A Study of a Patent Network with Respect to Structural Density and Diversity- Sector, Firm and Country Level Exploration-

Abstract

Innovation academics have been interested in the systematic perspective on innovation, which emphasizes the interaction between agents such as firms, universities, and governments. In recent years, to improve the understanding of the structural properties of knowledge through this interaction, and its effect on innovation performance, many studies have applied network analysis to the documents that codify knowledge, such as patents and papers. Many comparison studies using network analysis have focused on the effect of network position among comparable agents or its organizational structure. However, considering the resources that agents have developed and possessed such as technology and human resources, it is of primary importance for making strategies or policies for technology management. For this reason, many studies examine diversification, specialization, and the depth and breadth of the technology and the products that agents possess. However, the interactions and effects of the diversity and density of technology during its combination and convergence processes have not been examined thoroughly, despite their importance recently skyrocketing. In the digital economy, it is not just important for agents to construct a technology portfolio, but also to have the ability to converge existing technologies into their own technology pool. Therefore, this dissertation discusses the structural characteristics of the technology pool that an agent owns and technology network that reflects the technology convergence tendency for inventions. Moreover, the effect of the characteristics of the technology pool and network of innovation will be examined. Three kinds of patent co-classification networks are constructed at three levels (technology sector, firm, and country) using the International Patent Classification (IPC). To investigate the characteristics of technology convergence that has occurred in other industries, Korea's patent application data is used. To examine the relationship between technology structure and innovation performance at a firm and country level, patent data, innovation surveys, and economic data were merged. Examinations are conducted by focusing on the effect of the structural density and diversity of the patent network and pool. The comparison of the characteristics of technology convergence is based on the ICT that has led to Korea's growth. The result shows that ICT contributed to Korea's co-evolution, which is more likely to pivot on related technology convergence than unrelated technology convergence. The Relationship between the characteristics of the technology pool and network and innovation shows that diversity in the technology capability pool and the technology usage experience are positively related to the R&D expenditure and innovation output for both firm and country level studies, except for the effect on innovation output in the firm-level study. In contrast to technology pool indices, the technology network indices show different effects for firms and countries. The network density of a firm has a positive relationship with the R&D intensity and innovation output. However, a countrys network density does not significantly affect either variable. The network diversity shows a contrary result for innovation output. Firms network diversity has a negative relationship with innovation output, while a countrys network diversity has a positive relationship. Such findings and approaches based on the technology network shed light on policies and strategies for promoting innovation and the performance for different organizations.