Toward a network graph-based innovation cluster density index

Abstract

Innovation clusters have been studied and examined in many forms: ranging from qualitative observations to in-depth analytical models and frameworks to long-term studies tracking the actors and entities making up an innovation cluster's ecosystem. Economic development marketing in municipalities, regions, and countries often make representations about their cluster's strengths, yet rarely have strong empirical data to support and back their claims. A wide array of cluster mapping visualization tools exist, yet most focus on aggregate numbers of "nodes," and offer far less insights into their connectedness and relationships between the entities, the "edges" or links. The focus for this thesis is a bottoms-up perspective, with people as the core drivers of innovation. This paper seeks to supplement existing methods, by implementing an innovation cluster density index (CDI) as an indicator, ("Cluster Rank") enabling broader comparisons within clusters (intra-cluster), as well as the modeling of distributed virtual clusters (inter-cluster). This method proposes an empirical analytical approach, using complex network theory, commercially and open source available application program interfaces (APIs), and weighted network graphs as a framework, which integrates these elements to depict a new descriptor for clusters, the Cluster Rank. Implementation of the method in software is outside the scope of this thesis, but is separately being developed and is defined as a software platform using linked data technologies to build it ("Cluster Rank Engine"). The proposed Cluster Rank Engine is people-centric, and takes into account the embedded network effects, of people, derived from network graph analytics. It presents a bottoms-up view to intersect with the relatively top-down approaches currently in place. It identifies five key attributes, the "Penta Helix" and uses these as the core variables in modeling. Development of such a model enables the use of big data methods and algorithmic tools on the Internet to interrogate large distributed economic global datasets, query and extract the relevant pre-defined cluster attribute data, filter and process it to present a deeper analytically comparative lens of innovation clusters; both in terms of urban innovation mapping, cluster heat maps, etc. This method would enable, for example, the comparison of a biotech cluster in Cambridge with that of one in San Francisco ("Global biotechnology clusters map," n.d.), at discrete levels. The Cluster Rank for each discrete innovation cluster provides additional data beyond traditional graphical visualizations. Utilization of the Cluster Rank Engine across a wide range of clusters could then yield deeper statistically comparative data for a deeper understanding of cluster dynamics and cluster endurance over time, as well as serve as data input for a variety of graphical data visualizations. (Berkhin, 2002) Utility for such a solution is multi-fold: as an economic cluster modeling and tracking tool, an innovation lens on a given sector or geography, and as a tool for urban innovation mapping. At its fruition, it becomes a potentially predictive tool for network resilience and failure, to help better navigate decisions related to the growth of innovation clusters and/or the linking of remote clusters for a virtual cluster, to help make decisions for: resource allocations, partnership and contractual targets, angel and venture funding strategies.