Abstract: This paper presents DECO (Dresden Enron COrpus), a dataset of spreadsheet files, annotated on the basis of layout and contents. It comprises of 1,165 files, extracted from the Enron corpus.