Knowledge combination modeling: The measurement of knowledge similarity between different technological domains

Abstract

This paper proposes the DB-Combination model that considers three different knowledge combinations in depth (D) and breadth (B) based on similarities of two technological knowledge domains. We also investigate three methodologies A1, A2 and A3 to highlight the three knowledge combinations. To identify technological knowledge domains, citation analysis on patent information was used for A1 and A2 and pre-existing patent classification analysis was used for A3. And to measure the similarity between identified technological knowledge domains, text similarity measurements, existing intra-industrial citation tracing and IPC share similarity comparison were used for A1, A2 and A3 respectively. The usability of the model and methodologies were demonstrated through a case study on technological knowledge of the automobile industry and the aircraft industry. While these methodologies still need to be improved, it was demonstrated that the three measurements can highlight candidates of the three knowledge combination proposed in DB-Combination model. This research contributes to accelerate breadth knowledge recombination in a complex technology industry.