Patent citation network analysis: A perspective from descriptive statistics and ERGMs

Area of Science:

• Bibliometrics
• Network Science
• Patent Analysis

Background:

• Patent citation analysis is increasingly important for understanding innovation.
• Previous studies often focused on descriptive statistics rather than network mechanisms.

Purpose of the Study:

• To analyze patent citation networks from a statistical perspective.
• To explore the social and technical factors influencing patent citations using advanced statistical models.
• To provide insights into European patent citation patterns.

Main Methods:

• Collection and analysis of extensive patent and citation data.
• Descriptive analysis of most cited patents, innovative companies, and network structures.
• Application of Exponential Random Graph Models (ERGMs) to model citation networks.

Main Results:

• Identification of key structural properties within patent citation networks.
• Demonstration that social properties (homophily, transitivity) and technical aspects (language, categories) significantly affect citations.
• Analysis of sector-specific citation patterns and their relation to sector size.

Conclusions:

• ERGMs provide a powerful tool for modeling network mechanisms in large-scale patent data.
• Social and technical factors play crucial roles in shaping patent citation networks.
• This study offers novel insights into the dynamics of European patent citations.