Example: how a slight spelling variance affects the matching score.
For a long time, in Spanish, J and G (before e and i) have sounded the same, and were considered equivalent in writing. As an example, Gil was also written Jil. The parish person write Gil on the left, even as the father was clearly Mariano Jil (see how the J in Mariano Jil is the same J as in Joaquina).
However, if we test this search, https://www.familysearch.org/en/search/record/results?count=100&q.anyDate.from=1800&q.anyDate.to=1880&q.anyPlace=Guacar%C3%AD%2C%20Valle%20del%20Cauca%2C%20Colombia&q.givenName=jose%20mar%C3%ADa&q.recordCountry=Colombia&q.surname=gil
the sorting algorithm believes that for a name like Jose María Roa Gil*
MatchSimilarity("Jose María Gil", "Jose María Roa Gil"), and even Maria Mercedes Lozano Gil** is greater than
MatchSimilary("Jose María Gil", "José Maria Jil").
*where Roa is the father's lastname, Gil the mother's lastname, and when the person would have been generally known as José María Roa.
**(María Mercedes is female name, while Jose María is a male name).
——————————————————————————The search produces almost 8000 records. Among them, I have not seen one that finds the form Jose María Jil.It seems to me that if ?ge? = ?je? and ?gi?=?ji?, the sorting should place both Jose María Gil and Jose María Jil at the very top of the list.
Other names that are equivalent:
- Rengifo/Renjifo,
- Gimenez/jimenez,
- Jilberto/Gilberto,
- Mejía /Megia
and I forgot that Ximenez and Mexia are also present.
So, please consider including these letter equivalences when computing similarities between names and sorting by the resulting score.
——————————————————I understand that this could get very complicated for other languages, but for Spanish, with clear correspondence between script and sound, it is very simple.Thank you!