Even more polygraph foolishness

[Bottle or the Gun]

I'm dissapointed that the security of our country is going to rely on part in this 'technology'. Dissapointed but not at all surprised that once again someone is falling for the same old scam.

Lie-detector glasses offer peek at future of security
By R. Colin Johnson
EE Times
January 16, 2004 (2:05 p.m. ET)

Portland, Ore. — It may not be long before you hear airport security screeners ask, "Do you plan on hijacking this plane?" A U.S. company using technology developed in Israel is pitching a lie detector small enough to fit in the eyeglasses of law enforcement officers, and its inventors say it can tell whether a passenger is a terrorist by analyzing his answer to that simple question in real-time.

The technology, developed by mathematician Amir Lieberman at Nemesysco in Zuran, Israel, for military, insurance claim and law enforcement use, is being repackaged and retargeted for personal and corporate applications by V Entertainment (New York).

"Our products were originally for law enforcement use — we get all our technology from Nemesys-co — but we need more development time [for that application]," said Dave Watson, chief operating officer of parent V LLC (www.vworldwide.com). "So we decided to come out sooner with consumer versions at CES."

The company showed plain sunglasses outfitted with the technology at the 2004 International CES in Las Vegas earlier this month. The system used green, yellow and red color codes to indicate a "true," "maybe" or "false" response. At its CES booth, V Entertainment analyzed the voices of celebrities like Michael Jackson to determine whether they were lying.

Besides lie detection, Watson said, the technology "can also measure for other emotions like anxiety, fear or even love." Indeed V Entertainment offers Pocket PC "love detector" software that can attach to a phone line or work from recorded tapes. It's available for download at www.v-entertainment.com. Instead of color-coded LEDs, a bar graph on the display indicates how much the caller to whom you are speaking "loves" you. V Entertainment claims the love detector has demonstrated 96 percent accuracy. A PC version is due next month.

The heart of Nemesysco's security-oriented technology is a signal-processing engine that is said to use more than 8,000 algorithms each time it analyzes an incoming voice waveform. In this way it detects levels of various emotional states simultaneously from the pitch and speed of the voice.

The law enforcement version achieved about 70 percent accuracy in laboratory trials, according to V Entertainment, and better than 90 percent accuracy against real criminal subjects at a beta test site at the U.S. Air Force's Rome Laboratories.

"It is very different from the common polygraph, which measures changes in the body, such as heart rate," said Richard Parton, V's chief executive officer. "We work off the frequency range of voice patterns instead of changes in the body." The company said that a state police agency in the Midwest found the lie detector 89 percent accurate, compared with 83 percent for a traditional polygraph.

The technology delivers not only a true/false reading, but a range of high-level parameters, such as "thinking level," which measures how much as subject has thought about an answer they give, and "SOS level," which assesses how badly a person doesn't want to talk about a subject.

How it works

Nemesysco's patented Poly-Layered Voice Analysis measures 18 parameters of speech in real-time for interrogators at police, military and secret-services agencies. According to Nemesysco, its accuracy as a lie detector has proven to be less important than its ability to more quickly pinpoint for interrogators where there are problems in a subject's story. Officers then can zero in much more quickly with their traditional interrogation techniques.

V Entertainment is leveraging the concept to let consumers in on the truth telling, eyeing such applications as a lie detector that could be used while watching, say, the 2004 presidential debates on TV.

Called Ex-Sense Pro, the V software measures voice for a variety of parameters including deception, excitement, stress, mental effort, concentration, hesitation, anger, love and lust. It works prerecorded, over the phone and live, the company said. V Entertainment recommends it for screening phone calls, checking the truthfulness of people with whom you deal or gauging romantic interest.

The display can show each measured parameter in a separate window, with real-time traces of instantaneous measurements while flashing the overall for each parameter, such as "false probable," "high stress" and "SOS." Ultimately, the company plans to offer versions of its detectors for cell phones, dating services, teaching aids, toys and games.

EE Times

 

[teddygrahams]

It works prerecorded, over the phone and live, the company said.

How long until they claim it doesn't work as well with pre-recorded material ?

 

[Bottle or the Gun]

Polygraph Foolishness

At its CES booth, V Entertainment analyzed the voices of celebrities like Michael Jackson to determine whether they were lying.

And the verdict is?

I can't help but think that much of these devices (like the drug-detector) are intended to scare and deter the amateurs.

A few science-fiction novels about the polygraph:
The Truth Machine - James Halperin
Barking Dogs - Terence Green

 

[iain]

Besides lie detection, Watson said, the technology "can also measure for other emotions like anxiety, fear or even love." Indeed V Entertainment offers Pocket PC "love detector" software that can attach to a phone line or work from recorded tapes. V Entertainment claims the love detector has demonstrated 96 percent accuracy. A PC version is due next month.

I wonder how they objectively and accurately measure how much person A really loves person B, in order to rate the accuracy of their love detector.

In general though, I wouldn't be too quick about dismissing this sort of thing. Surely the charge against the lie detector is not that it doesn't work at all, but that it doesn't work reliably and can be fooled by someone who knows what they are doing.

It may be that a lie detector with a 90% accuracy is still useful in many situations; e.g. for an insurance company to detect fraud. I hope they would not be able to rely on lie-detector evidence in court, but I don't think that is what they want it to do.

In the case of insurance companies, they want to identify people to investigate more fully. The idea is that if you call up to report a claim and the gadget says you are lying about something, the company can ask more questions or check out that element of your story in more detail. If you get the all-clear your claim gets fast-tracked. I can see how the company would benefit from this. Not sure about customers (guess it depends on the rate of false positives and that sort of thing).

This is being used by several insurance companies in the UK today.

Edited to add : The big problem with airport security is going to be false positives, as always. If one in a million passengers is a terrorist and the device is 99.99% accurate, that's still 100 innocent people being wrongly identified for every one terrorist. And of course people will look at the 99.99% thing and say "there's only a 1-in-10,000 chance that you're not a terrorist", rather than the true figure of 99-in-100. This is less likely to be a problem for the insurance industry where a significant proportion of claimants are making fraudulent claims.

 

[Bottle or the Gun]

Polygraph Foolishness

I hope they would not be able to rely on lie-detector evidence in court, but I don't think that is what they want it to do.

Am I incorrect that under certain aspects of the US Homeland Security Act, that it is evidence enough or no evidence need be provided, citing National Security concerns? That it or something like it would be used to confirm what is already suspected? I think a better data-base would be more accurate than this device and others like it.

The novels I refer to in the previous post are examples of the changes that occur in society and people when an infallible polygraph is invented.

 

[crimresearch]

Polygraphs, and these devices, are not considered evidence for court proceedings, but I would expect that under the new security regulations, that this is exactly the sort of thing that could be sufficient to trigger an indefinite detention..scary stuff, especially when you consider that an error rate which locks up a certain percentage of innocent people, is liable to allow a certain percentage of those who are intent on wrongdoing to go through unchallenged.

Paul Nunis

 

[iain]

To clarify : when I said that "they" don't want to rely on lie detector evidence in court, I was referring to UK insurance companies.

Since the US Government seems happy to lock someone up indefinitely without a trial, access to lawyers or any evidence against them other than some bounty hunter saying "I've captured this terrorist, now can I have my reward please." I really can't say that anything they do would surprise me greatly

 

[juryjone]

Portland, Ore. — It may not be long before you hear airport security screeners ask, "Do you plan on hijacking this plane?" A U.S. company using technology developed in Israel is pitching a lie detector small enough to fit in the eyeglasses of law enforcement officers, and its inventors say it can tell whether a passenger is a terrorist by analyzing his answer to that simple question in real-time.

So if he nervously says yes, he is assumed to be lying and therefore can go on the plane?

 

[Flatworm]

The heart of Nemesysco's security-oriented technology is a signal-processing engine that is said to use more than 8,000 algorithms each time it analyzes an incoming voice waveform.

Is this supposed to be impressive? Is there really any advantage to using 8000 different algorithms, versus using the best one of available? I'm guessing they expect potential consumers to evaluate this statistic as if they were quoting clock speed or memory capacity ("Hey, look! This one has more algorithms!")

 

[michaellee]

Originally posted by Bottle or the Gun

...a state police agency in the Midwest found the lie detector 89 percent accurate, compared with
83 percent(accurate) for a traditional polygraph.

When I hear of or read about claims made about a product by its inventor or an advocate, nothing
pisses me off more than the usage of terms like "accurate" or "accuracy rate" as supposed evidence of
the products functionality. Lets look at the definition.

ACCURATE.
conforming exactly to fact; errorless.

Source: The American Heritage® Dictionary of the English Language, Fourth Edition
Copyright © 2000 by Houghton Mifflin Company.
Published by Houghton Mifflin Company. All rights reserved.

I understand this definition. My cat understands this definition. Why then, is the term used as in the above quoted
items? How in the world can something be "89 percent accurate"? If something is accurate, then it works 100 times
out of 100, not 99, 89, or 83 times. I also understand the meaning of "exactly to fact", and "errorless".

This "product", as well as the traditional polygraph, both do not work, other than as a means of interrogation in the
polygraphs case. If they actually did what they claim to do, they would perform without possibility of error.

Otherwise, I could say that my cat, every Sunday, picks the winners of NFL games, signifying with his left paw as I hold
up the newspaper, and this year did so correctly at a 76 percent accuracy rate. I guess my cat should pack his things and
head for Las Vegas because this proves his cat mind works when it comes to football prognostication.

The law enforcement version achieved about 70 percent accuracy in laboratory trials, according to
V Entertainment, and better than 90 percent accuracy against real criminal subjects...

Huh? In the lab, the thing was right 7/10 times, but when they tested "real criminals", it was correct 9/10 times?
They tested only "real criminals" and still could not figure 10/10 times the testee
was a criminal? When will all these pseudo-scientific, useless products, as well as the mind boggling, never ending
spin of their believers get the f... out of our lives?

The system used green, yellow and red color codes to indicate a "true," "maybe" or "false" response.

Classic. I have an idea. Remodel the product, but include only the "maybe" color code. Guess what?
Now the thing works with 100 percent accuracy!

Originally posted by iain

Surely the charge against the lie detector is not that it doesn't work at all, but that it doesn't work
reliably and can be fooled by someone who knows what they are doing.

Wrong. Read carefully. IT DOESN'T WORK AT ALL. How do I know? When being polygraphed, the examiner
asked me repeatedly "does the polygraph work?" and I answered "yes", then "no" alternatively-
the results were the same. Everytime.

 

[iain]

When I hear of or read about claims made about a product by its inventor or an advocate, nothing
pisses me off more than the usage of terms like "accurate" or "accuracy rate" as supposed evidence of
the products functionality. Lets look at the definition.

ACCURATE.
conforming exactly to fact; errorless.

I understand this definition. My cat understands this definition. Why then, is the term used as in the above quoted
items? How in the world can something be "89 percent accurate"? If something is accurate, then it works 100 times
out of 100, not 99, 89, or 83 times. I also understand the meaning of "exactly to fact", and "errorless".

This "product", as well as the traditional polygraph, both do not work, other than as a means of interrogation in the
polygraphs case. If they actually did what they claim to do, they would perform without possibility of error.

Otherwise, I could say that my cat, every Sunday, picks the winners of NFL games, signifying with his left paw as I hold
up the newspaper, and this year did so correctly at a 76 percent accuracy rate. I guess my cat should pack his things and
head for Las Vegas because this proves his cat mind works when it comes to football prognostication.[/b]

Michaellee,

I can see what you're saying, but I think you're mistaken. "90% accurate" does have a perfectly sensible and understandable meaning : that the result reflects reality nine times out of ten.

Whether this is desirable behaviour or not is another question entirely and the answer is that it depends what you want to do. A device that accurately predicted 9 times out of 10 what I should say to my wife when asked "does this make me look fat" to avoid upsetting or annoying her would be wonderful and I'd buy it in a second. A device that accurately predicted whether someone was guilty of murder nine times out ten would not be appropriate to decide whether to send them to prison or to the chair.

If your cat really could consistently predict the winner of NFL games with an accuracy signficantly greater than chance, you could win an awful lot of money. It doesn't prove anything about how the cat might be achieving this, but it does make sense to talk about the cat being 76% accurate.

Wrong. Read carefully. IT DOESN'T WORK AT ALL. How do I know? When being polygraphed, the examiner
asked me repeatedly "does the polygraph work?" and I answered "yes", then "no" alternatively-
the results were the same. Everytime.

This is an anecdote, not evidence. On the basis of one trial you are saying something doesn't work at all. I might similarly claim that heavier-than-air flight is impossible because one plane crashed. You might be right that it doesn't work at all; but your anecdote doesn't tell us that one way or the other.

Edited to add : If your cat really understands the definition of "accurate", that's one clever pussy you've got there.

 

[_Q_]

When I hear of or read about claims made about a product by its inventor or an advocate, nothing
pisses me off more than the usage of terms like "accurate" or "accuracy rate" as supposed evidence of
the products functionality. Lets look at the definition.

ACCURATE.
conforming exactly to fact; errorless.

Source: The American Heritage® Dictionary of the English Language, Fourth Edition
Copyright © 2000 by Houghton Mifflin Company.
Published by Houghton Mifflin Company. All rights reserved.

Is there maybe a little more to their definition than what you've quoted here?

1. Conforming exactly to fact; errorless.
2. Deviating only slightly or within acceptable limits from a standard.
3. Capable of providing a correct reading or measurement: an accurate scale.
4. Acting or performing with care and precision; meticulous: an accurate proofreader.

Source: The American Heritage® Dictionary of the English Language, Fourth Edition
Copyright © 2000 by Houghton Mifflin Company.
Published by Houghton Mifflin Company. All rights reserved.

If one looks up "accuracy" from the same source, one finds:

1. Conformity to fact.
2. Precision; exactness.
3. The ability of a measurement to match the actual value of the quantity being measured.

It seems to me that some of these definitions allow some more flexibility than an "all or nothing" evaluation of something.

Would you be content if such claims excluded the words "accurate" and "accuracy", but still stated plainly that a certain device, used under a given set of circumstances and in a given number of instances, yielded the correct answer in a particular percentage of those instances? Is it just the "a-words" that are a hot button?

If it's just the way you use the words compared to the way some people use the words that bothers you, you're not alone. Reading those definitions (and also their definition of "precise), I find that the way they kick "accuracy" and "precision" around makes me grit my teeth.


_Q_

 

[ktesibios]

I have to wonder what meaning phrases like "90 percent accurate" actually have.

It can't be "accuracy" as it's defined for measuring equipment, or "precision" as also defined for measuring equipment (they're not the same; the former refers to the maximum expected error of the measurement and the latter to the smallest change in the measured quantity detectable by the measurement).

If stories like these referred to concepts that actually have defined, relevant meanings- like "sensitivity" and "specificity"- it might not be quite so scary to think that gadgets like this could wind up being used to make decisions affecting the fate of actual human beings.

Even better would be a sign that the people developing and evaluating such devices are aware that, just as with a test for a medical condition, knowing the sensitivity and specificity of the test still requires you to know something about the incidence of the condition being tested for among the population being tested before you can make predictions about the number of true positives that escape detection and true negatives falsely detected.

Unfortunately, I don't have much confidence that there's any room for such knowledge in the media, or in the police mentality, nor among the secret-police creeps gadgets like this one are being pitched to.

Meaningless phrases like "90 percent accuracy" might do well enough for selling quack medical devices, but as a basis for public policy, or to justify to the public why their fellow citizens are disappearing into the Nacht und Nebel-

FEH!

 

[RPG Advocate]

...and its inventors say it can tell whether a passenger is a terrorist by analyzing his answer to that simple question in real-time.

Americans will love this device, then! Who needs CAPS-II when you've got this baby?

being repackaged and retargeted for personal and corporate applications by V Entertainment (New York).

Does anyone else find it strange that a company marketing a device that, if used, could change a criminal suspect's fate forever has the word entertainment in its name?

The heart of Nemesysco's security-oriented technology is a signal-processing engine that is said to use more than 8,000 algorithms each time it analyzes an incoming voice waveform. In this way it detects levels of various emotional states simultaneously from the pitch and speed of the voice.

Right, this is just a CVSA whose developer is making more preposterous claims than normal. Nothing revolutionary...unless it gets in the hands of law enforcement. Then it will be a revolutionary disaster. The only way I could see CVSA technology really working is if you could "train" the device to recognize certain waveforms associated with certain emotions in a specific person. Since training would take many, many trials, this is impractical for most people who would have a use for such technology.

The law enforcement version achieved about 70 percent accuracy in laboratory trials, according to V Entertainment, and better than 90 percent accuracy against real criminal subjects at a beta test site at the U.S. Air Force's Rome Laboratories.

Like the CVSA achieved such "accuracy"?

"SOS level," which assesses how badly a person doesn't want to talk about a subject.

It works prerecorded, over the phone

[color=ff0000] Bullsh*t Alert!!! [/color]

By the way, Bottle or the Gun, posting the entire text of an article, even if cited, is considered bad form here. Just post an excerpt from now on.

 

[Bottle or the Gun]

[color=ff0000] Bullsh*t Alert!!! [/color]

By the way, Bottle or the Gun, posting the entire text of an article, even if cited, is considered bad form here. Just post an excerpt from now on.

Well, sure. Know that now.