Kallaama: A Transcribed Speech Dataset about Agriculture in the Three Most Widely Spoken Languages in Senegal
Elodie Gauthier, Aminata Ndiaye, Abdoulaye Guissé
TL;DR
Kallaama tackles the scarcity of speech resources for Senegal’s national languages in agriculture by releasing a 125-hour transcribed agricultural speech dataset across Wolof, Pulaar, and Sereer, complemented by web-derived text corpora for Wolof and Pulaar and a Wolof G2P lexicon. The dataset targets Automatic Speech Recognition and related NLP tasks, enabling voice-based services for farmers with limited literacy and facilitating language technology development in under-resourced languages. The work documents data collection methods, dialect considerations, and challenges in transcription and orthography, while providing open licenses and data formats to ease integration into ASR pipelines. Overall, it lays a foundational resource for language-inclusive AI in agriculture in Senegal and invites broader expansion to cover the three languages more comprehensively.”
Abstract
This work is part of the Kallaama project, whose objective is to produce and disseminate national languages corpora for speech technologies developments, in the field of agriculture. Except for Wolof, which benefits from some language data for natural language processing, national languages of Senegal are largely ignored by language technology providers. However, such technologies are keys to the protection, promotion and teaching of these languages. Kallaama focuses on the 3 main spoken languages by Senegalese people: Wolof, Pulaar and Sereer. These languages are widely spoken by the population, with around 10 million of native Senegalese speakers, not to mention those outside the country. However, they remain under-resourced in terms of machine-readable data that can be used for automatic processing and language technologies, all the more so in the agricultural sector. We release a transcribed speech dataset containing 125 hours of recordings, about agriculture, in each of the above-mentioned languages. These resources are specifically designed for Automatic Speech Recognition purpose, including traditional approaches. To build such technologies, we provide textual corpora in Wolof and Pulaar, and a pronunciation lexicon containing 49,132 entries from the Wolof dataset.