Introducing Source Conflicts

This looks really well built out. I hope it's not just using exact matching on all fields. So many times a single source out of a handful will have a slightly different value, or even just a partial value instead of full value.

How does this interact with Source citation where the data is supported by the source, but the source doesn't list the exact value? I usually leave the data as null on those citations.

After processing about half of my source conflicts here are my findings:

* It's using a strict exact match. Profile field value of Minnesota, United States conflicts with Source citation value of Minnesota
* Location fields are not comparing the full location data. On places where I have a place name, or a street address that data is being ignored.

I don't like the idea of the use of ignored source when I've confirmed the data in the profile is correct. I may be ignoring the source, but in most of my cases so far I'm saying the data is not in conflict.

I am having some of the same issues as Jonathan Scott Krengel Almost every source that I confirm differs slightly - even census records so I am not confirming now. Kinda defeats the purpose.

I was taught by a curator here to write out "Spartanburg County" not just Spartanburg and it seems that if the source is from another platform that does not use the "county" distinction, it will red flag it. Spartanburg is a city and a county. Fields are not listing properly in some sources and now they red flag.

What happened to the ability to add profiles from sources? I wasn't able to do that when I have tried lately. That was a really usable feature as long as you commit to going back and adding sources and additional data.

I think Spartanburg is a county also

The idea of source conflicts is very good, but currently the implementation isn't very good or useful. Absolutely ALL profiles cause conflicts for birth place on confirmed matches against Swedish church records through MyHeritage/ArkivDigital, just because the priests wrote only the parish (and sometimes a very abbreviated version of the parish name) and not in which landscape or country the parish is located. Geni identifies ALL these as "conflicts" and suggests I remove all the details I have typed in. I got over 900 conflicts this week that I had to go over and "Ignore for all" on all of them. It would be really good to see REAL conflicts, but this is just a pain sitting hour after hour to click down all the "conflicts" which aren't conflicts. If I wouldn't have done this job now, I would never see any potential future REAL conflicts, and the whole inconsistency concept on Geni would be pointless.

I can't attach screenshots in this forum, but here are some examples:

"Norrköping, Östra Eneby, Östergötland, Sweden" is a typical birth place in Sweden, and it is entered as such on Geni. Norrköping is the city, Östra Eneby the parish (there are many parishes in Norrköping), Östergötland is the landscape and Sweden the country. The parish (entered in the county field on Geni) the central for genealogy in Sweden, and this is usually the only thing mentioned in the church records. But it is still very relevant which city, landscape and country it is located in, so all the fields are used on Geni. Sometimes parishes in different cities even have the same name.

The "source conflicts" on Geni compare "Norrköping, Östra Eneby, Östergötland, Sweden" strictly with the church records, which most of the time say anything like the following:

Östra Eneby
Ö. Eneby
Ö Eneby
Norrk. Ö Eneby
Norrk. Ö. Eneby
Ö. Eneby Österg.
Östra Eneby Östergötland

All of the above correctly describe the same place, but none of them is identified by Geni as the same place.

Another example:

"Linghem, Östra Skrukeby, Östergötland, Sweden" is a correct birth place entry on Geni, but in church records it may appear as follows:

Östra Skrukeby
Ö. Skrukeby
Ö Skrukeby
Ö. Skrby
Skrby
Ö.S.

All are correct entries hand written by a priest maybe 100-200 years ago, but none of them is identified as matching by Geni.

In some master profile I even saw a conflict with a profile listed with birth place "Budapest, Hungary" as not matching the birth record, which just said "Budapest".

So of all the around 900 inconsisencies around 850 were identified as conflicts even though the birthplace actually is the same on Geni as in the church record, just written in a shorter way by the priest. I don't think there is any automatic algorithm that can identify all possible abbreviations. Some parishes like "Västra Vingåker, Södermanland, Sweden" are often listed as just "V Vingåker" or even "V. V." or "W. W." in church records.

For all of these I just had to click "Ignore for all". All sources were attached to the right profile AND include correct information but not written in a complete enough way in the church records. I would absolutely not like to delete the source or change the birth place in the Geni profiles as everything is correct, not just identified as such by the Geni algorithms.

Then there were maybe 50 of my 900 conflicts that had the wrong birth place in the source. This is because the priests made mistakes. This often happened when a large family moved multiple times. After a move the priest in the new parish sometimes assumed all of them were born in the same place even they weren't or just made a mistake for other reasons like from reading the wrong line in a document. There is no doubt that the source shows the right family, because all the names, all the birth dates and most of the birth places are correct. So the source is connected to the right profile and includes a lot of correct information but just some errors in the details. Birth records are the primary source for birth places, but Geni/MyHeritage doesn't yet match with Swedish birth records, only household examination books.

Just in 2 cases of all my around 900 conflicts I found any information at all that was actually missing on the Geni profiles. Those profiles were in an American branch of our family and didn't have birth place entered on Geni at all.

So the big problem is different ways of writing birth places. Currently a lot less than 1% of the conflicts are real conflicts, so showing them does more harm than good to me right now.

Another problem is that lots of inconsistencies remain as orange exclamation marks in tree view after ignoring them. That must be a technical bug, and I have reported a ticket for that.

Niclas Carlenius where is the "ignore all" tab or button? That would be soooo much easier than what I am finding myself doing. Maybe a setting to just ignore them all.

Susanne Floyd I can't see any "ignore all" (to ignore all inconsistencies), just "ignore for all" for each conflict, so that specific conflict is ignored for all users.

Oh well, that would be good if we could just do that. I am worn out with them and as you note, not finding much useful. More that is not as refined or just a "hair off" the field. Thanks Niclas Carlenius.

Slightly off topic: Another problem with inconsistencies is that some of them are for claimed user profiles, where only that user has the right to "ignore for all". Most users are quite inactive and don't care or don't understand (even if I explain). Right now my only two inconsistencies are for two merged user profiles. One "has a child with a different last name", and the other one "has no birth surname". The different last name is correct (she got married after her child died). The missing birth surname is the same as the current surname and could be added if I had editing rights. I wish inconsistencies for user profiles would only be shown to that user and not to everyone in their family group who has no editing rights. I keep checking every day if there is a chance they reach their "8 months of no log in limit", so they can be declared inactive managers and I can edit. The same goes for the merge issues, which I have for these same two user profiles only (they had two different user profiles that got merged and now have one small data conflict each). In the past curators could help with that, but unfortunately their editing rights for private profiles were removed (unless the user him-/herself asks the curator, which they wouldn't care enough to do).

Is there any chance that Geni could change so that inconsistencies and merge issues could be shown only to users who actually have the editing rights to actually do anything about them?

Mike Stangel here is my great-uncle Farrell (Bud) Bergquist

Plenty more where that came from.

6 Inconsistencies - 6 Source Conflicts
Looked to see what the Source Conflicts were - then looked to see what the Inconsistencies were -- and the 6 Inconsistencies consisted of the 6 Source Conflicts
Seems misleading to me -- to my mind, either they should be separate categories, or they should be treated the same as the other categories in Inconsitencies -- or maybe say "X Inconsistencies including Y Source Conflicts"

The fact is that many settlements on the territory of the Russian Empire, the USSR and the post-Soviet republics have repeatedly changed their names and geographic referencing. So there will be discrepancies and inconsistencies here, as well as changes in the writing of personal names and surnames. Source Conflicts when building my tree are predictable.

I don't mind a "Source Conflict" if it is real. But when I am having to look at

Death on profile - Marion County, South Carolina, United States

Death on source - Marion, South Carolina, United States

I think it is overkill. Or if it is

Death on profile - Marion County, South Carolina, United States

Death on source - Blank or South Carolina, United States

That is a waste of my time to check. I am starting to not want to add even census sources on a profile because of conflicts and inconsistencies.

Another issue with the current implementation is that if the profile does not have a value in a field but a source does, and the profile is correct, there is no way to ignore the conflict because you can't select the empty value in the profile as the correct value.