Resumen de The OpenScore Lieder Corpus

Mark Robert Haigh Gotham, Peter Jonas

• The OpenScore Lieder Corpus is a collection of over 1,200 nineteenth century songs encoded by a dedicatedteam of mostly volunteers over several years. Having reported on the initial phase, motivations, design, andcommunity-oriented aspects of the project before [3], we present here the first, stable, large-scale release ofthis corpus specifically designed for MIR researchers, complete with comprehensive, structured, linked metadata. The corpus continues to be available under the open CC0 licence and represents a compelling datasetfor a range of MIR tasks, not least given its unusual balance of large-scale with high-quality encoding, and ofdiversity (songs by over 100 composers, from many countries, and in a range of languages) with unity (centredon the nineteenth-century lieder tradition)