Where does the "Other" detail in this Record Hint come from?
I was reviewing the profiles of some of my relatives just now and came across this:
The screenshot below shows the Record detail, so how come "Päätalo, Kuopio, Suomi" appears in the Record Hint?
Also, where does the "Danby by Middleham" come from (in the Record Hint)? It's neither the Event Place or Event Place (Original) in the "Record", and actually relates to Danby Hall, Middleham, which is about 50 miles from the parish of Danby (aka Danby in Cleveland or Danby with Castleton), where the event actually took place.
Answers
-
If you search on "placename standardization error" or similar terms, you will find dozens - perhaps hundreds - of threads covering the topic.
You have 2 issues here - one is the placename standardization algorithm that picked the first Danby it found in the database, without consideration for correctness. The other is the "standardization-on-the-fly" error that shows up in search results.
@N Tychonievich could you put this one on your spreadsheet for eventual correction, please? Thanks. URL: https://www.familysearch.org/ark:/61903/1:1:66GW-GGNN
0 -
Specifically, the Other: Päätalo, Kuopio, Suomi residence information, looking at the record itself, can only have come from Fryat, the spouse's residence. That is the only piece of information not otherwise accounted for. The search routine does a terrible job turning a single word place name into a full place name and you just can't trust what you read in a hint or in a list of search results. Instead you have to look at the actual record.
3 -
@Gordon Collett said
... Specifically, the Other: Päätalo, Kuopio, Suomi residence information, looking at the record itself, can only have come from Fryat, the spouse's residence. That is the only piece of information not otherwise accounted for. ...
An eminently sensible deduction, Mr Holmes 😉
What's deeply frustrating for me, though, is that I can't see how that's happened. If I use the LDS Standard Places facility, "Fryat" gives nothing - as in "No Results to Show".
"Frya" gives "Frya, Sør-Fron, Oppland, Norway" - but I don't think that has anything to do with it!
Working backwards, "Päätalo, Kuopio, Suomi" isn't even in the Standards database (these days?). On the other hand, "Päätalo, Oulu, Finland" is. (Yes, I'm expecting Finland to swap to Suomi, though why it did for Paul's I can't imagine.) But I can't imagine how anything arrived at either Päätalo from the data there.
If anyone reading this thinks, "What is Adrian on about - it's an error, let them fix it", well sorry, I have this delusion (it appears) that if I can direct someone in some direction, then it will assist "them" to fix the error, even if I've only got halfway there.
Unfortunately, I do feel this placename standardisation, be it "on the fly" or "background" or whatever, doesn't appear to behave in any way that I can understand. The only placename algorithms that I can see the effect of, are those where I enter the names into the FS Standard Placenames facility and see what it comes up with. But the results are so often so different from the background results that it's as if we have two (at least) separate algorithms, with the result that I feel that I can't help anyone because I can't predict anything... Sorry... 😯
2 -
Sorry (for once!) to disagree with you here, but, no, the standardization process has actually got it right (in the record / source) on this occasion. It has correctly picked "Danby, Yorkshire, England, United Kingdom" as the Event Place, so I would not want anything changed here.
The two points (errors) I am highlighting only appear in the Record Hint - and that what is/was baffling me. I now recall seeing this sort of thing previously - it happens from time to time with census sources, I believe. So, to make it clear, the issue I am trying to get to the bottom of is: from where "the program" gets the details that appear in the hint, but are not then to be found in the more important part: the record. Thankfully, when I attach the (marriage) record to my relatives' sources there will be no trace of the incorrect names that were added to the hint!
So, sorry, again for any confusion caused through my reporting this - but best to leave this one alone, @N Tychonievich, as it has been standardized just fine.
BTW - the place "Fryat" should read "Fryup", and while I see the logic of Gordon's suggestion, I would agree with Adrian that any actual connection with "Päätalo, Kuopio, Suomi" seems quite hard to figure out.
1 -
The only explanation I can think of is that at the time the hint was standardized on the fly, there was an entry in the database for "Päätalo, Kuopio, Finland", and this entry had an alternate name or other detail that somehow matched either the groom's or bride's indexed Residence Place. This entry must have since been edited or removed.
"Päätalo, Oulu, Finland" is currently in the Places database three times. Two (numbers 1906454 and 1906449) have entirely wrong coordinates: I can find nothing within ten miles of them labeled Päätalo. The third one (1906453) is within the boundaries of a place actually labeled Päätalo on the map, and matches what Google Maps comes up with if you ask it for "Päätalo, Finland". None of the three have any alternate names or other details (that I can see) that would explain any association with the English index entry.
There's also a fourth Päätalo in the database, in "Lappi, Finland"; the pin for that is in the middle of a body of water.
Kuopio, Finland the Lutheran parish has a bunch of alternate names, but none of them have any part that bears any resemblance to either "Head House" or "Fryat".
1 -
Here is a similar issue relating to what I am trying to get to the bottom of. There are a large number of records that are shown on the Results page indicating the location of individuals in the 1851 census was "Cleveland, Yorkshire, England, United Kingdom " (as shown in first screenshot), but when the record in opened up the incorrect place name no longer shows (see second screenshot).
Different circumstances, true (Result instead of Record Hint), but basically same issue involved - how is the data in the Record correct, but not on the other pages? What makes the incorrect information disappear once you click on the record - or, put another way, how does the incorrect detail appear on Results and Record Hints pages in the first place?
The "result" implies John Wrightson lived in Cleveland -
The actual record shows the correct location of Knaresborough (West Riding of Yorkshire) - nowhere close to the Cleveland area:
1 -
@Paul W said
" ... the issue I am trying to get to the bottom of is: from where "the program" gets the details that appear in the hint, but are not then to be found in the more important part: the record. Thankfully, when I attach the (marriage) record to my relatives' sources there will be no trace of the incorrect names that were added to the hint! ... "
So far as I remember this (or something similar) came up in GetSatisfaction days - the answer to a record query contained and displayed values that were not the same as the values in the record that the answer referred to.
I remember this because the techie guy who contributed to that thread didn't appear to take it as seriously as I did - the record was correct, so the issue was no more than an inconvenience. I got frustrated because, as I tried to explain, if the answer to the query is clearly not your chap (because it refers to Finland say and my chap lived in England!), then you might never click on the answer to go through and see the correct value on the record. No - when faced with 50 record answers, I am not going to click on each if some are clearly not of interest.
0 -
@Julia Szent-Györgyi suggested:
...The only explanation I can think of is that at the time the hint was standardized on the fly, there was an entry in the database for "Päätalo, Kuopio, Finland", and this entry had an alternate name or other detail that somehow matched either the groom's or bride's indexed Residence Place. This entry must have since been edited or removed. ...
That would be my thought as well. Though I thought that stuff got re-standardised automatically every so often - otherwise placename corrections are pointless as far as existing data goes. I suppose that re-standardisation run might not have happened yet? Of course, the worst-case scenario is that re-standardisation doesn't actually happen (any more)?
0
