There exist several collections of spreadsheets, e.g. the EUSES corpus, the ENRON corpus and the FUSE corpus. Unfortunately, these corpora do not provide information about faults and expected output values. For evaluating spreadsheet debugging approaches however, we need spreadsheets with known faults and also testing decisions for simulating user interactions. Therefore, we provide  special corpora for evaluating spreadsheet debugging techniques:

Info1 Corpus

This is a collection of exercises submitted by civil engineering students as part of an Excel course.

Integer Corpus

This is an artificially created corpus containing spreadsheets with Integer values only.

Modified EUSES Corpus

This modified version of the EUSES corpus consists spreadsheets with artificially seeded faults. The testing decisions and expected values are obtained from the original EUSES spreadsheets.

Payroll/Gradebook Corpus

This corpus is an Excel version of Ruthruff et al. ‘s Forms/3 spreadsheets.

Each corpus has its own advantages and disadvantages. Details about the corpora are given in the corresponding subpages. The format of the properties files which are used to simulate a user’s testing decision are explained here.