Fixing the No Fly List with Cross-lingual Name Matching

Or: How to detain Tsarnaev But Not Your Grandmother

Recent news stories trumpeting the unconstitutionality of the No Fly list (Washington Post) miss the point.  Sensationalized headlines like “No-fly list violates constitutional rights, judge rules” (Fox News) and “Federal judge rules US no-fly list violates Constitution” (Reuters) misinterpret U.S. District Court Judge Anna Brown.  The ruling states that while maintaining a No Fly List is constitutional, the procedure for allowing people to remove themselves is flawed.

Like it or not, there are people who seek to cause harm, and our government must be vigilant in preventing this. The real issue is not the existence of a No Fly List. It’s how best to prevent people who intend to do harm from boarding airplanes, while allowing law-abiding citizens the freedom to travel.  For better or worse, identification begins with a name.

A fundamental problem is names in languages that don’t use the Latin alphabet (such as Russian, Chinese, and Arabic) are rarely transliterated to a consistent English spelling.  For example, the first name of the Russian composer Tchaikovsky may be written as Pyotr or Peter.  The revolutionary leader of the People’s Republic of China may be written as Mao Tsetung or Mao Zedong.  And Arabic, Persian, Pashtun, and Urdu names may have hundreds or thousands of spelling variations.

Many government agencies, such as the Treasury Department’s Office of Foreign Assets Control, attempt to cope with this problem by creating lengthy lists of spelling variations for each individual name.  There are many problems with this approach.  Lists of spelling variations are inaccurate, unwieldy, and, most of all, ineffective.  Representing foreign names this way introduces uncertainty and removes transparency from the process of name verification, leading to both “false positives” and “false negatives.”

Basis Technology offers a powerful software solution for “No-Fly” lists and financial compliance watch lists which operate in a fundamentally different way.  By recognizing that names of global travellers—and of American citizens—may have origins in dozens of widely spoken languages, we have designed name matching software that does not rely on storing transliteration variants. It uses knowledge of the structure and culture of names in each country or language to accurately match names across languages and writing systems.

This strategy significantly boosts the accuracy of computerized watch-list checking systems. A smarter name matching system means not only stopping the next Christmas bomber (Omar Farouk Abdulmutallab) or Marathon bomber (Tamerlan Tsarnaev) before they cross our border, but also means protecting the rights of all law-abiding people, regardless of what name their parents gave them.