Configurable Pattern Recognition and Filtering Tool patent

The United States Patent and Trademark Office has issued a patent for the invention underlying Data Splitter :

"Configurable Pattern Recognition and Filtering Tool", United States Patent Number 6,687,694.

Check it out at the USPTO website:


Here's what it cost, in U.S. dollars :

12/31/1999 Provisional application fee 75.00
11/30/2000 Attorney - retainer 350.00
01/04/2001 Attorney / PTO app fee 705.00
05/01/2003 David Pressman book 40.00
06/04/2003 Extension fee 465.00
06/04/2003 Final review by attorney 350.00
Total: 2950.00

Plus about three hundred hours of labor spread over almost four years.

Initial investigation

Late in 1999, when the program started working well, I wondered:  is it patentable?   I made a few visits to the PTDL (Patent and Trademark Depository Library) in San Diego's main library.   There I found info on the required procedures and "prior art":  inventions in the same area that might render my idea unpatentable.   I found a number of similar inventions but nothing close enough to deter me.

The provisional application

This consisted mainly of a high-level software specification, something a competent developer could, in theory, use as a blueprint.   Mailed it on Dec. 31, 1999, about 5:30 pm, the last mail pickup of the twentieth century.   This set the clock running - the non-provisional application must be filed within one year of the provisional.

The non-provisional application

i.e. the real thing.   With the deadline looming a few months away, late in 2000, I started talking to patent attorneys, and selected the best one:  the only guy I found who seemed to have any understanding of software.   I sat and explained the thing to him for an hour or so and let him draft the claims.   We got it in the mail on the very last day, January 5, 2001.

... two years go by ...

The Office Action + the Amendment

Early in 2003 I received an "Office Action", listing the Examiner's reasons for rejecting all the claims in my application.   The rejection's basis was "obviousness":  an invention must be non-obvious to be patentable.   The rejection was accompanied by a slightly frightening stack several inches thick of "prior art" patents.   I was more disappointed than I should have been ...

I began the long slog through the Office Action and accompanying prior art citations, and after a while (a few months) realized that the Examiner had responded with what might be referred to as a "shoot from the hip" rejection (Pressman, 13/20).   The deadline for my response to the Office Action, or "Amendment", was early in June, so I spent the entire month of May, 2003, going point-by-point through the Examiner's objections explaining why my "Configurable Pattern Recognition and Filtering Tool" was non-obvious.   I mailed the Amendment, 30 pages or so, on the deadline:  June 4, 2003.

What is it with those me and those deadlines?

If I'd had more time I would have used it, but my gut feeling, placing that big envelope in the mail that night, was:  what I stuffed in the box was fundamentally right.


The Examiner bought my arguments and allowed (granted) the application in September, 2003.   The patent issued on February 3, 2004.

What I learned

If you think you have a patentable idea :

Even if you aren't planning to apply for a patent, Pressman's book is good reading for anyone working with technology.