Opening the Well is a crowdsourcing platform where volunteers are transcribing Gaelic language recordings. Together, we are building a corpus of transcribed speech which will enrich Scotland’s intangible cultural heritage, bolster Gaelic research and lexicography, and improve speech recognition technology in Gaelic.
Opening the Well builds on the successful ÈIST (Ecosystem for Interactive Speech Technology) initiative by advancing Gaelic automatic speech recognition (ASR) and expanding the language’s corpus and machine learning datasets. By establishing a dedicated crowdsourcing platform within the Tobar an Dualchais (TAD) ecosystem, the project enables large-scale transcription contributions from Gaelic speakers — resources that will enhance learning and accessibility for Gaelic learners and hearing-impaired users worldwide.
Inspired by Ireland’s Meitheal Dúchas.ie model, the platform empowers the Gaelic-speaking community to transcribe Gaelic oral narratives through an ASR-assisted interface. The resulting corpus will enrich Scotland’s intangible cultural heritage, bolster Gaelic research and lexicography and improve ASR performance.
Opening the Well is part of the ÈIST research project, which is funded by the Scottish Government.