Extraction of Principle Knowledge from Process Patents for Manufacturing Process Innovation

Abstract

Process patents contain substantial knowledge of the principles behind manufacturing process problems-solving; however, this knowledge is implicit in lengthy texts and cannot be directly reused in innovation design. To effectively support systematic manufacturing process innovation, this paper presents an approach to extracting principle innovation knowledge from process patents. The proposed approach consists of (1) classifying process patents by taking process method, manufacturing object and manufacturing feature as the references; (2) extracting generalized process contradiction parameters and the principles behind solving such process contradictions based on patent mining and technology abstraction of TRIZ (the theory of inventive problem solving); and (3) constructing a domain process contradiction matrix and mapping the relationship between the matrix and the corresponding process patents. Finally, a case study is presented to illustrate the applicability of the proposed approach.