search

DeveloperLiberationFront / Spreadsheet-Corpus-Paper

1 stars 0 forks source link

Papers debating the representativeness #20

Open Felienne opened 10 years ago

Felienne commented 10 years ago

Moving this here and out of the paper.

Has someone else raised public concerns about the representativeness of EUSES? Currently, the most used corpus for spreadsheet analysis is the EUSES corpus. This corpus contains spreadsheets obtained mainly by searching in search engines, so it contains spreadsheets that are publicly available. However, industrial case studies performed by us and other researchers have raised questions about the representativeness of this corpus~\cite{something}, as spreadsheets we encounter in industry tend to be bigger and be more complex. Hence, we believe there is a need for a corpus providing a more realistic test set for spreadsheet researchers.

Felienne commented 10 years ago

I have in one of my papers (I think ICSE 11) stated that Euses differed a lot from what I saw in practice. Does that count?

CaptainEmerson commented 10 years ago

I'm a little hesitant about criticizing EUSES directly -- I'd rather do it indirectly, and let the reader come to his or her conclusions. If there's another paper that's critical, it should come from someone other than us. :)