Advances in searching for prior art
Blog by Andrei Iancu, Under Secretary of Commerce for Intellectual Property and Director of the USPTO and Drew Hirshfeld, Commissioner for Patents of the USPTO
Patent examination, though inherently complex, is in large part a fact-finding mission with the goal of providing predictable and reliable intellectual property rights. The prior art search is the foundation for achieving this mission.
Our patent examiners’ ability to find the best prior art at the earliest possible time ensures both quality and timeliness, two primary goals of the USPTO’s Strategic Plan. Indeed, the best and most appropriate time to ensure patent quality is during the examination process, before issuance. This is why we have worked tirelessly to improve the search capabilities of our examiners so they can more readily identify patentable subject matter and the appropriate scope of patent rights.
As technology evolves and advances, so too must our examination process and the underlying tools and mechanisms, particularly those used for prior art search. To enhance examiners’ ability to efficiently find the most relevant prior art in a body of references that is expanding both numerically and globally, we have improved our processes, added search tools, leveraged the abilities of our highly skilled workforce, and promoted knowledge exchange, all while upholding the fundamental pillars of our world-class patent system. Our recent endeavors include:
- Launching our new Patents End-to-End (PE2E) Search Tool for examiners on a modern, web-based platform with a focus on performance and adaptability. This tool will allow for the flexible development of additional search functionality and access to over 70 million foreign documents with full image and text by April of 2021. A version of this tool will be available to the public later in FY 2021.
- Developing and testing promising new search capabilities that utilize artificial intelligence (AI), including an AI-based prototype search system to further assist examiners in finding relevant prior art as well as an auto-classification tool that leverages machine learning to classify patent documents using the Cooperative Patent Classification system.
- Conducting the Peer Search Collaboration Pilot, a valuable mechanism for learning and collaboration in which paired examiners independently search the same case and exchanged results and strategies.
- Establishing a Search and Classification Examiner position in every utility technology center to serve as an added resource for examiners by providing searching expertise and training.
- Offering continuing education classes on search techniques and strategies, database searching, and non-patent literature searching using discipline-specific examples.
- Training examiners on Global Dossier, which provides unparalleled access to an entire patent family in one location, English machine translations, and the ability to view all citations in a single list.
- Updating examiner performance appraisals to increase the emphasis on searching, both at the planning and conducting stages, and providing the most relevant prior art as early as possible in prosecution.
We continue to explore additional avenues to improve examiners’ access to prior art and ensure that their efforts are most effective. From making prior art in related applications more accessible, to assessing incoming applications and identifying attributes that can increase the quality of searches, from continuous and advanced training, to further exploring the potential of AI capabilities, we owe it to our patent applicants and stakeholders to leave no stone unturned. Our unwavering commitment to explore new tools and technologies is inextricably linked to our commitment to provide the utmost certainty and reliability in the patents we grant.