A domain-specific language for describing machine learning datasets

Joan Giner-Miguelez, Abel Gómez, Jordi Cabot

Abstract: Highlights•Data issues in ML raise the community’s interest in building data best practices.•This work proposes a structured language to describe machine learning datasets.•The language allows describing composition, provenance, and social concerns of data.•A structured format eases the dataset comparison and the replication of ML results.•The language is supported by DescribeML, a VSCode tool to aid in its usage.