Place Name Testing

Gordon Collett

I keep bringing up Norwegians because it’s just my wife and I researching her family and I have dozens of relatives working on my family. So my examples will come from there again. This time I will be stressing the quality checking for places. I’ve touched on this a bit but wanted to take a close look. I expect the same situation will arise for other countries.

First some background information to set the stage.

The Places database for Norway is very incomplete. As I have worked in various municipalities I would say that at best about 50% of the needed places within any municipality are included in the database. Of the places that are included, about 75%, as a rough guess, are entered incorrectly in that the place name does not include the municipality, just the place and county. This results in three types of standards in the database:

• Place, Municipality, County, Country
• Municipality, County, Country
• Place, County, Country

I am going to test out each of these situations for place names and see what the Score Details show.

Place, Municipality, County, Country

https://www.familysearch.org/tree/person/details/K2T1-FBK

All his places names in Vitals had appropriate values in the Places database. So everything is fine here, as expected.

Municipality, County, Country

https://www.familysearch.org/tree/person/details/LZGB-R7X

Here, her birth place of Sjørsand, Fusa, Hordaland, Norway, does not exist yet in the Places database and can only be linked to the less complete Fusa, Hordaland, Norway. The scoring routine is looking only at the linked standard and is ignoring that the actual place name does have four parts.

Place, County, Country

I created a temporary person that I have deleted.

Here, the person was born at Hatlestad, Lærdal, Sogn og Fjordane, Norway. Instead of keeping the full version, I used the version currently in the Places database which is just Hatlestad, Sogn og Fjordane, Norway. The routine cannot tell that what is missing is not the “City/etc” but rather the 3rd level of municipality.

If I put in his full birth place, I get the same notice:

Again the routine cannot tell what is actually missing from the place name and cannot tell that the actual place name really does have four parts.

It will takes years to get the Places database for Norway complete. In the meantime, it would be nice if the Quality Score routine could look at the full place name along with the linked standard.

From the first two examples, it looks like the routine does not yet compare profile place names to the tagged index sources.

MWL1

The same applies to our Hungarian ancestors. I usually see only three jurisdictions in places.

roberthparker3

I have created TF-3305 to create two new segments:

• Norway 1800-1899
• Hungary 1800-1899

I'm thinking we can only require three place jurisdictions for these segments while we wait for the place database to improve.

I have also created TF-3306 to consider the suggestion to take the original place into account.

AnneLoForteWillson

So I do have a question on the Places database and its connections. On my ancestor, Giulia Rosalia D'Agati KVJP-NQJ.

Her place of birth is listed as Piana dei Greci, Palermo, Sicily, Italy. but the error on that shows:

However, the Places Database is incorrect in this case, because the town was created in 1488 with the name of Piana dei Greci not in 1861 as listed.

I happen to recognize that this is a Places Database issue, NOT a Quality Scoring issue, but will there be some sort of guidance for folks who don't know, that refers them to the Places Database to correct the issue there?