Identifying technology clusters based on automated patent landscaping

Abstract

We introduce a new general methodological approach for accurately and consistently retrieving a large set of patents related to specific technologies. We build upon the automated patent landscaping algorithm by incorporating a tractable amount of human supervision to improve the accuracy and consistency of our results. We demonstrate the efficacy of our approach by applying it to six novel and representative technologies: additive manufacturing, blockchain, computer vision, genome editing, hydrogen storage, and self-driving vehicles.