Suggestion re AI problem with visitors & Lodgers
I have been trying your "AI Research Assistant" a couple of times a week. Sometimes it is very helpful. However, I have found a suggestion that can cause problems.
For the 1871 England Census the family of John Cuthbert LJG3-L3M and Jane Cuthbert (nee Woodcock MF2S-R6R shows them with 2 of their daughters plus a married couple of lodgers and their son John Jones. The AI Research Assistant suggested that John Jones was the son of the Cuthberts, since the word 'son' was after his name. I dismissed the suggestion, as I checked the original on Find My Past.
However, there are many households in the UK Census documents which have more than one family (whether as lodgers or visitors). Is there any way the AI can be refined to note that a son or daughter may be the child of a lodger or visitor?
A new researcher may have erroneously added John Jones to the Cuthbert family instead of realizing he was the son of the Jones couple who were lodging with them.
Thank you.
Answers
-
@GrahamJanet3 - there appears to be a Community Group specifically for this sort of question - we are told that monitoring of those groups tends to be closer and engineers may participate.
I would say that you raise an interesting question. My first response is, as always, that researchers should check the image every time and think - as you have done. If the researcher can't see the image (eg because it's behind a paywall) and there appears to be an oddity (as evidenced here by the clash of surname - for which there are several possible explanations) then in my opinion, they need to get out of those profiles and leave them to those who can see the images. No-one should imagine that they are obliged to follow a hint.
As for the exact circumstances, like you, I have seen all sorts of blended households. For instance, there might be a married daughter in the household with her own children, who are described as "son" or "daughter". These will be the grandchildren of the head of the household. It doesn't help that there is often a slight lack of consistency on the original pages.
Maybe the only safe way is to detect the break of family and tell the researcher that they're on their own after the break of family!
1