More on "peer review" of patents
The boingboing blog wrote:
Patent office will ask the public to "peer review" inventions
The US Patent and Trademark Office has launched "Peer to Patent," a community patent peer review project. The USPTO is overloaded with patent filings, so it does little or no investigation into patnets [sic: patents] before rubber-stamping them, expecting that the courts will sort out who invented what. This changes the patent system from something that promotes invention to something that rewards companies who aggressively sue inventors.
Peer to Patent aims to address this by encouraging the public to review patents, to determine whether they are valid based on the at-large expert knowledge about what has already been invented and what is a new, useful, nonobvious invention. IBM has agreed to have its patents vetted by the public as a guinea pig in the project.
David A. Utter wrote:
Critics of the US Patent Office get their chance to put their expertise where their blogs are and participate in a new peer review initiative aimed at finding patents that are obvious creations and thus not qualified for a patent.
Ideally, all patented inventions would be novel and nonobvious. But the issuer of patents in the United States, the US Patent and Trademark Office, has been accused many times of rubber stamping patents that end up harming other inventions due to the patent covering some obvious function.
**
TechDirt also had some discussion of the peer to patent program:
One of the ideas that's been floated to solve the patent office's resource crisis is a system of peer reviews, much like that of an academic journal.
It appears that the theory is slowly being put into practice
since patents are only supposed to be granted on ideas "non-obvious to those skilled in the art," it makes sense to actually ask those skilled in the art for their opinions.
In comment,
--> peer to patent involves identification to the PTO of what are considered pertinent prior art references. In peer review at an academic journal (a concept apparently not understood by Mike at TechDirt) a referee is most concerned with determining that a paper's conclusions flow from the data (at least a plausible interpretation of the data) and that the data are taken under appropriate conditions. Although the topic might come up at times, a referee is generally not looking for anticipation by prior work, and is not generally going to say "don't publish because it's all been done before" or because it's "obvious." More frequently, a referee will seek inclusion of a relevant reference, not an anticipating one. Further, at a journal such as Science, a submitted article is more likely to be rejected as not being of great enough interest to readers, NOT because it's anticipated, obvious, implausible, or not "good" science. Moreover, when an issue of (missed) prior art actually does come up, journal editors will go into all kinds of contortions to deny the prior art, making many complaints about the USPTO on "missed art" look like kindergarten chatter in comparison.
Thus, the system of "peer reviews" contemplated by the USPTO is NOT much like that of an academic journal.
--> In fact, the peer to patent system is NOT a peer REVIEW at all. The prior art, identified by the "peers," is given to the examiner, who conducts the review.
--> The legal concept of obviousness is measured against the hypothetical construct of one of ordinary skill in the art.
***
Back on February 11, I had placed on techdirt:
Some posts suggesting that peer review would be bad for the Patent Office:
http://ipbiz.blogspot.com/2006/02/scientist-asks-if-peer-review-is.html
http://ipbiz.blogspot. com/2005/12/issues-of-peer-review-co-authorship.html
http://ipbiz.blogspot.com/2005/10/2005-medicine- nobel-strong-argument.html
And, if you want to go further back than Hwang-gate, the 2005 Nobel, and Jan-Hendrik Schon, check out how the Royal Society "reviewed" the papers of Benjamin Franklin back in 1750.
***
[As a footnote to this post, the above-linked reference to an amazines article entitled "How did Woo Suk Hwang trick the scientific world about stem cells?" prepared on Feb. 18, was completely unsearchable on Google on May 19. One could type in any of the key words of the article and NOT find the article in a search. It had previously been searchable. With Google, things come and things go. Unlike the issue with the USPTO's 406,302 number, the actual amazines article is still on the amazines site (so far), but does it really exist if you can't find it with Google?]
Patent office will ask the public to "peer review" inventions
The US Patent and Trademark Office has launched "Peer to Patent," a community patent peer review project. The USPTO is overloaded with patent filings, so it does little or no investigation into patnets [sic: patents] before rubber-stamping them, expecting that the courts will sort out who invented what. This changes the patent system from something that promotes invention to something that rewards companies who aggressively sue inventors.
Peer to Patent aims to address this by encouraging the public to review patents, to determine whether they are valid based on the at-large expert knowledge about what has already been invented and what is a new, useful, nonobvious invention. IBM has agreed to have its patents vetted by the public as a guinea pig in the project.
David A. Utter wrote:
Critics of the US Patent Office get their chance to put their expertise where their blogs are and participate in a new peer review initiative aimed at finding patents that are obvious creations and thus not qualified for a patent.
Ideally, all patented inventions would be novel and nonobvious. But the issuer of patents in the United States, the US Patent and Trademark Office, has been accused many times of rubber stamping patents that end up harming other inventions due to the patent covering some obvious function.
**
TechDirt also had some discussion of the peer to patent program:
One of the ideas that's been floated to solve the patent office's resource crisis is a system of peer reviews, much like that of an academic journal.
It appears that the theory is slowly being put into practice
since patents are only supposed to be granted on ideas "non-obvious to those skilled in the art," it makes sense to actually ask those skilled in the art for their opinions.
In comment,
--> peer to patent involves identification to the PTO of what are considered pertinent prior art references. In peer review at an academic journal (a concept apparently not understood by Mike at TechDirt) a referee is most concerned with determining that a paper's conclusions flow from the data (at least a plausible interpretation of the data) and that the data are taken under appropriate conditions. Although the topic might come up at times, a referee is generally not looking for anticipation by prior work, and is not generally going to say "don't publish because it's all been done before" or because it's "obvious." More frequently, a referee will seek inclusion of a relevant reference, not an anticipating one. Further, at a journal such as Science, a submitted article is more likely to be rejected as not being of great enough interest to readers, NOT because it's anticipated, obvious, implausible, or not "good" science. Moreover, when an issue of (missed) prior art actually does come up, journal editors will go into all kinds of contortions to deny the prior art, making many complaints about the USPTO on "missed art" look like kindergarten chatter in comparison.
Thus, the system of "peer reviews" contemplated by the USPTO is NOT much like that of an academic journal.
--> In fact, the peer to patent system is NOT a peer REVIEW at all. The prior art, identified by the "peers," is given to the examiner, who conducts the review.
--> The legal concept of obviousness is measured against the hypothetical construct of one of ordinary skill in the art.
***
Back on February 11, I had placed on techdirt:
Some posts suggesting that peer review would be bad for the Patent Office:
http://ipbiz.blogspot.com/2006/02/scientist-asks-if-peer-review-is.html
http://ipbiz.blogspot. com/2005/12/issues-of-peer-review-co-authorship.html
http://ipbiz.blogspot.com/2005/10/2005-medicine- nobel-strong-argument.html
And, if you want to go further back than Hwang-gate, the 2005 Nobel, and Jan-Hendrik Schon, check out how the Royal Society "reviewed" the papers of Benjamin Franklin back in 1750.
***
[As a footnote to this post, the above-linked reference to an amazines article entitled "How did Woo Suk Hwang trick the scientific world about stem cells?" prepared on Feb. 18, was completely unsearchable on Google on May 19. One could type in any of the key words of the article and NOT find the article in a search. It had previously been searchable. With Google, things come and things go. Unlike the issue with the USPTO's 406,302 number, the actual amazines article is still on the amazines site (so far), but does it really exist if you can't find it with Google?]
posted by Lawrence B. Ebert at 6:04 AM
0 Comments:
Post a Comment
<< Home