Unsolved:Nemesysco

From HandWiki

Nemesysco is an Israeli company that markets discredited voice analysis programs for fraud prevention, the detection of emotion, stress management, and other purposes. The use of the company's products by insurance companies, call centers, banks and airports has raised questions about whether the underlying technology has any validity.

History

The company was founded in 2000 by Amir Liberman, an Israeli businessman with no college education or formal training in mathematics, computer science or speech science.[1] Despite the founder's lack of experience in the field and what one journal described as the program's "amateurish" programming,[1] the company's products have been put into service by insurance companies Admiral Group and Esure,[2] and by the Waupun, Wisconsin police department.[3] The company describes its software as "honesty maintenance" software, with applications including its use in human resourcing during interviews to determine the honesty of the applicant.[4]

But after spending 2.4 million pounds on tests over three years, the Department for Work and Pensions revoked its funding for the program, saying that it "couldn't confirm that it represented any good value for money."[5]

The program was also used for a study looking for links between the vocal performance of company executives and their firms' future financial performance.[6]

Assessments

Government, academic and peer-reviewed studies have found major flaws in "layered voice analysis," the technique used by Nemesysco's products. A report for the U.S. Department of Justice assessing the program's ability to detect deception found that it "failed consistently," with high rates of both false positives and false negatives.[7] A University of Florida study conducted a year earlier had found that LVA "misclassified the low-stress and truthful samples with great frequency."[8]

A third paper, published in the International Journal of Speech, Language and the Law, classified Nemesysco's programs as "charlatanry" and said that independent analyses had authoritatively debunked the company's claims because "the evidence against them is just too overwhelming to motivate any more reliability tests."[1] The paper was pulled from the journal's website—but not withdrawn—after Liberman threatened to sue for libel, claiming that the researchers did not actually use his company's products.[9]

That paper's findings were defended by David Beaver, a professor of linguistics at the University of Texas at Austin, who called Nemesysco a "bullshit lie detector company run by a charlatan."[10]

A subsequent review, published in Homeland Security Affairs, found results similar to those of the three previous studies, showing that Nemesysco's LVA software was "unable to detect deception above chance levels."[11]

References

  1. 1.0 1.1 1.2 Eriksson, Anders; Francisco Lacerda (2007). "Charlatanry in forensic speech science: A problem to be taken seriously". The International Journal of Speech, Language and the Law 14 (2): 169–193. doi:10.1558/ijsll.2007.14.2.169. ISSN 1748-8885. https://semanticscholar.org/paper/3e4d4907d833ade7248a87ac82a0741f25fc1419. 
  2. Hunter, Teresa (2003-12-16). "'We can tell if you're fibbing'". The Daily Telegraph. https://www.telegraph.co.uk/finance/personalfinance/insurance/2871936/%27We-can-tell-if-you%27re-fibbing%27.html. Retrieved 2009-02-02. 
  3. Maney, Kevin (2003-09-03). "The truth is out there, and lie-detection technology just might find it". USA Today. https://www.usatoday.com/tech/columnist/kevinmaney/2003-09-02-maney_x.htm. Retrieved 2009-02-02. 
  4. "App Tells You How You Feel". Wall Street Journal. 2014-03-10. https://online.wsj.com/news/articles/SB10001424052702303824204579421242295627138. Retrieved 2014-03-12. 
  5. Lomas, Natasha (2010-11-11). "DWP kills funding for benefit fraud-finding tech". ZDNet (CBS Interactive). http://www.zdnet.com/dwp-kills-funding-for-benefit-fraud-finding-tech-3040153890/. Retrieved 2012-07-24. 
  6. <Mayew, William J.; Mohan Venkatachalam (2012-01-17). "The Power of Voice: Managerial Affective States and Future Firm Performance". The Journal of Finance 67 (1): 1–44. doi:10.1111/j.1540-6261.2011.01705.x. https://semanticscholar.org/paper/812afa87b492d9565fcd44fabb55ff77f33d07c7. 
  7. Damphousse, Kelly R. (2007-03-31). Assessing the Validity of Voice Stress Analysis Tools in a Jail Setting. Washington, D.C.: National Criminal Justice Reference Service. https://www.ncjrs.gov/pdffiles1/nij/grants/219031.pdf. Retrieved 2012-07-19. 
  8. Hollien, H.; J.D. Harnsberger (2006-02-28). Voice Stress Analyzer Instrumentation Evaluation. Gainesville, Florida: Institute for Advanced Study of Communication Processes. http://www.voicestress.org/VoiceStress(FloridaStudy).pdf. 
  9. Cho, Adrian (2009-02-13). "Brouhaha Over Controversial Forensic Technology: Journal Caves to Legal Threat". Science (American Association for the Advancement of Science). Archived from the original on 2012-06-19. https://web.archive.org/web/20120619104953/http://news.sciencemag.org/sciencenow/2009/02/10-01.html. Retrieved 2012-07-19. 
  10. Beaver, David (2009-04-30). "Industrial bullshitters censor linguists". Language Log. University of Pennsylvania Institute for Research in Cognitive Science. http://languagelog.ldc.upenn.edu/nll/?p=1390. Retrieved 2012-07-22. 
  11. Elkins, Aaron C.; Jay Nunamaker; Judee Burgoon (2012-03-01). "Vocal Analysis Software for Security Screening: Validity and Deception Detection Potential". Homeland Security Affairs 8 (DHS Centers of Excellence Science and Technology Student Papers). http://www.hsaj.org/?special:fullarticle=supplement.4.1. Retrieved 2012-07-22. 

External links