« Sour Grapes | Main | Driving »

December 22, 2008

Funked Koo Kung Fu

IBM inventors Fred Koo and Ting Leung got their database query modification patent application sua sponte kicked in its Bilski on BPAI appeal. The decision shows how moronically the CAFC In re Bilski ruling is being applied. Not that something better could have been expected.

Ex parte Fred Koo and Ting Y. Leung (BPAI 2008-1344)

An independent claim 101ed a la Bilski:

1. A method for optimizing a query in a relational database management system, the method comprising:
    evaluating the query to determine whether a sub-expression of the query is being joined to itself and whether a predicate of the query comprises an equality test between a same column of the sub-expression;
    determining whether a first row set producible from a first set of references of the query to the sub-expression is subsumed by a second row set producible from a second set of references of the query to the subexpression; and
    reforming the query to eliminate the joining of the sub-expression to itself based on evaluation of the query and determination of whether the first row set is subsumed by the second row set.

Relational databases execute in a machine known as a computer. The above claim goes to changing a search/query of the database. Analogous to the physical world, a database is an organized set of records, and a query is, well, a query - an instruction to a searcher to find records in the database. The claim therefore goes to stating the search properly, so that the right records may be found.

Rumor once was that the patent office knew a thing or two about technology. That's no longer so apparent.

Claim 1 is rejected under 35 U.S.C. § 101 because the claimed invention is directed to non-statutory subject matter. We note that claim 1, given its broadest reasonable interpretation, does not require computer or machine implementation. Claim 1 is not directed to a machine, manufacture, or composition of matter. Accordingly, the issue is whether this claim, which covers a method for optimizing a query in a relational database management system, involves a new and useful "process," and thus, recite patentable subject matter under 35 U.S.C. § 101. Giving the claim limitations their broadest reasonable interpretation, we conclude that claim 1 is unpatentable under section 101 because it seeks to patent an abstract idea.

Nothing abstract about proper wording. Are spelling checkers an abstract idea? And, suddenly, computers are not machines.

In view of In re Bilski, 2007-1130, slip op. (Fed. Cir. Oct. 30, 2008) we must consider the statutory sufficiency of the recited claims under 35 U.S.C. § 101. Under Bilski, "the machine-or-transformation test is the only applicable test and must be applied, in light of the guidance provided by the Supreme Court and this court, when evaluating the patent-eligibility of process claims." (Id., slip op. at 29).

Claim 1 does not recite any steps that necessarily involve machine implementation. While the preamble of claim 1 recites a "system," the "system" of claim 1 is not recited in terms of hardware or tangible structural elements. Rather, the "system" could be a software system, where the elements of claim 1 are implemented solely in software or algorithms. Thus, the nominal recitation of a "system" in the preamble does not transform claim 1 into patentable subject matter under § 101.

Computer software necessarily runs on computer hardware. Computer hardware makes for a machine. Computers nowadays are ubiquitous machines. Cell phones, TV cable boxes, and microwave ovens are all forms of computers, as well as the laptops and desktop workstations more often thought of as "computers." There is no such thing as "a software system." Software necessarily implies operational residence in a computer. Software is incapable of independently worming its way into the earth or crawling under the sofa. Only someone completely computer-ignorant could write something like that.

Each of these steps does not call for any transformation of an article to a different state or thing, nor does it require any transformation of data or signals.

A query is a set of instruction data, a search specification. "Reforming the query" means changing the search instruction. Change = transform.

Hard to believe this is the 21st century, and not the 14th. The courts and USPTO seem awfully busy getting stupid about software.

Posted by Patent Hawk at December 22, 2008 8:29 PM | § 101

Comments

Two problems hawk. Reforming could refer to making an entirely new query. And it likely does, since a query cannot really be "shaped" or "molded" or "changed", it either is a certain query or it is a different certain query. Second, reforming the query could result in the same query.

Posted by: 6 at December 23, 2008 12:27 AM

PH,

As Ex parte Koo shows, the BPAI is either clueless or more likely, devious in applying Bilski's "machine or transformation" test.

Posted by: EG at December 23, 2008 5:48 AM

Even though computer software necessarily runs on computer hardware, the claim does not transform anything except the abstract query, nor does it tie the method to any particular hardware or (non-signal) computer-readable medium leaving the descriptive material of the method claim nonstatutory.

So whether or not the BPAI really is clueless or devious or whatever else, this appears to be the proper interpretation given the case law supplied by the CAFC.

Posted by: VernoWhitney at December 23, 2008 6:46 AM

BPAI got it right. Bilski got it right. The person who got it wrong is the one who drafted those claims. You get what you pay for, IBM.

Posted by: Defector at December 23, 2008 9:50 AM

Maybe this claim might me more acceptable:
1. A method for directing a relational database management system to optimize optimizing a query, the method comprising:
directing the relational database management system to:
(I) evaluate evaluating the query to determine: (a) whether a sub-expression of the query is being joined to itself, and (b) whether a predicate of the query comprises an equality test between a same column of the sub-expression;
(II) determine determining whether a first row set producible from a first set of references of the query to the sub-expression is subsumed by a second row set producible from a second set of references of the query to the subexpression; and
(III) reform reforming the query to eliminate the joining of the sub-expression to itself based on: (i) evaluation of the query, and (ii) determination of whether the first row set is subsumed by the second row set.

Posted by: jj at December 24, 2008 6:45 AM

jj,

Can you explain what about your rewritten claim would distinguish over the claim that was summarily dismissed by the CAFC with In re Bilski?

In your explanation, please point out why the original claim failed In re Bilski and your claim does not.

Posted by: breadcrumbs at December 24, 2008 8:03 AM

I would think that jj's claim is tied to a "particular machine" - namely a "relational database management system"
I don't know if this would clear the Bilski hurdle - all this Bilski/Benson googly-gook makes no sense to me.
I mean the original claim from Benson was tied to a particular machine (i.e. the shift register) and transformed the article (i.e. the shift register) into a different article (i.e. the shift register) to a different state.

So Benson is self-contradictory, and Bilski was forced to follow the Benson precedent.

BTW I have NO problem if the courts or legistlators want to exclude SW for POLICY reasons (I actually disagree), but this Benson/Bilski stuff is PURE nonsense.

Posted by: Moshe at December 24, 2008 8:23 AM

Moshe

Todah (thanks) for speaking truth to power.
The In re Bilski majority opinion (and some of the minority opinions too) is/are abominations against science and rational thinking. How could a re-entrant shift register (the one in Benson Claim 8) not be a particular machine? One gets the notion that the Bilski majority opinion was a political hack (a "compromise") attempting to get to what they, the majority thought was "justice", by any means possible, including turning blind eyes to clear as day contradictions. It did not take long for the Bilski irrationality to explode across the scene in cases like the present Ex parte Koo. We all know that Koo intended for this to be a machine-implemented process for optimizing a query. OK, so he didn't use some magic form-over substance words. So shoot him, right? Wrong. It's a miscarriage of justice. The first of many to come.

Posted by: step back at December 24, 2008 10:43 AM

How about this:

1. A method for optimizing a query for a relational database management system, the method comprising:

receiving the query in a computer system implementing an optimization algorithm, the optimization algorithm:

evaluating the query to determine: whether a sub-expression of the query is being joined to itself and whether a predicate of the query comprises an equality test between a same column of the sub-expression;

determining whether a first row set producible from a first set of references of the query to the sub-expression is subsumed by a second row set producible from a second set of references of the query to the subexpression; and

reforming the query to eliminate the joining of the sub-expression to itself based on: (i) evaluation of the query, and (ii) determination of whether the first row set is subsumed by the second row set.

Posted by: Defector at December 24, 2008 10:56 AM

Defector,

No need to go through all these extreme contortions. Instead, insert before the "comprising" transition something like: "said method being implemented by a digital machine and comprising:". Now the reason we would use "digital machine" rather than just machine (or machine-implemented) is because the Bilski court said it must be a "particular" machine. A digital machine is a particular machine.

Happy Holidays to all
And to all a Bilski-less New Year :-)

Posted by: step back at December 24, 2008 11:57 AM

step back,

Did "digital machine" save Benson?

Posted by: breadcrumbs at December 24, 2008 12:44 PM

...or should I ask "Would..." (if Benson were before the CAFC today)?

Posted by: breadcrumbs at December 24, 2008 12:45 PM

It "should".
But "would" is a whole other question because we may be dealing with wooden minds :-)
Have you seen today's Sundance decision by Moore J. et al?

Posted by: step back at December 24, 2008 1:10 PM

step back,

Do you have a link?

Posted by: breadcrumbs at December 24, 2008 3:28 PM

Posted by: linky for Sundance at December 24, 2008 4:27 PM

Merry Christmas, everyone.

step back,

I have reviewed the decision in the link provided. The case seems pretty straight forward. Was there a particular facet that intrigued you?

Posted by: breadcrumbs at December 25, 2008 6:38 AM

Merry Christmas to all!

Posted by: S. Nicholas at December 25, 2008 6:49 AM

Breadcrumbs,

First off, Happy Holidays.

Second off, yes, Judge Moore talking about how only a person with experience in the truck covers art can expound on what is obvious to the ordinary artisan, and then substituting herself in (a person with no experience of record in said arts) ahead of the district court judge, and making the call. The district court judge is the one who had first hand exposure to all the witnesses and could best judge credibility and apparent experience with the art. An appeal court is not supposed to act as a trial court. Yet the CAFC does it time and again, with no shame. (Forgive them --for they know exactly what they are doing.) Cheers.

Posted by: step back at December 25, 2008 10:54 AM

Nice point step back...

Except,(perhaps)...

There is a distinction or two being made.

1) A correction was made as to the propriety of the "expert" witness. I thought this was a fair reading and an appropriate correction to what may be considered unethical posturing by Covington & Burling LLP to pose as expert someone who clearly was not.

2) Notwithstanding the lack of "expert" witness, and the dubious posturing by counsel, I believe justice was served by the CAFC making the appropriate call (de novo review) as a matter of law regarding the claims. I do not see a fault in their logic (concerning obvious).

Perhaps there IS an issue with (and definitely the perception of an issue exists):

"we conclude, however, that no such testimony was required because there are no underlying factual issues in dispute as to obviousness. The technology is simple and neither party claims that expert testimony is required to support such a holding."

and

"The factual inquiries underlying our determination of obviousness are not in material dispute regardless of the admissibility of the testimony of Mr. Bliss."

It does appear at first blush that the judges make much ado about Mr. Bliss lacking the "required" expertise for his testimony to have anything but a misleading effect; and then, seemingly claim that NO expertise is actually "required" because "the technology is simple".

Without actaully being at the trial, it can be tempting to take this somewhat contradictory result and apply it in a less than complimentary light.

Perhaps as you indicate, something is afoul. Perhaps, the lack of any dispute about the underlying facts means that the "expert" simply was not necessary, and the panel is NOT claiming to be 'experts' in this particular (simple technology) art.

Posted by: breadcrumbs at December 25, 2008 1:24 PM

Breadcrumbs,

Take a look at IPBiz's take on Sundance:
"Contemplate the CAFC line: Secondary considerations of nonobviousness—considered here by the district court—simply cannot overcome this strong prima facie case of obviousness. [?]"

Posted by: step back at December 25, 2008 3:57 PM

Thanks step back,

After reading IPBiz's take, I'm uncertain of the issue you see. It is difficult to come to a conclusion regarding the secondary considerations when I haven't taken them in the context that the court has.

I still don't see holes in the logic as applied by the court (except as possibly posted above).

Posted by: breadcrumbs at December 26, 2008 8:34 AM