The Patent Quality Arguments of Ocean Tomo
IPBiz mentioned the paper of Ocean Tomo personnel James Malackowski and Jonathan Barney in posts on Sunday, July 13 and on Friday, July 18.
The post of July 13 noted text by Joff Wild of IAM concerning an appropriate response to the Ocean Tomo paper from hypothetical critics-->
However, what the critics would be better off doing is to read through the whole article very carefully to work out whether the research that the two authors have done is flawed. If it is, they need to explain why and how the flaws distort the real picture. In other words, they need to address the message rather than the messengers.
While accepting the general premise that those advocating a decline in patent quality have not presented good arguments, IPBiz notes problems in the Ocean Tomo position advocating an increase in patent quality. In the following, IPBiz notes some deficiencies in the Ocean Tomo paper-->
The first sentence of “Patent Quality” states:
The question “What is Patent Quality?” is
one that has long been answered not by the
authors of this paper but by the marketplace
itself.
While IPBiz would agree with the general concept that arguments made by certain
legal academics of a lack of patent “quality” are flawed, IPBiz finds a lack of depth in
the arguments from the Ocean Tomo people.
A rather obvious and eye-catching flaw with the paper is the citation practice. While the authors make out well, in fact better than the US Supreme Court does, the shoddy worksmanship does attest to a lack of quality in the paper. Footnote 20, a self-citation to author Barney, illustrates that the authors are capable of doing a complete citation:
J. Barney, “A Study of Patent Mortality Rates: Using Statistical Survival Analysis to Rate and Value Patent Assets,” AIPLA Quarterly Journal, 30, no. 3, pp. 317—352 (2002).
However, one sees in footnote 10 that the Supreme Court does not fare as well:
KSR International Co. v. Teleflex, Inc. (04-1350), U.S.
Supreme Court (Decided April 30, 2007).
In July 2008, there is no excuse to cite the April 2007 KSR case in such a way. Academics are also victimized by the poor citation style of the authors:
Footnote 9: See, e.g., C. Shapiro, Navigating the Patent Thicket—Cross-
Licenses, Patent Pools and Standard-Setting, University of Cali-
fornia at Berkeley (March 2001).
Footnote 19: K. Moore, “Worthless Patents” (July 2004). George
Mason Law & Economics Research Paper No. 04-29
[Carelessness is also shown at the end, with the final section titled: Conclustion]
X
More problematic than the bad citing is the false citing. Of the patent allowance discussion,
initiated by Quillen and Webster, the authors paraphrase the Quillen and Webster argument without citing Quillen and Webster:
For one thing, it turns out that patent
allowance rates are devilishly difficult to define given
the common practice in the U.S. of filing multiple
continuation applications attempting to patent essen-
tially the same invention that may already have been
previously rejected by the Patent Office.
X
Aside from not crediting Quillen and Webster, the authors don’t discuss the various
papers showing that continuation applications are generally not multiple attempts to patent an already rejected invention. So, in this case, the authors are wrong on the cite and are wrong on the substance.
Of the patent quality debate itself, the authors identify two themes:
The first broad theme centers on the concept of an
overburdened patent office that is understaffed and
ill-equipped to maintain pace with burgeoning new
technologies and a perceived flood of new patent fil-
ings. This argument is usually heralded by a shocking,
if not amusing, parade of patent horribles—bogus,
dubious or outright wacky patents—which are of-
fered up as evidence of a patent system run amuck.
A particular parade favorite these days is the now
infamous ‘peanut-butter and jelly sandwich’ patent
(U.S. Pat. No. 6,004,596).
[The authors neglect to note that the peanut-butter and jelly sandwich died
as a result of re-examination.]
A more apt theme, addressed later on below, argues that the patentability
standards and/or applications of the patentability
standards by the Patent Office are too low, resulting
in the proliferation of many dubious patents on little
more than trivial improvements.
X
Of the first theme, the authors write:
This by now familiar rhetoric and anecdotal
evidence seems logical enough, even compelling at
times. But, several key underlying numerical and
statistical trends simply do not support the asserted
conclusion that the Patent Office is overburdened or
that quality of U.S. patent examination is flagging.
In fact, some of the statistical evidence we have
analyzed seems to suggest that patent examination
quality has either remained steady or has even im-
proved somewhat over the past five years.
X
The authors point out that current % increases in patent applications are comparable to some increases in past years. Of the pendency problem, the authors write:
In fact, a growing backlog of cases under peak
usage conditions is the expected and desired result
of a patent examination system that is maintaining
standards and quality controls at steady levels.
X
Of course, a growing backlog can arise for other reasons.
The authors note arguments that overburdened examiners may lead to
poor patent quality:
Examiners, the argument goes, being
highly pressed for time and resources, do not have
adequate opportunity to conduct sufficiently thor-
ough prior art .
The authors attempt to negate this by observing an increase in cited
art:
While this significant growth in prior art citations
may not be conclusive evidence of increased search
thoroughness or search quality, it is certainly sug-
gestive evidence and is plainly inconsistent with
the notion that search thoroughness and search
quality have somehow declined in recent years.
The authors neglect to mention that cited art on an issued patent can
arise from two sources: the examiner (892 form) and the applicant (1449 form).
In the absence of knowing “which” source has created the “significant” growth, conclusions that examiners are conducting more thorough searches are unwarranted.
The authors address arguments asserting the granting of overly broad claims:
Examiners granting patents with overly broad
claims is another commonly cited anecdotal factor
in support of the flagging quality and overburdened-
patent-office argument.
The authors attempt to negate this by comparing the average word count in an
independent claim in 2007 (160.1) to the word count in 2003 (153.3). This increase in average
word count in 2007 is supposed to indicate that claims are becoming more complicated (less broad).
The validity of this correlation is not proved.
As to impact of KSR, the authors suggest that KSR caused the recent drop in patent allowance rate:
Patent allowance rates for the
first quarter of 2008 (the first post-KSR statistics
available) have dropped to a new low of 44 per-
cent.
But patent allowance rates had been dropping before KSR.
Speaking of KSR, the authors alluded to common sense:
Even common sense tells us that the most ubiquitously
available search tools today (e.g. Google®) can access
references that a decade ago would have remained
undiscovered. The statistical data appears consis-
tent with the conclusion that examination search
thoroughness and search quality is as good or better
today than it was five years ago.
In the sense that Google has increased access to 19th century
books, such as those on the U.S. Civil War, IPBiz would agree. However,
Google and Google Scholar have not really allowed access to 20th
century science papers that otherwise would have remained undiscovered.
There were plenty of science databases ten years ago, including CA and STN,
that permitted routine access to the scientific literature. So "common sense"
does not lead IPBiz to the conclusion of the Ocean Tomo authors. Further, in 2008,
searching the internet for science papers leads an investigator typically to a
request for payment, not directly to the publication. As IPBiz recently noted, the
journal Science requires people to pay to view mere "letters to the editor."
[See Biofuels, land and fire.]
There is another problem with internet searching. Information changes
{ Internet Publishing: Online Today, But What About Tomorrow Or Where Have You Gone, 406,302? } or simply vanishes [the Sikahema effect].
The Ocean Tomo authors have presented no convincing arguments to support
the proposition that examination search thoroughness and search quality is as good or better
today than it was five years ago. At best, they have equated "more" with "better", and have
not even identified the source of "more."
There are flaws in the article which may distort the true picture.
The post of July 13 noted text by Joff Wild of IAM concerning an appropriate response to the Ocean Tomo paper from hypothetical critics-->
However, what the critics would be better off doing is to read through the whole article very carefully to work out whether the research that the two authors have done is flawed. If it is, they need to explain why and how the flaws distort the real picture. In other words, they need to address the message rather than the messengers.
While accepting the general premise that those advocating a decline in patent quality have not presented good arguments, IPBiz notes problems in the Ocean Tomo position advocating an increase in patent quality. In the following, IPBiz notes some deficiencies in the Ocean Tomo paper-->
The first sentence of “Patent Quality” states:
The question “What is Patent Quality?” is
one that has long been answered not by the
authors of this paper but by the marketplace
itself.
While IPBiz would agree with the general concept that arguments made by certain
legal academics of a lack of patent “quality” are flawed, IPBiz finds a lack of depth in
the arguments from the Ocean Tomo people.
A rather obvious and eye-catching flaw with the paper is the citation practice. While the authors make out well, in fact better than the US Supreme Court does, the shoddy worksmanship does attest to a lack of quality in the paper. Footnote 20, a self-citation to author Barney, illustrates that the authors are capable of doing a complete citation:
J. Barney, “A Study of Patent Mortality Rates: Using Statistical Survival Analysis to Rate and Value Patent Assets,” AIPLA Quarterly Journal, 30, no. 3, pp. 317—352 (2002).
However, one sees in footnote 10 that the Supreme Court does not fare as well:
KSR International Co. v. Teleflex, Inc. (04-1350), U.S.
Supreme Court (Decided April 30, 2007).
In July 2008, there is no excuse to cite the April 2007 KSR case in such a way. Academics are also victimized by the poor citation style of the authors:
Footnote 9: See, e.g., C. Shapiro, Navigating the Patent Thicket—Cross-
Licenses, Patent Pools and Standard-Setting, University of Cali-
fornia at Berkeley (March 2001).
Footnote 19: K. Moore, “Worthless Patents” (July 2004). George
Mason Law & Economics Research Paper No. 04-29
[Carelessness is also shown at the end, with the final section titled: Conclustion]
X
More problematic than the bad citing is the false citing. Of the patent allowance discussion,
initiated by Quillen and Webster, the authors paraphrase the Quillen and Webster argument without citing Quillen and Webster:
For one thing, it turns out that patent
allowance rates are devilishly difficult to define given
the common practice in the U.S. of filing multiple
continuation applications attempting to patent essen-
tially the same invention that may already have been
previously rejected by the Patent Office.
X
Aside from not crediting Quillen and Webster, the authors don’t discuss the various
papers showing that continuation applications are generally not multiple attempts to patent an already rejected invention. So, in this case, the authors are wrong on the cite and are wrong on the substance.
Of the patent quality debate itself, the authors identify two themes:
The first broad theme centers on the concept of an
overburdened patent office that is understaffed and
ill-equipped to maintain pace with burgeoning new
technologies and a perceived flood of new patent fil-
ings. This argument is usually heralded by a shocking,
if not amusing, parade of patent horribles—bogus,
dubious or outright wacky patents—which are of-
fered up as evidence of a patent system run amuck.
A particular parade favorite these days is the now
infamous ‘peanut-butter and jelly sandwich’ patent
(U.S. Pat. No. 6,004,596).
[The authors neglect to note that the peanut-butter and jelly sandwich died
as a result of re-examination.]
A more apt theme, addressed later on below, argues that the patentability
standards and/or applications of the patentability
standards by the Patent Office are too low, resulting
in the proliferation of many dubious patents on little
more than trivial improvements.
X
Of the first theme, the authors write:
This by now familiar rhetoric and anecdotal
evidence seems logical enough, even compelling at
times. But, several key underlying numerical and
statistical trends simply do not support the asserted
conclusion that the Patent Office is overburdened or
that quality of U.S. patent examination is flagging.
In fact, some of the statistical evidence we have
analyzed seems to suggest that patent examination
quality has either remained steady or has even im-
proved somewhat over the past five years.
X
The authors point out that current % increases in patent applications are comparable to some increases in past years. Of the pendency problem, the authors write:
In fact, a growing backlog of cases under peak
usage conditions is the expected and desired result
of a patent examination system that is maintaining
standards and quality controls at steady levels.
X
Of course, a growing backlog can arise for other reasons.
The authors note arguments that overburdened examiners may lead to
poor patent quality:
Examiners, the argument goes, being
highly pressed for time and resources, do not have
adequate opportunity to conduct sufficiently thor-
ough prior art .
The authors attempt to negate this by observing an increase in cited
art:
While this significant growth in prior art citations
may not be conclusive evidence of increased search
thoroughness or search quality, it is certainly sug-
gestive evidence and is plainly inconsistent with
the notion that search thoroughness and search
quality have somehow declined in recent years.
The authors neglect to mention that cited art on an issued patent can
arise from two sources: the examiner (892 form) and the applicant (1449 form).
In the absence of knowing “which” source has created the “significant” growth, conclusions that examiners are conducting more thorough searches are unwarranted.
The authors address arguments asserting the granting of overly broad claims:
Examiners granting patents with overly broad
claims is another commonly cited anecdotal factor
in support of the flagging quality and overburdened-
patent-office argument.
The authors attempt to negate this by comparing the average word count in an
independent claim in 2007 (160.1) to the word count in 2003 (153.3). This increase in average
word count in 2007 is supposed to indicate that claims are becoming more complicated (less broad).
The validity of this correlation is not proved.
As to impact of KSR, the authors suggest that KSR caused the recent drop in patent allowance rate:
Patent allowance rates for the
first quarter of 2008 (the first post-KSR statistics
available) have dropped to a new low of 44 per-
cent.
But patent allowance rates had been dropping before KSR.
Speaking of KSR, the authors alluded to common sense:
Even common sense tells us that the most ubiquitously
available search tools today (e.g. Google®) can access
references that a decade ago would have remained
undiscovered. The statistical data appears consis-
tent with the conclusion that examination search
thoroughness and search quality is as good or better
today than it was five years ago.
In the sense that Google has increased access to 19th century
books, such as those on the U.S. Civil War, IPBiz would agree. However,
Google and Google Scholar have not really allowed access to 20th
century science papers that otherwise would have remained undiscovered.
There were plenty of science databases ten years ago, including CA and STN,
that permitted routine access to the scientific literature. So "common sense"
does not lead IPBiz to the conclusion of the Ocean Tomo authors. Further, in 2008,
searching the internet for science papers leads an investigator typically to a
request for payment, not directly to the publication. As IPBiz recently noted, the
journal Science requires people to pay to view mere "letters to the editor."
[See Biofuels, land and fire.]
There is another problem with internet searching. Information changes
{ Internet Publishing: Online Today, But What About Tomorrow Or Where Have You Gone, 406,302? } or simply vanishes [the Sikahema effect].
The Ocean Tomo authors have presented no convincing arguments to support
the proposition that examination search thoroughness and search quality is as good or better
today than it was five years ago. At best, they have equated "more" with "better", and have
not even identified the source of "more."
There are flaws in the article which may distort the true picture.
posted by Lawrence B. Ebert at 2:54 PM
0 Comments:
Post a Comment
<< Home