Loading datasets...
License:
NOODL-1.0
Steward:
Institute of African Digital HumanitiesDataset ID:
cmk1bdl0q39v6mb07gee6udp2
Release Date: 1/5/2026
Format: MP3, WAV, TSV
Size: 3.44 GB
The dataset consists of paired audio and text resources on spoken French from the Republic of the Congo. The audio files were extracted from longer recordings of semi-guided interviews conducted in Brazzaville, and orthographic transcriptions were added. The long audio recordings and their corresponding TRJS transcription files were automatically clipped alongside their respective transcriptions. The dataset comprises ten folders containing audio files and ten audio/text mapping files.
Licensing
Nwulite Obodo Open Data Licence 1.0 (NOODL-1.0)
https://licensingafricandatasets.com/nwulite-obodo-licenseRestrictions/Special Constraints
You agree :
Forbidden Usage
You agree not to use the data for:
Intended Use
The audio-text alignment in this dataset enables speech recognition models to be trained or evaluated for the development of a more inclusive and representative ASR model for French.
Global linguistic governance does not formally recognise Congolese French as a language or language variety. The locale 'fr-CG' (Congolese French) used in this dataset is an attempt to reclaim the term 'Congolese French' for this variety of French. Despite the French variety used in the Congo not having formal recognition, people from the Republic of Congo who speak French can be identified by French speakers from other countries, particularly in Africa, as speaking a specific variety of French. In some cases, it may even be possible to identify someone as Congolese on the basis of their accent and/or linguistic repertoire. Numerous documentation and description efforts have been carried out by researchers in local universities and abroad with regard to the Congolese variety of French.
Although Congolese French may be subject to internal variation, there is insufficient information available to substantiate the existence of particular sub-varieties or their peculiarities. Nevertheless, the audio recordings in this dataset were made in Brazzaville with people who had lived there long enough to recount the origins of their settlement.
The alphabet used in the transcription of audio recordings is that of standard French: a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z.
This dataset was compiled from transcriptions of audio recordings made by a research group at the Université Marien Ngouabi in Brazzaville as part of a project led by Professor Edouard Ngamountsika. Researchers conducted audio recordings in some parts of Brazzaville using prompts such as 'How did you come to settle in this place?'. The recordings were then transcribed using a multimodal annotation framework, including indications of silence and noise.
The dataset consists of spontaneous spoken Congolese French collected in urban settings, primarily through informal sociolinguistic interviews. The texts cover everyday conversational interaction, interview discourse, personal and situational narration, and metacommunicative speech related to the recording process.
Total size is 3,27 GB
This dataset comprises audio clips and audio/text mapping files. There are 13,344 audio clips totalling 6 hours, 8 minutes and 12.286 seconds, as well as 44 audio/text mapping files totalling 13,346 lines.