The innovation aims to develop a patent-valuation system that considers the totality of a United States patent's ?prior-art? patent-citations. Using network mathematics of eigenvector centrality, a patent's intrinsic value is calculated based on recursively weighting associations between a patent and other patents (its backward and forward citations). This methodology establishes the values of the patents relative to each other so we have a uniform evaluation standard for all patents in the United States Patent & Trademark Office (USPTO). This algorithm has the potential to bring order to the patent-valuation problem. At a particular point in time, every patent within the entire universe of United States patents receives a score which can be ranked in a myriad of ways, such as: within a respective patent portfolio (a firm portfolio, an inventor portfolio, a patent-agent portfolio), within a technology classification (e.g., 705/310: DATA PROCESSING/Intellectual Property Management), universally within the entire network, and so on. Monitoring the diffusion of a patent's marginal score over time (i.e., a patent's Schumpeterian shock) also makes it possible to predict a potential lifetime value of a patent.
The broader/commercial impact of this project is to develop an objective metric for the value of patent innovations (i.e., to deliver transparent and objective patent-network dynamics and in the long term become like the credit-rating FICO® score for patents). The successful completion of this project would provide additional tools and products for those engaged in intellectual-property management, specifically the patent holders and other organizations that strive to estimate a patent's current and potential value. This meta-innovation about technology will benefit all facets of commercialization ecosystems, including: identification of viable entrepreneurial opportunities, assessment of internal/external technological venturing, evaluation of patent-assertion prospects, and overall management of radical innovations.
Patent Rank is an objective, transparent measure of a patent's intrinsic value. Using eigenvector network centrality of "prior-art" citations, we can define the value of a patent, monitor its value over time, and predict its lifetime value. Similary, we can monitor the value of other entities/actors in the network (e.g., inventors, firms, technologies, lawyers, examiners). Hence, Patent Rank can measure network dynamics. During Phase I/I(b), the technical feasibility has been established, market response has been favorable, and market conversations have helped refine our roadmap for continued R/R&D toward commercialization. We have developed a brand image (Patent-Rank.com) and an early-stage prototype (MyPatentAssets.com) to enable us to have meaningful conversations in the market while demonstrating the utility of the technological innovation Patent Rank. We compete in patent-valuation services, a subset of patent analytics which requires a patent-data platform. (Analytics can only be as good as your data - garbage in, garbage out). We are a fiscally-bootstrapped, entrant startup that initially will target a un(der)served "value-novice" market while establishing credibility in the more sophisticated "intellectual-asset management" market. We have received external funding from Kentucky's Innovation Network to support market development, patent development, data-platform development, additional R&D, rent, and equipment purchases. We anticipated receiving an additional $500K+ of similar funding if we are awarded Phase II. Disappointingly, we have gone "all-in" with this venture and were notified that will will not receive the Phase II award. We believe this decision was unwarranted based on the obtuse comments of the review team, and we outlined the concerns of the review process through a formal appeal process. The appeal was declined by the NSF. We were also very disappointed with the appeal process as seven facts were presented to the appeal team, and the reply failed to address these factual concerns. In sum, the seed funding of the Phase I grant has helped move this ambitious project forward. However, the inability of the NSF to correctly ascertain the potential of this novel "meta-innovation" has been very disappointing. Intellectual Merit: As Google's PageRank algorithm has brought 'order to the web,' we propose that Patent Rank will bring 'transparency to intellectual property.' This meta-innovation objectively answers the question "what is the value of a patent" but peripherally can answer other important questions: "who are key actors in an innovation ecosystem, what patent is most closely related to my product/patent/idea, what are trends in innovation ecosystems, what are financial implications on patent innovations in the market?" Leading mathematicians and computer scientists in network theory are developing artificial-intelligence algorithms (eigenvector network centrality, Shannon's entropy, latent semantic analysis, record linkage) to "measure network dynamics" which, coupled with statistical rigor, will enable knowledge workers in intellectual-property management to make more informed decisions while reducing asymmetry of information in the market. Qualified inference requires a comprehensive data platform achieved by integrating and scrubbing BIG DATA from many sources. Our refined algorithms will: (1) capture a patent's intrinsic value SOONER by including application citations; (2) monitor the VALUE of ACTORS in the innovation ecosystem while controlling for any propensity BIASES; (3) identify MISSING "prior-art" links; (4) create a more transparent, objective measure of network value from which market value can be deduced. Broader Impacts: The 'promotion of the progress of science' (U.S. Constitution Article 8, Section 1) and the 'encouragement of learning' (1790 Copyright Act) are vital pillars of this country's strategy for economic development and are embedded within the mission of the National Science Foundation. Since inception, over 8,625,000 utility patents / 675,000 design patents / 24,000 plant patents have been issued by the United States Patent and Trademark Office. In 2013, more than 270,000 patents were issued, of which, over 75,000 were assigned to top-50 multinational companies. Within two years, this innovation network will surpass 10 million patents. Patent innovations fuel the economic-development engine through proper commercialization. An objective tool to diagnose the health of this innovation pipeline by measuring dynamics of the network of patents, inventors, firms, lawyers, examiners, technologies, and countries would be instrumental in nurturing and fostering innovation in sophisticated R&D research labs and in an entrepreneur's garage. Organizing the wealth of information available about patent innovations, creating artificial-intelligent algorithms to extract meaning from this evidence, and presenting these patent-valuation calculations to the market has the potential to reduce asymmetry of information and bring 'transparency to intellectual property.
