Question for Duncan Campbell re: Word-Spotting Capabilities

[Pete Bentley]

At Tue, 03 Aug 1999 00:09:41 BST, Duncan Campbell writes:
>In summary : Voice message (i.e., phone call) selection can be done on 
>called or calling telecommunications address (including the so-called "wild 
>card" criteria, selecting all message from a particular city suburb and/or 
>at a particular time); or by individual speaker recognition.   

Are there any ballpark figures on the accuracy and relative cost of
individual speaker recognition?  Assuming the accuracy is pretty high, 
then the cost of maintaining each "of interest" individual within the
system is the only factor which affects which targets this system is
used against...(if it costs US$1m per person per year, then it would
only be deployed against "high value" targets).

Logically, if word spotting is still unattainable, and the spooks are
aware of that, then they would concentrate their resources on bringing
down the cost of individual speaker recognition and on the use of
non-telecoms means to identify the individuals they are interested in.

>Word 
>spotting is not available, but as computational power increases, topic 
>spotting by running continuous speech recognition engines on a per-channel 
>basis will become affordable, at first for high value targets

Seems to me that topic spotting is a more useful goal anyway.  Even if 
you have 100% accuracy in word spotting, you will generate too many
false positive hits when that word appears out of context (eg "Boy did 
the Yankees bomb tonight").  Presumably both word *and* topic spotting 
are easy to defeat by careful use of aliases, pseudonyms,
abbreviations etc.

Pete.