>1.9M people is a massive number of people I would say incredibly massive. I mean, to get onto that "terrorist" list (if it is of US origin) most probably it

>1.9M people is a massive number of people I would say incredibly massive. I mean, to get onto that "terrorist" list (if it is of US origin) most probably it means that the person must have traveled to or within the US by plane (or applied for a Visa, or similar), otherwise the fields of passport_id and country_of_issuance can be empty and (false or real) matches only depend on name, surname and date of birth. In Italy something that is similar to the US SSN is the Codice Fiscale that can be thought as a condensed string based on Surname, Name, Sex, Date of birth, and place of birth (each italian comune, think of a municipality has its own alphanumeric code). At the time it was introduced, there were of course a number of "collisions" due to homonimy or partial homonimy (the algorithm uses only some consonants of Surname and Name, making collisions more likely) and a provision for an "exception" was made, altering last character (which is a control character and normally is a sort of hash of the other data). This allowed up to 26 people born in the same municipality on the same day and with same (or similar) surname and name to get a distinct codice fiscale. For strangers the code for place of birth is instead of municipality the much more generic country of birth. The amount of collisions with foreigners staying in Italy or however needing the codice fiscale (coming from a few countries) zoomed to higher than that and now there are quite a few "non-standard" codice fiscale where one (or more) of the numeric characters has been replaced by a letter character according to a substitution algorithm 0=L, 1=M, etc. starting from the last (rightmost) number. This has raised the number of possibilities from 26 to (I believe) 128 or so and for the moment the system still works. However, last data I could find, is 2016 and there were 36,000 "collisions" on 94,000,000 total, (in 2000 it was some 24,000 on 80,000,000 or so). I wonder what kind of probabilities there are, if you have a common enough name and you are born in a country for which there are many entries in that list to be "picked" for homonimy, if the passport info is either empty or not used.