Fri May 14, 2004


My post about voting networks in the Eurovision led to a followup from Danyel Fisher, a grad student at Irvine who studies social networks. His weblog is has lots of interesting stuff, including a better-informed version of a post I’ve been meaning to write for a while about fingerprint databases. When the U.S. announced that it was going to fingerprint visitors entering the country, I began to wonder when the vast size of these databases was going to run up against the problem of false positives. Although we think of fingerprints as unique, the matching process is prone to error (like everything) and, for a large enough scale, your prints may be essentially identical to someone else’s. Daniel’s post links to a story where exactly this happened, in the case of the Spanish investigation into the train bombings. A perfect match turned up in Portland, Oregon.

Danyel links to a paper On the Individuality of Fingerprints (pdf). I also know of—but haven’t read—Simon Cole’s Suspect Identities, a study of the emergence and institutionalization of standards for fingerprinting.