Scholarlink

<p>Regional diversification and innovation depend on how well existing knowledge bases align with emerging scientific domains. Conventional measures of relatedness, however, rely on static classification systems that struggle to capture the interdisciplinary and dynamic nature of modern knowledge production. This study introduces <i>semantic proximity</i> as a novel, text-based approach to assess the cognitive alignment between regional scientific activity and new fields. Drawing on scientific publications from European NUTS-3 regions between 2010 and 2021, we employ BERTopic and BERT-based similarity to map regional knowledge spaces and measure their proximity to quantum science, an interdisciplinary field of high strategic importance. Using negative binomial regression models, we find that regions with higher semantic proximity to quantum science generate significantly more quantum-related publications. By moving beyond rigid taxonomies, the concept of semantic proximity provides a more flexible and fine-grained tool for analyzing regional knowledge evolution. Our findings extend evolutionary economic geography and offer actionable insights for policymakers seeking to foster smart specialization and guide regional innovation strategies in emerging technologies.</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

Mapping regional knowledge spaces via BERTopic and BERT similarity: the semantic proximity of quantum science in the overall scientific publication landscape

  • 期刊论文

  • 发表日期：

    2025年9月29日

查看全文
  • Keungoui Kim,
  • Dieter F. Kogler,
  • Jisoo Hur

摘要

Regional diversification and innovation depend on how well existing knowledge bases align with emerging scientific domains. Conventional measures of relatedness, however, rely on static classification systems that struggle to capture the interdisciplinary and dynamic nature of modern knowledge production. This study introduces semantic proximity as a novel, text-based approach to assess the cognitive alignment between regional scientific activity and new fields. Drawing on scientific publications from European NUTS-3 regions between 2010 and 2021, we employ BERTopic and BERT-based similarity to map regional knowledge spaces and measure their proximity to quantum science, an interdisciplinary field of high strategic importance. Using negative binomial regression models, we find that regions with higher semantic proximity to quantum science generate significantly more quantum-related publications. By moving beyond rigid taxonomies, the concept of semantic proximity provides a more flexible and fine-grained tool for analyzing regional knowledge evolution. Our findings extend evolutionary economic geography and offer actionable insights for policymakers seeking to foster smart specialization and guide regional innovation strategies in emerging technologies.