Re: Another thing that goes wrong with naming

In article <HCEGKCLCHALDGMOIAHFDOEAACNAA.oml@...>, Owen Lewis
<oml@...> writes

>One last thought. There are five man named Dafydd and six men named Jones in
>a Company (of about 120) in the 1st Battalion, The Royal Welch Fusiliers
>(about 600 souls altogether). Consequently and in familiar use, one Jones
>will be referred to only by 442, the last three digits of his
>machine-assigned ID. For many purposes within the Company - indeed within
>the Battalion - this becomes his known name.

I'm sorry, but I can't resist posting this any longer :(

    http://www.the-tropics.net/jokes/catagories/military/the_spy.htm

and, whilst I'm at it

    http://www.ruspoli.com/folklore/numberedjokes.html

So in recompense for those links; to get back, at least vaguely, on
topic and back to the thread -- you may recall that some time ago now
there was a discussion about photo ID and provisional driving licences.

<URL:http://www.chiark.greenend.org.uk/pipermail/ukcrypto/2003-May/02557
6.html> et seq

The opinion was expressed then that the system requirement might have
been usefully expressed as the privacy-enabling "we wish to record the
identity of the person who is capable of driving a car" -- and take a
photo of successful candidates; rather than the Big Brother approach of
"we wish to check the identity of people taking the test"; which is the
way that it works at the moment.

Anyway, this recent case illuminates how the current system is panning
out:

    http://news.bbc.co.uk/1/hi/england/london/3404333.stm

To summarise, Adolphus Sorhaindo made about 50K GBP from a scam whereby
he took theory and practical driving tests on behalf of others (possibly
as many as 143 and 99 respectively). He paid 6K in fees from just one
credit card... but the system wasn't set up to query this.

Originally he just carried a photograph of himself signed on the back to
attest it was someone else. When the scheme was tightened up to include
photos on the provisional licence he just took his chances (but his
clientele was now reduced to black males because of his skin colour!)

Not surprisingly, with the occasional false moustache, he was mistaken
for the photograph of someone else and hence was allowed to take the
tests. This went on for three years until a driving examiner recognised
him as being a repeat.. [note that you get a lot more cues from a person
in the flesh than comparing with a photo]

Standard reference for an experiment showing that this is expected is:
    R Kemp, N Towell, G Pike, "When Seeing Should Not Be Believing:
    Photographs, Credit Cards and Fraud," in Applied Cognitive
    Psychology, v 11 no 3 (1997), pp 211-222.
which is discussed in Ross Anderson's "Security Engineering" book.
(Basically, it's an elegant experiment that shows that people are very
poor at matching photos and faces).

Anyway... if the photograph on the licence was of the person who passed
the test, then every time in future that the licence was used then there
would be a chance of discovery (and it is likely that the holder would
not believe that this chance was as small as it actually is -- reducing
the demand for Mr Sorhaindo's services).

I understand that various US states work it this way -- taking photos at
the time of test (perhaps they could afford to distribute polaroid
cameras to test centers ? and we thought we'd save some pennies by
making the punters pay for the snaps ?)

If it is a question of cost, then I'm not sure that we're going to find
the money for the roadside kits to create the biometrics for Owen to
hash into a unique identifier ! and, though this may sound picky, cost
is _always_ an issue in these things and you have to take a holistic
view of that as well as listing every benefit you can think of!

--

-- 
richard                                              Richard Clayton

They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety.         Benjamin Franklin