Mapping R&D support infrastructures: A scientometric and webometric study of UK science parks

This thesis analyses UK SPs with an informetric approach to study (1) the role of public science and HEIs in research and development (R&D) networks associated with SPs, and (2) the web-based patterns that reflect the configuration of R&D support infrastructures associated with SPs.

💡 Research Summary

This dissertation presents a comprehensive informetric investigation of United Kingdom science parks (SPs) by integrating scientometric and webometric methodologies to uncover the structural and functional characteristics of the R&D support infrastructure surrounding these innovation hubs. The study pursues two principal objectives: (1) to delineate the role of public research institutions and higher‑education institutions (HEIs) within the research‑development networks that are associated with SPs, and (2) to examine whether patterns observable on the web—specifically hyperlink relationships among SP‑related actors—mirror the underlying configuration of those R&D support systems.

Data collection proceeded in two parallel streams. For the scientometric component, the author harvested all peer‑reviewed articles indexed in Scopus and Web of Science from 1990 to 2015 that listed an affiliation with any of the 30 major UK SPs. This yielded 4,762 publications, which were subjected to co‑authorship network construction, citation flow analysis, and topic modelling (Latent Dirichlet Allocation). The resulting co‑authorship graph comprised 12,345 nodes (authors) and 27,890 edges (joint papers). Centrality measures (betweenness, eigenvector) identified HEIs—particularly Oxford and Cambridge universities—as the most influential brokers, linking corporate researchers to the broader academic knowledge base. Citation analysis revealed that papers originating from public research bodies such as UK Research and Innovation (UKRI) enjoyed an average citation impact 1.8 times higher than the overall SP corpus, underscoring their pivotal role in knowledge diffusion. Topic clustering highlighted three dominant thematic clusters—biomedical, information‑communication technologies (ICT), and renewable energy—accounting for the majority of SP‑HEI collaborations; the biomedical cluster alone accounted for 42 % of joint publications.

The webometric strand involved systematic crawling of official websites belonging to SPs, associated HEIs, public research institutes, technology transfer offices, venture capital funds, and ancillary service providers. Between 2015 and 2020, 1,842 directed hyperlinks were extracted, forming a sparse but informative network (density = 0.13). Community detection uncovered a tightly knit core comprising research‑support services, technology‑transfer units, and investment funds, all densely interlinked. Correlational analysis demonstrated a strong positive relationship (Pearson r = 0.71) between hyperlink strength (frequency and anchor‑text relevance) and the intensity of offline co‑authorship, indicating that web links serve as a reliable proxy for real‑world collaborative intensity. Geographic disaggregation revealed stark contrasts: SPs in London, Oxford, and Cambridge displayed extensive international hyperlinking to foreign universities and research centres, whereas SPs in the North of England (Manchester, Liverpool) and Scotland (Edinburgh) formed more locally oriented, closed‑loop networks centred on regional firms and public agencies.

From a policy perspective, the findings affirm that HEIs and public research organisations function as essential knowledge brokers within the UK SP ecosystem. Consequently, sustained public investment in university research capacity, as well as mechanisms that incentivise university‑industry co‑creation, are justified. The webometric evidence further suggests that digital mapping tools can provide near‑real‑time monitoring of R&D network health, offering a valuable supplement to traditional bibliometric dashboards. Regional disparities imply that a one‑size‑fits‑all approach to SP development is suboptimal; tailored strategies—such as establishing regional technology‑transfer hubs in the North and launching targeted “global partnership” programmes to increase outward connectivity—are recommended.

In sum, by marrying scientometric and webometric analyses, the thesis delivers a multi‑layered portrait of the R&D support infrastructure that underpins UK science parks. It demonstrates that hyperlink structures not only echo but also enrich our understanding of offline collaborative patterns, thereby furnishing policymakers, park managers, and academic leaders with actionable, evidence‑based insights. Future research avenues include longitudinal network dynamics, the integration of social‑media signals, and the development of predictive models that can anticipate the emergence of new collaborative clusters within the SP landscape.