opposite of a false positive

[Marc Adler]

I'm translating a patent on biometric authentication, and I need to know what the opposite of a false positive is.

Assuming that in this context a false positive is mistakenly recognizing a phoney as the real thing (e.g., Mr. A successfully duping the system to believe he's Mr. B), I need to know what the opposite situation is called, i.e., when the system mistakenly rejects Mr. B thinking that he's not Mr. B, even though Mr. B is the person the system is supposed to let in.

A false negative? A true positive? The Japanese corresponding to "false positive" is the unwieldy 他人の受け入れ率 and the Japanese corresponding to the term I'm looking for is the equally unwieldy 本人の拒否率.

If anyone can suggest any other terms for these phases, I would open to hearing those, too.

Thanks in advance.
--
Marc Adler
http://www.linkedin.com/in/adlerpacific

[Terry Gallagher]

Well, "false negative" is an established term, and may accurately reflect the meaning you're looking for, but may still be too cryptic.

How about "falsely (or mistakenly) identify" for "false positive," and "failure to recognize" for "false negative"?

Terry Gallagher

Eastham, MA USA

[Jarrad Shearer]

Mark,

As Terry stated, false negative is the opposite of false positive. I
usually hear these terms used in the context of virus detection. In
your context (biometrics) you could use false acceptance for FP and
false rejection for FN.

Interestingly my CISSP textbook lists those terms in the biometrics
section when referring to FPs & FNs.

Cheers,
Jarrad.

> --
> You received this message because you are subscribed to the Honyaku Mailing list.
> To unsubscribe from this group, send email to honyaku+u...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/honyaku?hl=en?hl=en
>

[Marc Adler]

On Fri, Jul 23, 2010 at 5:41 PM, Jarrad Shearer <grumpy...@gmail.com> wrote:

 

usually hear these terms used in the context of virus detection. In
your context (biometrics) you could use false acceptance for FP and
false rejection for FN.

Ah-ha! Those are good.

[Mark Spahn]

To undestand any term that ends in 率,

you need to know what number the numerator is,

and what number the denominator is:  the rate

is calculated by dividing what by what?

Does 本人の拒否率 mean

(1) (number of authorized persons who are rejected)/

(number of all authorized persons),

or does it mean

(2) (number of authorized persons who are rejected)/

(number of all persons who are rejected) ?

Or maybe what is being counted in both numerator

(the top number) and denominator (the bottom number)

is not the number of persons of a certain type but rather

the number of "Please let me in" attempts.

 

Similar remarks apply to the ambiguity of

他人の受け入れ率 .

 

As a medical term, _positive_ means

"demonstrating or proving the presence or existence of

a condition, symptoms, bacteria, etc.",

and _negative_ means

"not demonstrating or proving the presence or existence of

symptoms, bacteria, fracture, etc.",

and in an authentication context, "true" would mean

a correct identification and "false" would mean

an incorrect identification, i.e., a misidentification. 

So the four possibilities are:

 

true positive = identifying an authorized person as an authorized person

false positive = identifying an impostor as an authorized person

true negative = identifying an impostor as an impostor

false negative = identifying an authorized person as an impostor

 

Also relevant here is the vocabulary of

http://en.wikipedia.org/wiki/Type_I_and_type_II_errors .

 

To me, none of these terms has a meaning that is immediately

obvious.  I always have to puzzle out and reassemble

the meanings of _true/false_ and _positive/negative_.

And grasping the meaning of a Type I error or

a Type II error causes even more time-consuming

brain-churning.

Anyone who comes up with snappier terms for these

concepts -- terms whose meaning is conveyed directly --

would be doing a good public service.

-- Mark Spahn  (West Seneca, NY)

 

[Mark Spahn]

"Anyone who comes up with snappier terms for these

concepts -- terms whose meaning is conveyed directly --

would be doing a good public service."

 

Jarrad Shearer suggests:

you could use "false acceptance" for FP [false positive]

and "false rejection" for FN false negative.

"Ah-ha! Those are good."

Yes!  A good public service.  Thank you.