Just got an update from Ancestry.com saying that the Scottish 1881 and 1901 censuses will be coming soon. So that completes the set-- "all" they need to do now is proofread it all




Regards,
Sarah
Moderator: Global Moderators
Although I suppose the OCR program should be much better at reading typescriptemanday wrote:The question is - are they using the 1881 typed version or the original documents?
If they are using the typed version, and still come up with some of the howlers from before, then we'll know they are using OCR
You'd think, but it depends on how much they paid for itSarahND wrote:Although I suppose the OCR program should be much better at reading typescriptemanday wrote:The question is - are they using the 1881 typed version or the original documents?
If they are using the typed version, and still come up with some of the howlers from before, then we'll know they are using OCROne can hope anyway!
Don't think they caught a Snark, though, did they?sporran wrote:Hello,
David Webster and I have received separate replies from Ancestry, and both have contained a statement that Ancestry do not use OCR in censuses. Please take notice.
This topic is getting to be like Lewis Carroll and "The Hunting of the Snark":
"What I tell you three times is true".
Regards,
John
That is indeed the case, but I still have the dilemma that I continue to see outputs that I just cannot believe are the result of the extensive training and considerable quality control that the same communication from Ancestry went out of its way to describe, if it is the case that human beings only are involved, not just in the occupation field, but the place of birth field.sporran wrote:Hello,
David Webster and I have received separate replies from Ancestry, and both have contained a statement that Ancestry do not use OCR in censuses. Please take notice.
This topic is getting to be like Lewis Carroll and "The Hunting of the Snark":
"What I tell you three times is true".
Regards,
John