ELIAS
(Enhanced Linguistic Intelligence and Analysis System)
Beginnings
ELIAS, an artificially intelligent voice print and linguistic library got its start when Der Linguist was in seventh
grade. His grandparents had friends in South Africa who wanted to give their son an international living experience. They arranged to have their son live with Der Linguist's family for a year. Der Linguist became
fascinated by the fact that both he and his South Africa friend spoke English, but not the same
English. The South African's pronunciation of words was different and he used
different words to refer to the same thing. This new perspective of
dialectic and regional language differences in the same language, intrigued Der Linguist. He decided to structure his 7th grade science fair project around the dialectic differences
in the English language. In addition to the obvious differences in the
English between South Africa and the United States, he documented dialectic differences in the Boston area where we lived.
Der Linguist graduated from High School speaking German, French,
Spanish, Dutch, Norwegian, and Hebrew with native fluency. Though he had
been accepted at Middlebury College to study Russian and
other Eastern European languages, the selective service lottery made it clear that he was headed for the military. Given
the inevitable, he enlisted in the US Army to become a Russian
linguist. This started his path to the Defense Language Institute and on to Berlin with the
Army Security Agency where he served as a Russian, Czech, and German linguist with a psychological warfare team targeted against Group Soviet Forces Europe and the secret police of East Germany, Czechoslovakia, and Hungary.
While stationed in Germany, he continued his hobby of recording phrases and words of native
speakers in different cities and regions. During this time he was able to make city and regional language recordings in Germany, The
Netherlands, Belgium, Austria, Switzerland, France, Luxembourg, Spain,
and Norway. After the wall came down, he continued the collection of
recordings in an additional 21 countries.
In total ELIAS now contains sample recordings of native speakers from 2,500 cities
and towns in 30 Western and Eastern European countries. After mastering Japanese, Japanese recordings from 92
Japanese cities and towns were also added.
Building a Computer Based System
With the advent of high speed computer processors and inexpensive RAM, Der Linguist teamed with a talented group of computer programmers and language analysts to digitize the tapes and build a computer based language / speech
pattern data base and voice analysis tool.
Today, ELIAS is a high speed computer based analysis tool
employing sophisticated algorithms and optimization branching
techniques to listen to a recorded message and then map the dialectic
inflections, intonations, vocabulary, and speech patterns to a town, an
educational level, an approximate age, an emotional state, and
psychological state of the speaker.
In exchange for providing the U.S. National Security Agency with a complete copy of ELIAS, Der Linguist and his team were given the NSA's collection of specific voice prints that were collected over the past 40 years from foreign nationals who had served in the military and/or secret police organizations in the former Soviet Block countries. Der Linguist and his team were also given individual voice prints of criminals and 'person's of interest' that were recorded by the police/intelligence agencies of 20
other foreign governments. Through such cooperation, ELIAS now contains specific voice prints of 8.6 million individuals who are of European, North American, South American, or Japanese origin.