Proposal3

Part III: The “Peer to Patent” Community Patent Reform Proposal

While there are those who would advocate doing away with patents altogether, this would require both a constitutional amendment and, at the same time, risk eliminating an important fundament of national competitiveness.  The Community Patent idea is a realistic alternative for reform that requires minimal statutory change while narrowing the gaps in the patent system’s filter: it increases the likelihood that good inventions will pass through while blocking unworthy inventions.  This section outlines that proposal in detail, describing both the process and the technology that might enable it.

The peer-to-patent system relies on well-known social software techniques to manage expert scientific juries and to advise on patent applications. These experts advise on prior art and can also be used to assess the utility, novelty, non-obviousness and enablement of inventions; with many pairs of eyes applied to the problem instead of one they are able, collectively, to give applications a thorough and meaningful review with the benefit of a much wider array of collective experience and intelligence.

The system replaces review by a lone examiner with peer production of review by on-line experts.  As we know from other examples of peer produced and collaborative information review and filtering, any peer production system – whether on-line or face-to-face – has to meet several criteria in order to enable the group to work together on solving the problem assigned to it.[1]  Members of the collective require a clear sense of the goal or the problem to be solved.  They have to be able to divvy up tasks in connection with solving that problem and achieving the goal.  Those tasks, ideally, should be modular and short, rather than requiring an extended time investment.  This increases the likelihood of participation.  Participants in a peer production system, unlike in a firm or other hierarchical organization, need to be able to self-assign their tasks or roles in the group based on their own estimation of expertise.[2]  Far better for me to designate what I am good at since I am in the best position to have this information.[3]  The group needs to have a clear awareness of the tasks and roles of participants in the group in order for the collective to function well together.  Experience with group and organizational dynamics, whether on-line or off, shows that peer production functions where successful participation confers status upon members of that group.  Hence the group needs to evolve mechanisms, which may be cultural, of conferring status on those people who participate well and shoulder their burden with regard to the goal of the group.[4]

This experience tells us that we need a modular system implemented through software that creates groups to review patents, breaks down their work into clear stages and tasks, which they can assign to themselves and where they can participate without an undue burden.

The system demands a means for participants to rate each other and their participation.  The system comprises four stages:

• electronic filing
• prior art advising
• patentability advising
• final determination

Electronic Filing

The first step is on-line filing.  This provides the opportunity to optimize inputs into the system to facilitate peer review.  Patent applications are already electronically filed[5] and sorted according to the patent classification system (USPC) that categorizes based on the classes of technology claimed in the patents.[6] The USPC ensures that applications can be routed to the right examiner.

Under a peer production system, applications could be “tagged” or labeled, not only according to the imposed classification scheme, but also by the community.  This kind of community self-tagging – or what is sometimes called a “folks-onomy”[7] – might make it easier to ensure that experts can later self-select to examine inventions in their area of expertise. Such a folksonomy might make labeling more granular and precise to speed up the process of self-assignment.[8]  We already have tagging and labeling software that allows Internet users to label content for easier retrieval, indexing and searching.[9]

The application needs to include a short summary of no more than a paragraph, again, to facilitate review.  This only requires providing, in addition to the written description,[10] a short, plain-English abstract to describe the invention. 

Once received the patent coordinator has twenty-one days to review the application for completeness and publish the application for peer review. This is a perfunctory check to make sure all parts of the application have been completed and contain the necessary information.  It is also an opportunity to ensure that the application relates to patentable subject matter.  Where the patent examiner is unsure of that determination, the application publishes to the system for review.

The patent office website will allow participants in the peer review system to subscribe to RSS (really simple syndication)[11] feeds to receive the name and abstract of any new inventions filed, allowing them to know that an opportunity has arisen for review in their area of expertise.

Once published, the peer review process to propose relevant prior art and assess novelty commences.  Eventually, parallel peer review processes to assess both novelty and obviousness would take place over the next three to five months, drastically accelerating the process of review from the average  2-4 years[12] that the patent office currently requires to perform patent review.  By speeding up the review process, we also speed up the time for scientists to publish and publicly discuss innovation without the fear of triggering a statutory bar.[13]  While the normal process will require three months, additional weeks may be added where a specific question arises from the peer jury requiring back-and-forth information exchange with the inventor managed by the patent coordinator as interlocutor.

Prior Art and Novelty Review

The first significant and substantive area where peer review can be of use is the determination of novelty as required by the statute.  The law outlines a series of earlier types of invention or publication, whether by a third party or by the inventor himself, that can preclude patentability.[14]  Essentially, the statute inquires whether the invention is new as compared to the prior art.  Prior to filing for this patent was there a prior patent?  Prior to filing for this patent was there a prior printed publication that defeated the uniqueness and newness of the claimed invention?  Prior to filing for this patent was there a prior public use?  The statute seeks to ensure that, not only is the invention new with regard to what came before, but that the inventor did not sleep on his rights[15] by failing to file an application more than one year after publicly promoting the invention.

It is illogical to have one person – with access to limited information – determining novelty when we can harness the collective intelligence and experience of thousands to answer these questions. Many technological advances are not described in commonly available academic publications or those sources to which the patent examiner has easy access. It is also illogical to turn to  a single or even a handful of private firms to conduct this review when those with the deepest experience in any given area of innovation and bring their expertise to bear.[16] 

The novelty determination is ideally-suited to peer review because it enunciates a clear goal, requires only minimal participation to answer and lends itself to self-selection on the basis of expertise.  While a patent examiner might have to search for prior art for hours, an expert knows instantly whether an invention is reminiscent of earlier work or avenues of research.  Designed right, the software can make participation for a network of scientific and innovation experts clear and easy.

As Eric Raymond, hacker “anthropologist” explained, with many eyeballs “all bugs are shallow.”  Just as a community of open source programmers is well-suited to spotting mistakes in code, the peer to patent community is equipped to address whether an invention is novel or resembles something seen before. 

A prior art novelty review is an opportunity to get more public input into the patent system and introduce citizen consultation, the common and legally-required practice of every government agency,[17] into the intellectual property review process.[18]  This is akin to the practice of scientific peer review, which is not only practiced in the private sector, but in common use in government as well.[19]  The National Science Foundation and the National Institutes of Health both use peer reviewers to determine if research is novel and represents a contribution to its field.[20] 

The novelty peer review should solicit input in both quantifiable as well as qualitative form in order to provide information that is both useful and manageable for the patent coordinator.  This requires taking advantage of graphical and visual aids to make it easier for the expert to participate faster and for the group and the patent coordinator to make sense of the comment. While participation can take the form of text-based comments on prior art it can also be expressed by means of sliders to show the expert’s opinion as to the invention’s novelty. Push-pins can be placed on an electronic map of innovation to indicate where the invention sits with regard to other inventions that have come before.[21] 

The system will also need to promote “lonely patents” by advertising under-subscribed patents for review to the network of experts.  Experts will receive RSS or email notifications of patents awaiting reviewing.  Reviewing one patent application will generate a prompt: “Would you like to review another?”  Again, if the system is designed to optimize inputs and facilitate participation, it can reduce the burden of reviewing a patent for novelty and commenting on prior art.  The National Science Foundation currently relies on a network of over 50,000 reviewers.[22]  The National Institutes of Health relies on outside review groups and advisory councils from the scientific community to review over 70% of its applications.[23]  The Environmental Protection Agency grant selection process relies heavily upon “Science Review Panels” which are peer review groups chosen and managed by an outside scientist.[24]

Crucial to the design of the system is a social reputation scheme whereby participants rate each other and their participation in the process.  The social reputation scheme will form the cornerstone of the next step: advising on the patentability determination.

Determining Obviousness and Enablement

While the novelty review demands the short attention of many eyeballs to identify whether something is new vis-à-vis what has come before and suggest relevant prior art to the patent examiner, the process of determining whether an invention is truly innovative and represents the “flash of creative genius”[25] requires a more probing and thorough examination that will require the advice and assistance of a smaller group of experts over a longer period of time.

The obviousness determination is the core of the patent examination.[26]  The standard of non-obviousness is not how hard the inventor worked to obtain the invention but if “the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art[27] to which said subject matter pertains.”[28]  Enablement[29] requires that the inventor clearly describe the invention to give the public the benefit of the knowledge obtained and to clearly delimit the metes and bounds of the patent.

The obviousness and enablement review can commence three weeks into the novelty review to allow time to sort out blatantly unqualified applications while ensuring rapid completion of the peer review process.  Obviousness and enablement review will comprise six weeks of on-line “expert jury” counsel to the examiner by smaller groups of 12 to 24 persons having “ordinary skill in the art” as required by law and assisting the USPTO with the review.

We start from the proposition that convening such a jury will depend upon having a social reputation system in place to select the members of the jury who demonstrate the ordinary skill in the art.  We know from other experience with on-line commons-based peer production projects[30] as well as off-line group organizational projects[31] that the ability to measure and communicate status is crucial to fostering participation.[32]  Status and reputation are essential to building the trust in the community necessary for iterative interaction.[33]  It is also crucial for determining qualifications for participation and for creating an incentive to ongoing collaboration.  The currently prevailing wisdom is that social reputation software is the way to foster and find such expertise.

Developing a robust software system to support a social reputation system requires innovating and adapting from existing reputation tools, such as those pioneered by EBay.[34]  We imagine the system will start by allowing experts to self-assign their own areas of expertise.  I am a materials scientist.  I am a derivatives expert.  I am a bridge engineer.  I know about Expressed Sequence Tags or retailing inventions or plastics. The expert completes a biographical form with a list of publications, research and grants received, as applicable.  The biographical form will be customized for different communities to capture that information which is most relevant to assessing knowledge in that industry.  In addition to manually inputted information, the software will keep track of each person’s participation in the system and the ratings accorded by others to his postings (taking a lesson from Slashdot).

Every time the expert logs on, he will be asked to rate the participation of others in the network.  He will rate on two scales: rating the person and rating his participation.  The first rating system will identify the person to be rated by name, allowing the rater to view that person’s profile.  This first measure will assess the person’s standing in the community.  The collaborative rating can be combined with other automated status rankings.  In other words, the community in that industry may decide that certain prizes or grants or educational qualifications or expertise count for added points.   This ranking system works like other social software tools, such as Friendster, that permit direct ranking and rating of members of the network.  Hence the community, rather than rating the individual, can determine the rating criteria according to which the system should do the rating.

In addition to rating the person, the system will also allow the community to rate participation in the community patent process as an incentive to participation. We are assuming that participants will rate postings divorced of personally identifiable information. Each time a user logs on to the system (or with regular frequency via email), he will be asked to review another person’s comment in the same way that Amazon asks readers to assess “was this book review helpful” or Netflix asks movie watchers to rate the reviews provided by other users.  The system will provide positive feedback for more and better participation in much the same way that eBay provides feedback points and other sites, such as Slashdot, assign positive karma to those who participate well.

This two-track status system will take account of reputation and participation.  It will need to be complex and dynamically evolving to keep track of new information provided by participants about measures of reputation within that sector. 

Each participant rates as many participants as he wants with the system regularly prompting the user to rate more participants and more postings.  Participation might require a minimum of three ratings and incentives can be built into the software, as Slashdot does,[35] to encourage ongoing rating and ranking.

As with any good social reputation system, the software should show the person being rated the evolving information about his reputation.  Being able to visualize one’s standing in the community creates further incentive to constructive participation. It also allows members of the group of experts to see the group and sense the community of which they are a part: who is in that group, who is participating, how much and in what ways.

These visualization techniques can both be applied to the community of experts and used to enable them to participate more easily.  In addition to the map of the community, on-screen visual maps of the subject matter of that community can help inventors to locate the inventions they are reviewing within the context of the prior art.  One project currently under development[36] uses natural language search technologies to build a visual map of the landscape of innovation.  With such data readily available, the peer to patent system should provide a map of inventions with one axis representing novelty and the other access representing how related inventions are to each other.

Together, the community can create this map of innovation in particular areas of science and industry.  Each participant can rate an invention based on how much of an advance it constitutes over the prior art and how obvious or non-obvious it is.  This allows participants to comment via a visual interface quickly and easily and provides the public and industry with the benefit of more information about the state of the art.

The Final Determination

The patent examiner will make the final determination of validity with the benefit of this expert citizen participation.  While eventually, with testing and refinement and further use, we can imagine deferring decisionmaking to the group entirely, in the initial deployment, the Patent Office needs to run and coordinate the process and provide an added check on abuse of discretion.  With two competing systems in place, namely examiner review and peer review, courts have the benefit of a wider array of information in reviewing any challenged patents.  The public, peer review system provides a check on the examiner’s discretion and the examiner provides a check on any abuses by the group.

The patent examiner coordinates the back-and-forth colloquy that is essential to the patent application process, feeding the group’s questions to the inventor for refinement and response.  This preserves the rights of the inventor to hone the application for compliance with the requirements and, at the same time, preserves the public’s rights to demand more carefully and narrowly drafted specifications.[37]