Automating asylum

In order to process a backlog of thousands of asylum claims, the German government will pilot the use of speech recognition software this week. Originally developed for financial security checks in telephone banking, the software assesses speech patterns to establish claimants’ place of origin. As Deutshe Welle reports, lingustic assessments have been part of the verification process in Germany since 1998 where lingustic data was sent to linguists to analyse. Even then, the system is flawed, with a recent study by the University if Essex’s Prof. Monika Schmid highlighting the inaccuracy if this approach. The study found that German citizens identified recordings of German expats as not being of German origin due to language attrition. Schmid goes on to say:

“We have argued that in order to do so reliably, an analyst must have a solid background in linguistic analysis and be able to take into account a wide range of factors. For example, people will adapt the way they speak to the speech patterns of their interlocutors. 

“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.”