IP: NSA patents voice recognition technology

[Dave Farber <farber () cis upenn edu>]

> From: Declan McCullagh <declan () well com>
>
> [I once worked on a project at Carnegie Mellon that used the Sphinx voice
> recognition system developed in the school of computer science. (I did the
> user interface and database.) It's hardly a surprise that the US government
> wants to be able to do this, and probably funded that CMU project, though I
> no longer remember one way or another. Hmm. I wonder what the NSA could use
> this patent for? --DBM]
>
> ===
>
> http://www.independent.co.uk/news/Digital/Features/spies151199.shtml
>
>             The US National Security Agency has
>             patented a new technology for monitoring
>             millions of telephone calls, so watch out, it's
>             now even easier for the spooks to eavesdrop
>             on your conversations
>
>
> http://164.195.100.11/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&;
> u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='5,937,422'.WKU.&OS=PN/5,937,422&RS=
> PN/5,937,422
>
> United States Patent
>                                                                    5,937,422
> Nelson ,   et al.
>                                                              August 10, 1999
>
>
> Automatically generating a topic description for text and searching and
> sorting text
> by topic using the same
>
>                                   Abstract
>
> A method of automatically generating a topical description of text by
> receiving the text containing
> input words; stemming each input word to its root form; assigning a
> user-definable part-of-speech
> score to each input word; assigning a language salience score to each input
> word; assigning an
> input-word score to each input word; creating a tree structure under each
> input word, where each
> tree structure contains the definition of the corresponding input word;
> assigning a definition-word
> score to each definition word; collapsing each tree structure to a
> corresponding tree-word list;
> assigning a tree-word-list score to each entry in each tree-word list;
> combining the tree-word lists
> into a final word list; assigning each word in the final word list a
> final-word-list score; and choosing
> the top N scoring words in the final word list as the topic description of
> the input text. Document
> searching and sorting may be accomplished by performing the method
> described above on each
> document in a database and then comparing the similarity of the resulting
> topical descriptions.
>
>
> Inventors:
>           Nelson; Douglas J. (Columbia, MD); Schone; Patrick John
> (Elkridge, MD);
>           Bates; Richard Michael (Greenbelt, MD)
> Assignee:
>           The United States of America as represented by the National 
> Security
>           (Washington, DC)
>
> [...]
>
> FIELD OF THE INVENTION
>
> This invention relates to information processing and, more particularly, to
> automatically generating a
> topic description for text and searching and sorting text by topic using
> the same.
>
> BACKGROUND OF THE INVENTION
>
> Identifying topics of text has been an area of study for several years, and
> identifying such in
> unconstrained speech has been an area of growing interest. The latter of
> these two areas, however,
> seems to be more difficult since much of the information conveyed in speech
> is never actually spoken
> and since utterances frequently are less coherent than written language.
>
> The standard method of electronically searching for a document related to a
> particular topic is by
> using keywords. In a keyword search, a user selects a small set of words
> (i.e., the keywords) which
> may be expected to occur in documents related to the topic of interest. The
> documents are then
> searched for occurrences of the keywords. Documents containing the keywords
> are then presented
> to the user. A disadvantage of this method is that relevant documents that
> do not include the
> keywords will not be retrieved.
>
> [...]
>
>
>
> --------------------------------------------------------------------------
> POLITECH -- the moderated mailing list of politics and technology
> To subscribe: send a message to majordomo () vorlon mit edu with this text:
> subscribe politech
> More information is at http://www.well.com/~declan/politech/
> --------------------------------------------------------------------------