Duplicate Merging as a Service
"The FamilySearch shared tree uniquely strives to have just one public profile for every deceased person who has ever lived."
I love this principle and its implementation! However, duplicates in the World Tree abound. I was wondering if it is possible to review and merge possible duplicates as a service to the World Tree? I have often wondered what happens when ordinances are performed twice or even thrice for duplicates; or when they are performed on profiles who are really two different people merged in one (I've seen it happen! in fact I have been defending one profile in particular for long...).
I was also wondering separately, if anyone knows, what programming language would be best to write in for using the API and making programs to help with research? To my understanding, JSON requests are used to fetch data from the tree. Thank you for your time in responding!
Comments
-
In my opinion it would be a great service to merge duplicates, but feel generally it's best to stick to your own family or to a limited area in which you have a particular expertise and good experience working in Family Tree, and in which there is sufficient documentation to support the merge. I would also suggest than all merging needs to be supported with sources. So not just a merge and nothing else, but really have research behind it will full analysis of the family involved.
2 -
I am suspicious of how an automatic process would determine that one set of information is more valuable than another. Example, when was person has birth 1856, death "deceased" and the other has birth 21 Dec 1857, death 5 March 1904, would the more detailed dates be kept? Would both be kept as alternate info? And if people warbled about their age every census, would all those birth years be kept? Of course I had no ancestors who did this ... (rolling eyes).
Something that contributes to this problem is that a user "FamilySearch" appears to be creating a whole new family set for lots of birth records, or other records. THAT has caused most of my merges, to be honest. #2 is when I search for parents as a part of adding them, "none found" shows so I create new records. Then, miraculously, the notice of possible duplicates appears. Sigh.
I'm not sure we want system-merging, but I don't want to invoke an argument.
1 -
I am sure Fernando was not requesting an automatic process, just if there was a way for volunteers to just sit and work on manually merging duplicates. Automated merging, as you suggest, is just asking for a disaster.
I do want to clarify one thing, I'm pretty sure that if you check those individuals "contributed" by "FamilySearch" you will find that they all have creation dates in the spring of 2012. This was the initial population of Family Tree from the older databases including the IGI which contained millions of names. In the IGI, people were not linked in extended families. Each line in the IGI consisted of a single individual, a single individual linked to one spouse, or a single individual linked to parents. That is the form they came into Family Tree.
This means that that a family consisting of five children, all married, would have these IGI entries:
- Child 1 with mother and father
- Child 1 with spouse
- Child 2 with mother and father
- Child 2 with spouse
- Child 3 with mother and father
- Child 3 with spouse
- Child 4 with mother and father
- Child 4 with spouse
- Child 5 with mother and father
- Child 5 with spouse
That means two copies of each child, one copy of each spouse, five copies of the mother, and five copies of the father. If twenty different relatives had each submitted this family to the IGI, then multiply those numbers by twenty. (My gg-grandfather was in the IG over a hundred times).
The actual submitters to the IGI and other databases were not transferred into Family Tree but the contributor name had to have something so they put in "Family Tree."
1 -
@FernandoAlexanderAM you are in good company. On Family Tree there are many contributors who spend most of their time doing merging (else splitting). I am one.
One way you can start is to choose a surname in your own ancestry. Use Find and Exact Match on the surname to list them all, and look for profiles with no father. Those profiles are likely to be duplicates.
1 -
@Gordon Collett You are right about the 2012. I was half way through merging the parents yesterday and still have the 2 fathers to merge, so I could check. Yes, the independent family group was created 2012.
0
