontoon

• Common Voice Communities

Discussions

• Common Voice on Matrix

• Common Voice on Discourse

• Common Voice on Discord

• Common Voice on Telegram

Contribute

https://commonvoice.mozilla.org/abb

• Speak

• Write

• Listen

• Review

Acknowledgements

Datasheet authors

Emmanuel Ngue Um , Thomas Eyoum Ndando , Emmanuel Ndando , David Miyenga Miyenga

Citation guidelines

Ngué Um E, Ngo Tjomb EEC, Dibengue FL, Banum Manguele BM, Abo Djoulde B, Nyambe MA, Atangana Eloundou BM, Ngami Kamagoua JS, Mpouda Avom J, Nyobe Z, Eloundou Eyenga EG, Likwai AP (2025) Speech Technologies Datasets for African Under-Served Languages. Proceedings of the Eight Workshop on the Use of Computational Methods in the Study of Endangered Languages, edited by Lachler J, Agyapong G, Arppe A, Moeller S, Chaudhary A, Rijhwani S, Rosenblum D. URL Association for Computational Linguistics (ACL).

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