How the Dynamics between Inventors and Patent Examiners Determines Innovation

The aim of this project is to investigate how exchanges between inventors and patent examiners affect the development of technology and its value once patented, and more generally the translation of research to use. The patent examination process may involve rejections of applications’ claims based on obviousness or novelty, while inventors may respond with arguments for inventions’ distinctiveness; they also may adjust technologies in response. This research investigates this process through innovative AI-assisted analysis of all publicly available patent applications at the United States Patent and Trademark Office. It measures how patent applications evolve via the examination process; it measures examiners’ and applicants’ interpretations and counterinterpretations of patent claims; and it models how these interactions affect what gets patented, the long-run value of patented technology, and the strategies examiners and applicants learn through multiple examinations. Results from this study inform improvement in translation of research to patents and from patents to technologies that improve our lives. The research is conducted in three interrelated studies of the patent examination process. Project 1, “Invention Evolution,” measures how patent claims are transformed during prosecution, utilizing a blend of AI and machine learning techniques such as natural language processing, and qualitative coding. Project 2, “Pluralistic Interpretation,” applies a combination of natural language processing and qualitative coding to measure how examiners interpret prior art and application claims, and how applicants offer counter-interpretations in their rebuttals. Finally, Project 3, “Learning to Patent,” integrates findings from the previous two projects to address questions about examiner and inventor learning. It fits panel multinomial logic models to explain how examiner and applicant interactions change emergent technological claims; how these interactions, in turn, relate to the long-run success of patents; and finally, it models how parties develop strategies that lead to more efficient and effective patent execution. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.