Peer-to-Patent has released a report on the results of its one-year pilot. Click here to download the full report in pdf format.

Here are the highlights of the report from a press release today:

Launched on June 15, 2007 by New York Law School Professor Beth Noveck together with a network of corporate and academic collaborators and in cooperation with the USPTO, Peer-to-Patent is the first social networking project with a direct link to decision-making by the federal government. Under traditional practices, USPTO patent examiners bear the sole burden of identifying and relating information pertinent to patent applications. Under Peer-to-Patent, expert volunteers were permitted to assist in these efforts at the www.peertopatent.org Web site.

With the consent of participating inventors, patent applications were posted to the Peer-to-Patent site where the expert reviewers discussed the applications and submitted bibliographic information, known as prior art, relevant to determining if an invention was new and non-obvious, as the law requires to obtain a patent. At the conclusion of the review period, this prior art was forwarded to the USPTO patent examiner for consideration and use in their further search efforts.

Major companies such as IBM, Microsoft, Hewlett-Packard, Sun Microsystems, Intel, and GE, companies whose patent portfolios account for nearly one-third of the patents issued to the top 30 U.S. patent holders in 2007, all submitted patent applications to the Peer-to-Patent process. Other patent applications were submitted by Red Hat, Cisco, and Yahoo!, as well as smaller firms.

Data from the first year of the Peer-to-Patent pilot shows that an open network of reviewers can improve the quality of information available to patent examiners and that such citizen-reviewers are capable of producing information relevant to the patent examination process and are willing to volunteer time. Initial results based on a survey of patent examiners from the USPTO suggest that information provided by the public is beneficial to the examination process.

Findings from the first-year report include:

* Peer-to-Patent attracted more than 2,000 peer reviewers.

* The first 23 office actions issued during the pilot phase showed use of Peer-to-Patent submitted prior art in nine rejections

* On average, citizen-reviewers contributed 6 hours reviewing each patent application in the pilot

* Although USPTO rules permit third-party prior art submissions on pending applications, the average number of prior art submissions on Peer-to-Patent applications was 2,000 times that of standard rule-based submissions.

* Ninety-two percent of patent examiners surveyed said they would welcome examining another application with public participation, while 73% of participating examiners want to see Peer-to-Patent implemented as a regular office practice.

* 21% of participating examiners stated that prior art submitted by the Peer-to-Patent community was "inaccessible" directly to USPTO examiners.

* Prior art submissions by Peer-to-Patent reviewers were four times as likely to include non-patent literature (any document that is not a patent, including Web sites, journals, textbooks, and databases) as compared to prior art submissions by applicants.

"As the first example of harnessing public knowledge to improve a government process, the first year of Peer-to-Patent was an unquestioned success," Noveck said. She added: "While the impact of this project on patent quality will take longer to assess, the early indications are certainly promising."

We've been following the peer-to-patent project with interest and have linked here on our company weblog to several blog posts and  a YouTube video presentation describing the project and its promise.

What do you think can be learned from the Peer-to-Patent pilot program?

• Print
• Comments
• Trackbacks

EFF, the Electronic Frontier Foundation, announced on their blog today the release of Patents and the Public Domain: Improving Patent Quality Upon Reexamination a report finding that there are at least three ways to work toward improving patent quality, including through reexamination:

1. Encourage Organized Efforts to Have a More Thorough Review of Patent Applications

* Increase third party efforts to improve prior art searches and to facilitate a stronger understanding of the world of prior art.

2. Support Organized Efforts After Patents Have Been Issued

* Reexaminations should be limitless in time and with no financial harm limitations (in many countries, public interest organizations like EFF are only allowed to challenge issued patents for a short period of time after they have issued, even though often a patent’s threat to innovations may not be immediately obvious).

* Don't Streamline. Maintain a variety of options for reexamination requesters.

3. Increase Access to Information in the Patent Office

* Update information in the patent office’s online databases regularly.

* Conduct and publish regular evaluations of effectiveness and performance within the Patent Office.

In addition to providing an organized effort through the Patent Busting Project in furtherance of the second goal, EFF is working in conjunction with Mozilla, Yahoo, and the Internet Archive to create a database that will work to achieve the first goal. This Prior Art project is still under development.

The wiki page set up for this contains the project plan for a software prior art initiative under consideration and welcomes constructive input. At the bottom of that wiki page, they point to four related projects:

1. Software Patent Institute
2. Open Source as Prior Art
3. Patent Commons
4. Peer to Patent Project

The Software Patent Institute Database of Software Technologies is contained exclusively in the IP.com Prior Art Database. While limited searching can be performed for free as a "guest", robust and unrestricted searching along with all-inclusive document downloads is available in weekly, monthly, and yearly intervals. Get started searching SPI data today - click here for more information.

• Print
• Comments
• Trackbacks

The Peer-to-Patent project announced that expert, volunteer reviewers from the general public have cooperated with the United States Patent and Trademark Office (USPTO) to identify “prior art” that has been used in Office actions to reject at least one claim in each of five applications.

The Peer-to-Patent project is an initiative of New York Law School in cooperation with the USPTO aimed at opening the patent examination process to public participation, and improving the quality of patents. Under the Peer-to-Patent pilot program, inventors agree to have their patent applications posted for up to four months on the www.peertopatent.org Web site. Expert volunteers from the public then discuss the applications and submit prior art they think might be relevant to determining if an invention is new and non-obvious, as the law requires.

The program’s first 19 patent applications—including applications from GE, Hewlett-Packard, IBM, Intel, Microsoft, Red Hat, and Sun Microsystems Inc., all companies that volunteered to be part of this pilot—have been examined. For every 500 patent applications published in 2007, the USPTO received only one third-party prior art submission. In the Peer-to-Patent pilot, volunteer reviewers supplied nearly four prior art references for each pilot application - a trend which could significantly help the USPTO if Peer-to-Patent expands beyond a pilot phase.

So far, of the first 19 first office actions sent by the USPTO, five patent applications received non-final rejections that relied specifically upon prior art submitted through Peer-to-Patent. In these instances, submissions of prior art from the public enabled the USPTO to reject claims the examiners believe fail the novelty and non-obvious requirements that prohibit patents on inventions that have already been invented or would have been obvious. The USPTO response time on these office actions is notable. Because the USPTO agreed to examine patent applications in the pilot ahead of other applications, the time between application filing and the onset of examination shrank from four to two years.

USPTO’s Commissioner for Patents John Doll noted, “I hope other patent applicants look at the processing statistics from this pilot program and realize that Peer-to-Patent review might be a win-win situation for them. We are encouraged by the initial success of the pilot, and we believe it holds potential as a source of relevant prior art.”

In sending the program’s first non-final rejection, the USPTO examiner used prior art and commentary submitted by Steven Pearson, a senior software engineer at IBM, to reject claims of an HP application. The second non-final rejection relied upon prior art and commentary submitted by Rob Cameron, a Professor of Computer Science at Simon Fraser University, to reject claims of an IBM application. As a result, Pearson and Cameron have been awarded the title “prior artist” on the pilot Web site.

“It is a privilege to participate in this important project,” Steve Pearson, IBM software engineer, said. “I’m confident these early results will help validate that this community approach can have a meaningful impact on the examination process and the quality of patents. Hopefully this will encourage the participation of more domain experts in this pilot program.”

Professor Cameron believes “it is to everyone’s benefit—inventor, investor and the public at large—to make the best possible effort to ensure that issued patents are properly placed in the context of related prior art. In the future, I think that an open, mediated review process following the trail blazed by Peer-to-Patent should become an integral part of best practice patent examination.”

HP and IBM will now have the opportunity to respond to these first office actions and persuade the examiners their claims are new and non-obvious.

“We’re very pleased with this initial outcome,” Manny Schecter, Associate General Counsel for Intellectual Property at IBM said. “Patents of questionable merit are of little value to anyone. We much prefer that the best prior art be identified so that the resulting patent is truly bulletproof. This is precisely why we eagerly agreed to sponsor this project and other patent quality initiatives. We are proud of this result, which validates the concept of Peer-to-Patent, and can only improve the quality of patents produced by the patent system.”

Launched in June of last year, Peer-to-Patent opens the patent examination process for the first time, enabling the public to contribute prior art and commentary relevant to the examination of pending patent applications assigned in TC 2100 technologies with the goal of improving patent quality. On average, each posted application has garnered a community of 14 reviewers who have submitted five instances of prior art per application.

See also: www.peertopatent.org

• Print
• Comments
• Trackbacks