Germany wants to streamline the process for granting asylum to refugees by using speech analysis software to determine where they come from, according to German newspaper Die Welt (via Deutsche Welle).

Like many countries in Europe, Germany is dealing with a massive influx of refugees from other countries. With so many people seeking asylum, many of whom lack proper paperwork, it can be nearly impossible for officials to accurately verify their point of origin, which is an important factor in determining if someone is eligible for refuge.

One solution to this problem is dialect analysis, which looks at the way applicants speak to try to pinpoint where they come from. Dialects are basically regional variations of the same language, and they can often be a clear indicator of the speaker’s background.

For example, in the U.S. you can roughly guess what part of the country someone is from if they refer to soft drinks as soda, pop or Coke. The same can be said for people who use words like y’all or youins.

According to Deutsche Welle, Germany has been using linguistics experts to analyze refugee dialects for nearly 20 years, but this requires a high level of knowledge and training, and it can greatly slow down the asylum application process. To solve this problem, Germany hopes to at least partly automate the linguists’ role with speech analysis software.

Germany’s Federal Office for Migration and Refugees, or BAMF, says it will test out a speech analysis program that will compare voice recordings of asylum seekers to a library of speech samples. BAMF will begin testing the software within the next two weeks, and the organization has plans to expand the program by 2018. “The idea is to record a separate speech sample from asylum seekers and to carry out an automatic dialect analysis,” explained Julian Detzel, a member of BAMF’s Global Strategy digitization and IT program management.

Some linguists are skeptical of the software, however, and worry that asylum seekers could be misidentified. “I don’t see how automated software can distinguish whether a person uses a certain word or pronounces it in a particular way because this is part of their own repertoire or because they were primed to do so by the interviewer or interpreter,” Monika Schmid, a professor of linguistics at the University of Essex told Deutsche Welle.

Another complicating factor in dialect analysis is the fact that language is constantly changing, and it will be difficult for BAMF to obtain the volume of data needed to train the software and keep it up to date, especially for rare dialects that are not easy to track down.

According to Deutsche Welle, BAMF had 45 experts who understood a total of 80 different languages as of last year, and the organization conducted 411 analyses of refugee dialects in 2015.

Photo: Jürgen Matern – Own work (JMatern_071104_8454-8458_WC.jpg), CC BY-SA 3.0, Link