How can we prevent importing more duplicates of the British Royal family?

Question

Or, for that matter, other well-known European dynasties?

Asking this because two new merges for duplicates of Queen Elizabeth II (plus merges for members of her family) have just been found. A few days ago many merges were set up for kings of Denmark and their families, also following a gedcom import with very few new profiles in it.

I understand the Gedcom import tool is designed to find duplicates and direct the user to skip them, the problem with royals is that their family name is not such a well established concept; so, it's much more difficult for a program to pick duplicates, because of the different possible LNABs.

The question is if, or where, a message could be added telling people, before they download their Gedcom, that European Royal families (or other very notable families) are probably already on Wikitree and that they should double-check that the family is not already there before importing? (Maybe it is already there and I haven't seen it. I never used the Gedcom features).

Thanks !

 

Comments

You're assuming that people know who they've got in their gedcom.

But many people have gedcommed in famous figures without knowing who they were and/or without knowing they had them.

---

Right, I hadn't thought of that.

The case is hopeless then, isn't it?

Answer 1

We're looking at entirely changing the way GEDCOMs are currently imported.  It's a work in progress at the moment so I don't have a time frame on when it would be implemented, but it's a change that should prevent things like this from happening.

Comments

That's GREAT news, Eowyn!

---

Yes, great news !!

---

If Gedcoms are permitted at all, they should be restricted to the submitter and persons no more distant than 3rd cousins.  Or some such relationship.

Answer 2

Gedcom downloads continue to be the nemesis of the WikiTree system. Someone has to make a decision between growth or accuracy; otherwise it shall remain a never ending annoyance.

Comments

One of the problems, I'm sure, is that people who upload Gedcoms, don't do the checking after the upload, and either don't understand about 'skipping' duplicates, or just cannot be bothered, and hit the OK button.

Maybe there should be a percentage where the Gedcom is discarded when possible duplicates are detected.

---

It may come to that if the powers to be are willing to address the problem; which I doubt as growth of the system is the main objective. You can not have faith in people always doing things properly. Automated detection systems are inherently plagued with problems as well. If member education does not work, I am afraid nothing will.

Answer 3

Suggestion: WT should block creation of pre-1700 profiles from gedcom imports. Most pre-1700 profiles from gedcom imports are probably either going to be duplicates or incorrect.

Comments

Actually, the Gedcom help page says pre-1700 profiles are automatically skipped already. I agree that most pre-1700 American profiles are likely to be duplicates, and probably not only in America, but there are some areas of the world where this is far from being the case. There is hardly no way to allow import of gedcoms from under-represented countries, because many people import profiles with hardly any locations, or names of places without country, etc.

---

That's correct, that pre-1700 individuals are automatically skipped.

---

It's simple enough.  Pre-1700 ancestors of new American users are likely to be duplicates.  Pre-1700 ancestors of non-American users, much less so.

That might have to be American+Canadian.

---

Case in point: My family is from Bohemia, records go back in some cases to the late 16th century - and basically nobody is already on WikiTree.

---

When you count the work required in fixing a Gedcom profile after it has been created, I'm firmly convinced it takes less work to create a profile from scratch.

Answer 4

We have a pre 1700 qualification, an honor code, etc.  Perhaps we should have a lesson and quiz before a member is approved to upload a Gedcom.

Everyday there are questions from new members about Gedcoms, it appears they have not looked at the Help section for Gedcoms.  

Maybe an instruction that before you upload a Gedcom you must have entered 5 profiles in good form manually. Second, Gedcoms might be broken down into smaller files.  And require that the clean-up of each upload be completed before uploading another.

These restrictions might discourage folks who have 5,000 on their tree from starting, but those who are serious and not just interested in numbers would continue and fewer duplicates and errors would exist,

Answer 5

I'm not sure if this is doable, but maybe:

Make a list of incorrect names of commonly duplicated people.

So if people keep importing Sachsen-Coburg_und_Gotha-4 as King George VI and the duplicate finder isn't catching it, add "King George VI" to the list. When someone uploads a gedcom, the normal duplicate checker can be run, and then the person can be compared against this special list. If there is a match to a name on the special list, don't allow the profile to be imported.

Comments

The problem is that there are lots of possible LNABs and it's sometimes hard to figure out what people will come up with. The latest - or one of the latest - set of imported duplicates had "d'Angleterre" as LNAB. And the profiles were all private... hats off to the Euro Aristo folks who managed to FIND the profiles !

There was also an import of a big chunk of the Danish dynasty - with the LNAB Unknown (which makes sense, in a way).