Identifying Patent Precedents for Engineering Design: An Iterative Heuristic Method of Mining Text, Citation and Inventor Information

Abstract

Engineering designers often investigate patents as design precedents to learn about relevant prior arts, seek design inspiration, or assess the novelty of their own new inventions. However, patent retrieval relevant to the design of a specific product or technology is often unstructured and unguided, and resultant patents do not accurately capture the prior knowledge base. This paper proposes an iterative and heuristic methodology to comprehensively search for patents as precedents for the design of a specific technology or product. The methodology integratively mines patent text, citation and inventor information to identify relevant patents; in particular, the search keyword set, citation network, and inventor set are expanded through heuristic learning from the patents identified in prior iterations. The method relaxes the requirement for search keywords while improving patent retrieval accuracy. We apply the method to the identification of self-propelled spherical rolling robot patents. Engineers can use this repeatable method to retrieve patent sets representing the design precedents of their specialized interests.