Wanted by the FBI for questioning: Adil Pervez. Or is it Adel Pervaiz? Adil Pervaz, maybe?
As it turns out, Adil Pervez wasn't wanted after all. On Jan. 7, the FBI called off a manhunt for Pervez and four other Middle Eastern men. But the fact that alternative name spellings for the five suspects were posted online immediately after the alert points to an increasing need for understanding complexities of foreign names.
"This is something we have to deal with on a daily basis," says Justin Lindsey, the FBI's chief technology officer. "Names will be entered one way or another depending on the ports they enter, for example. It's very complex."
To get a taste of the difficulty, type "Libyan strongman" into Google. Among the unrelated results, you'll see 50 or so spellings of Mohmmar Qadaffi. Or is it Muammar Gaddafi? (In Arabic script there are only two variations.)
Alternative spellings create problems when trying to cross-check names against government watch lists. While "Adil Pervez" might raise a flag, "Adel Pervaiz" might not.
Financial firms must also come to grips with foreign names. The Patriot Act requires companies to cross-check customer names against watch lists posted by the Office of Foreign Assets Control. The Securities and Exchange Commission is formulating a similar customer-identification program to apply to brokerage firms. Multimillion-dollar fines await those who don't comply.
Software firms are racing to provide the most effective tools for foreign-name searches. "The ability to deal with foreign names," said IDC analyst Steve McClure, "is big business."
"Last year was our best ever," says Jack Hermansen, CEO of Language Analysis Systems, which provides foreign-name searching and recognition products. While 75 percent of the company's business comes from the government, Hermansen said, he's seen a sharp increase in demand from banks, airline reservation systems and data-mining companies.
Basis Technology, a Boston firm better known for its Asian localization software, has increasingly focused on the Middle East. Recently it hired CIA staff and set up an office in Washington, D.C. "I've been spending more time in Washington than in Tokyo," said CEO Carl Hoffman.
Hoffman has been urging the U.S. government to adopt Unicode-compliant systems, which makes it possible to display and process foreign names, not only as they're written in the English alphabet, but also as they appear in their native scripts. "Increasing numbers of multinational corporations have computer systems that work in multiple languages," he said, suggesting it's time for the United States to catch up.
Basis recently paired up with Language Analysis Systems to offer the government Cartouche, a name-matching system for searching and retrieving names across multiple languages.
Another firm, Intelligent Search Technology, offers ISTwatch, which lets companies quickly -- and relatively cheaply -- search and match individuals against OFAC and other terrorist lists. Companies can use the software-generated reports to show government officials they made a good-faith effort to comply with the regulations, said IST CEO Richard Wagner.
But some warn against too much reliance on automation.
"You still need linguistic and historical understanding," said Arabic expert Thomas Milo. "Otherwise foreign names become a blurred mass of variety."
While a search engine will tell you there are 200 ways to spell "Mohammed," Milo said, you still need a human expert to determine what a particular spelling reveals about a person's background.
The question, said Cord Hart, director of the Center for Asian Crime Studies, is why it's taken so long for the United States to address the issue of foreign names.
"You would think that in our multicultural society we would have done a lot better a lot sooner in dealing with foreign names," Hart said. "Now we've ratcheted up our capabilities, but we've still got a ways to go."
