Patent Analysis for Systematic Innovation: Automatic Function In-terpretation and Automatic Classification of Level of Invention using Natural Language Processing and Artificial Neural Networks

Abstract

With advances in computing power and the processes of globalization, the analytical and engineering science skills that contribute to innovation are becoming a commodity, and the activities of research and development—and innovation—are being outsourced. These trends leave the creative and systems integrative skills of engineering design as the value-added part of innovation. This paper presents a framework to address this challenge, termed mass innovation, which can be defined as expanding and diffusing innovation activities to the general population through connecting inventors and entrepreneurs with the engineering tools and services needed to assess and realize their novel design concepts. As part of mass innovation, this paper presents the development of an approach for automatic function interpretation, and an example is given, in the context of sustainable design, of the application of automatic function interpretation and automatic classification of level of invention. The method for automatic function interpretation is based on text extraction, natural language processing using a parser, and semantic definition of functional requirements and design parameters. The classification of level of invention is based on an artificial neural network model using inputs based on patent citation measures.