Network analysis of international patent data: technology cohort, temporal dynamics and the role of trade

Access to online electronic database and international patentdata harmonization has enabled more researchers to workwith patent data. This thesis joins the more recent studiesto apply network methodologies to patent data analysis inorder to understand the multi-variable interactions in innovation, knowledge flows, and technological trends.The first chapter focuses on technology cohort in the patentfamily networks and citation networks to investigate how technologies are integrated in utility invention and the patternsover time. The consistent technological groups found in thisanalysis are compared to the authority-defined system andprovide a more complete coverage.Chapter 2 focuses on community evolution over time. Thestabilized Louvain method is adopted to improve consistencyand stability of community detection. The majority mappingalgorithm is incorporated for community tracking across timeslices. A new method is developed to identify sets of centralnodes. A case study of patent filed by applicants in Germanyis used to demonstrate and verify the method and the results.In the last chapter, I apply the methods developed from thefirst two chapters to the pharmaceutical sector. Motivatedby the needs to encourage pharmaceutical R&D and promoteglobalized innovation, I focus on the effects of being centralwith a favorable balance in international market on being central in global pharmaceutical R&D collaboration. The descriptive results and a further regression analysis shows positiveeffects from export “Coreness” on co-invention centrality. Despite causality is not fully resolved, our results show a fundamental relationship between knowledge production andtrade in pharmaceuticals.