Arabic poses a challenge because words often are not separated by spaces
Arabic poses a challenge because words often are not separated by spaces. In addition, conjugations, gender and number can be incorporated into the word. This results in many spelling variations on one term. For example, the Arabic word for "airplane" is much different from "the airplane" or "airplanes."
"Government computer systems generally work in English," said Carl Hoffman, chairman and chief executive officer for Basis Technology Corp. "In an age where we deal with people all over the world, English-only computers don't cut it anymore."
Basis Technology sells analyzers to software companies — many of which have contracts with the government — to defense contractors and directly to the government if agencies have specific needs.
Glenn Nordin, assistant director for language intelligence policy for the Defense Department, said this type of software highlights the problems that arise when Arabic words are translated into the Roman alphabet.
"Anything that can help us sort out the names — place names and personal names — in the streams of traffic is absolutely essential," Nordin said.
Everette Jordan, director of the FBI's National Virtual Translation Center, said that before these tools were available, searches were done manually and cognitively. The Basis Technology software will speed up searches in documents with large amounts of information, especially names.
"The variations in spelling do a number on you," Jordan said. "If you're doing a search, you have to know each different kind of spelling.... Ten years ago, you could only dream of something like this — maybe even five years."
NEXT STORY: SBA decertifies HUBZone companies