Why am I getting lots of tree matches when the named person has different parents?
If the name is the same but the parents do not match then it is not a Tree Match, I have to go through each match flagged up to check them and it takes time.
Why am I getting lots of tree matches when the named person has different parents?
If the name is the same but the parents do not match then it is not a Tree Match, I have to go through each match flagged up to check them and it takes time.
Am also experiencing this problem, very irritating as wastes time. Should have a way of forcing a 'No Match' condition when this occurs.
I would like to have the settings for tree matches editable by Myself !
It would seam that for now this is set at a match in the name only.
This needs to be changed to match and i think that this may be suitable:-
Name plus the following;-
1- Parents names
2- If there are no parents then the name of the spouse
3- If "2" & "3" are null then use children's names
What do you think?
Possibly also taking the Birth Location in to account as well.
The problem I experienced was where the subjects name were the same, there was no birth date in both cases, but birth location was given for one side, and for both sets of parents, the names were the same, but the sibling (brother) names differed. And the system insisted that it was a possible match.
I think there must be some kind of options for the tree matches, decided by me as the user:
In Sweden before 1894, the children got their last name from the fathers first name with the ending -son or -daughter, so a match of the fathers first name make no sense, (all Sven Larsson has a Father named Lars) it must be a match of the fathers last name + Mothers name (last or first) if there should be a chance of getting the matches being correct.
This is of course not the case in country's where the child would inherit their fathers last name. - Thats why I need to decide myself how to get matches, or the matches should concider "country of birth"
Also there is a lot of matches today that have the "estimated birth between 1670 and 1730", but if my ancestor is born in 1850, i still get the suggestion that it is a match. I think it is great that people without a year of birth gets this estimate, but it must be considered for the matches - only match if the birth year is within the estimate.
Ivan
The problem with using a place name is that the name can change with time.
We are supposed to use the place name as it was at the time of the event.
You can have a person born in 1910 in Presony, Hungary and die in 1930 Bratislava, Czecoslovakia (today the Slovak Republic)
You have to take into account all the National boundary changes and with that place name changes.
England changes to GB in 1600 and GB to UK in 1820
I also have to contend with New England / USA.
Randolph
I cannot accept your argument, as the place names can always be stated in form 'old name/new name'.
I am working with entries from what is variously referred to as Old Austro-Hungarian Empire / Kingdom of Serbs, Croats and Slovenes, Kingdom of Yugoslavia, Socialist Republic of Yugoslavia or Republic of Croatia, so I understand what you are worried about, but IMHO it's not a problem.
What about it? Wikis are 'guidelines' and can be updated, please look at the issue of solving the problem, you are deviating from the original issue.
In the case of the problem that I am experiencing, if the algorithm handling matches, were to use the location as a sub-criteria in order to decide if there were a possible match or not, then there would be no erroneous match, but because it is not, it is insisting on matching two non related individuals who were born in different locations (different villages, never mind different countries).