Scholarlink

<p>This study develops a patent-based scientometric framework for examining how persistent inventive activity is embedded within corporate co-inventor collaboration backbones. Integrating Vibrant Inventor Detection with Minimum Spanning Tree (MST) analysis, the framework links inventor-level persistence to backbone-level network structure. Vibrant Inventor Detection identifies inventors who sustain above-average recent activity and continuity over time, while MST extracts a minimum-distance co-inventor backbone by transforming co-patenting frequency into reciprocal distance, thereby prioritizing frequently repeated collaboration ties. Using patent documents for BYD and Tesla from 2005 to 2024, we compare how vibrant inventors are positioned within each firm’s intra-firm co-inventor network. The results suggest that BYD’s high-frequency collaboration backbone is more consistent with a centralized and hierarchical configuration dominated by a single inventive core, whereas Tesla’s backbone is more consistent with a distributed and modular configuration composed of multiple interlinked clusters. In BYD, vibrant inventors mainly occupy coordinating positions connected to peripheral collaborators; in Tesla, they play more integrative bridging roles across clusters. Rather than proposing universal categories of corporate innovation, this study offers a comparative scientometric approach for analyzing how persistent inventive talent is organized within different high-frequency collaboration backbones.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Centralized versus distributed innovation: mapping corporate R&D collaboration backbones through vibrant inventor and MST analysis

  • 期刊论文

  • 发表日期：

    2026年7月16日

查看全文
  • Dar-Zen Chen,
  • You-Fu Lee,
  • Chun-Chieh Wang

摘要

This study develops a patent-based scientometric framework for examining how persistent inventive activity is embedded within corporate co-inventor collaboration backbones. Integrating Vibrant Inventor Detection with Minimum Spanning Tree (MST) analysis, the framework links inventor-level persistence to backbone-level network structure. Vibrant Inventor Detection identifies inventors who sustain above-average recent activity and continuity over time, while MST extracts a minimum-distance co-inventor backbone by transforming co-patenting frequency into reciprocal distance, thereby prioritizing frequently repeated collaboration ties. Using patent documents for BYD and Tesla from 2005 to 2024, we compare how vibrant inventors are positioned within each firm’s intra-firm co-inventor network. The results suggest that BYD’s high-frequency collaboration backbone is more consistent with a centralized and hierarchical configuration dominated by a single inventive core, whereas Tesla’s backbone is more consistent with a distributed and modular configuration composed of multiple interlinked clusters. In BYD, vibrant inventors mainly occupy coordinating positions connected to peripheral collaborators; in Tesla, they play more integrative bridging roles across clusters. Rather than proposing universal categories of corporate innovation, this study offers a comparative scientometric approach for analyzing how persistent inventive talent is organized within different high-frequency collaboration backbones.