Patent data driven innovation logic

Abstract

Innovation research is conventionally conducted with creativity techniques such as TRIZ, Mind Mapping, Brainstorming, etc. (Dewulf, Baillie 1998). Patent research is typically used to research novelty or prior art, and legal studies. This thesis is at the intersection of creativity techniques, and patent data analysis. It describes how to utilise patent data for distilling Innovation Logic and conducting innovation research. Using the patent research tool PatentInspiration (© AULIVE Software NV), the 4 different stages of the Innovation Logic approach have been subjected to text analysis in patent literature. The specific text patterns were identified and documented on several case studies, with one case study across the whole thesis: the toothbrush. The opportunities and limitations of Patent Data Driven Innovation Research have been documented and discussed. This methodology has been demonstrated within a proposed structural approach to problem solving, technology marketing and innovation research. Furthermore, the potential of artificial idea generation and artificial creativity was examined and debated for the purpose of computer aided creativity.

This thesis examines and confirms three claims: CLAIM 1: PROPERTIES AND FUNCTIONS CAN BE ADJECTIVES AND VERBS IN PATENT LITERATURE CLAIM 2: PATENT DATA ANALYSIS AUGMENTS THE FULL INNOVATION LOGIC PROCESS CLAIM 3: ARTIFICIAL INNOVATION METHODS CAN BE FUELED BY PATENT DATA

Patent data can be text mined, acting as a global brain consisting of over 100 million invention documents. It is possible to use this existing data to reverse engineer thinking methodologies, allowing scientists and engineers to solve new problems, invent new products or processes, or find new markets for existing technologies. Patent Data Driven Innovation Logic will demonstrate a systematic innovation approach that combines the force of contemporary data mining methods on patent literature, with a structured innovation research methodology.