Another WIP notebook in need of additional documentation… This one explores the stats around volunteer correction of OCR errors in @TroveAustralia’s newspapers. More to come!
@wragge This is all fascinating. I'm particularly interested in Output 24, which (unsurprisingly) shows that the correctors gravitate toward family notices. But the other categories that show a high proportion of corrections is unusual. (Also, what are “Detailed Lists, Results, Guides”? Things like sports scores?)