Feb. 27th, 2012

walkitout: (Default)
I don't actually know how Ancestry.com generates hints, however, it really seems to be pretty agnostic to non-surname ordering. That is, it's perfectly happy to match Anna Jane Smith to Jane Anna Smith. I can understand why that's a good idea and I rely upon it quite heavily, especially with United States Census Records.

The match is pretty fuzzy on dates as well. It'll definitely take anything within a year or two, and often is willing to take a wider range than that, even.

Unfortunately, the fuzzy match interacts really poorly with Frisian patronymics, and here's why. A very common naming pattern might look like this:

Dad: Lyckle Jans Surnamema (<-- a little Frisian joke there: take a plausible surname, stick "ma" on the end of it, and it sounds a little Frisian)
Son: Jan Lyckles Surnamema

In one of these families, there will be a ton of kids, so Jan Lyckles will have siblings like Grietje Lyckles Surnamema and Beerendje Lyckles Surnamema or Wybe Lyckles Surnamema or whatever. Dad probably has a bunch of siblings, too, named things like Grietje Jans Surnamema and Beerendje Jans Surnamema or Wybe Jans Surnamema. They'll have kids, too, named things like (in the last case) Jan Wybes Surnamema or Lyckle Wybes Surnamema.

As you can see (once you wrap your brain around it all), it's actually really easy to figure out who is related to who and approximately how. However, Ancestry.com's loose spelling match and agnosticism with respect to non-surnames means that it'll mix all these guys up really gratuitously -- even matching fathers to sons, despite a 20 year or greater difference in birth years.

Frustrating.

September 2025

S M T W T F S
  1 2 3 456
78910111213
14151617181920
21222324252627
282930    

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 5th, 2025 05:53 am
Powered by Dreamwidth Studios