We use community detection analysis to investigate the structure of Bengaluru's ICT cluster's inter-organizational network during the period 2015–2017. Building on the knowledge sourcing literature, we conjecture that cluster firms primarily build knowledge-seeking horizontal linkages with technologically similar companies, and that this splits the network into multiple technological communities within which firms are tightly connected, but between which linkages are scarce. We further propose that community-spanning firms which build horizontal linkages that bridge technological communities are more likely to conduct radical innovation than their peers. We finally argue that no relation exists between technological proximity and community formation in the network of vertical buyer-supplier relations. Using a voltage-based algorithm for community discovery, we draw empirical support for these predictions. We discuss the implications of our findings for Bengaluru's upgrading potential.

中文翻译：

---

班加罗尔 ICT 集群剖析：社区检测方法

我们使用社区检测分析来调查 2015-2017 年期间班加罗尔 ICT 集群的跨组织网络的结构。基于知识采购文献，我们推测集群公司主要与技术相似的公司建立寻求知识的横向联系，这将网络分成多个技术社区，企业在这些社区内紧密联系，但之间的联系很少。我们进一步提出，与同行相比，建立横向联系以弥合技术社区的跨社区公司更有可能进行激进创新。我们最后认为，在垂直的买卖双方关系网络中，技术邻近性和社区形成之间不存在任何关系。我们使用基于电压的社区发现算法，为这些预测提供经验支持。我们讨论了我们的发现对班加罗尔升级潜力的影响。