Inequality in Global Scientific Research: Implications for Novelty and Innovation

研究人员跨越国界进行合作，科学家移居国外以促进他们的职业生涯，新发现通过范围和地点日益国际化的期刊和会议广泛传播。政策制定者和科学家的假设是，全球化有利于研究并提高知识生产，特别是对资源较少国家的研究人员来说。有广泛的研究表明，科学家的经验和观点的多样性，特别是在多学科合作中，往往会在公开合作的问题解决项目中产生更好的解决方案。然而，研究人员可能会过早地为他们的工作找到现成的、但可能不是最优的解决方案，因为科学知识--无论是通过同行评议的期刊还是全球人际网络--更容易大量和按需获得。扩大现成知识的影响是，只有少数几个国家能够获得维持尖端科学研究所需的资源。那些这样做的国家可能会越来越多地影响具有时事重要性的东西，并隐含地为科学领域设定议程，潜在地使科学话语同质化，扼杀创新。该项目旨在通过开发方法、指标和分析方法来提高对社会有益的科学发现的比率，这些方法、指标和分析方法可以识别由于同质化而随着时间的推移可能不那么新颖的领域，以及落后于同行的国家。该项目将使用整合自然语言处理和社会网络分析的创新分析工具，在全球范围内模拟科学影响力和新颖性。自然语言处理揭示了科学媒介(即学术出版物中的文本)中包含的影响和知识，而社会网络分析从结构上跟踪了国家之间和时间上的这种影响。该项目有五个主要目标：1)利用科学出版物的摘要创建跨不同领域的国际影响力网络；2)对这些网络应用基于网络的不平等衡量标准，以显示不同领域和国家随着时间的推移的趋势；3)对这些领域中的每个国家采用新颖性分数；4)将国际影响力网络中日益不平等的衡量标准与新颖性分数联系起来；以及5)确定影响力不平等以及新颖性上升或减弱最严重和最不严重的领域和国家。该项目的指导性假设是，在一系列科学领域，影响力越来越多地来自高度发达国家中较小的派别。它们在其他国家对科学的不平等影响导致了世界各地许多领域研究的同质化。随着一个领域的影响力日益集中，研究主题变得越来越跨境相似，全球创意的新颖性下降。这一奖项反映了NSF的法定使命，并通过使用基金会的学术价值和更广泛的影响审查标准进行评估，被认为值得支持。

Researchers collaborate across national borders, scientists move abroad to further their careers, and new discoveries are widely disseminated via journals and conferences that are increasingly international in scope and locale. The presumption held by policy makers and scientists is that globalization benefits research and enhances knowledge production, especially for researchers in countries with fewer resources. There is extensive research to suggest that diversity of experiences and perspectives in scientists, specifically in multidisciplinary collaborations, tends to yield better solutions in openly collaborative problem-solving ventures. However, researchers may prematurely coalesce on readily available, but perhaps suboptimal, solutions to their work, as scientific knowledge—whether embodied by peer-reviewed journals or global interpersonal networks—is more readily accessible in high volume and on demand. Amplifying the influence of readily available knowledge is that only a few countries are able to garner the resources needed to sustain cutting-edge scientific research. Those countries that do may increasingly influence what is topically important and implicitly set the agenda for scientific fields, potentially homogenizing scientific discourse and stifling innovation. This project aims to improve the rate of socially beneficial scientific discovery by developing methods, indicators, and analytic approaches that identify fields that may be less novel over time due to homogenization, as well as countries that lag behind their peers. This project will model scientific influence and novelty on a global level using innovative analysis tools that integrate natural language processing and social network analysis. Natural language processing uncovers influence and knowledge contained in the medium of science at scale (i.e., the text found in academic publications), while social network analysis structurally traces this influence between countries and over time. The project has five key objectives: to 1) create international networks of influence across diverse fields using the abstracts of scientific publications, 2) apply network-based measures of inequality to these networks to demonstrate trends over time in different fields and countries, 3) apply novelty scores to countries in each of these fields, 4) relate measures of growing inequality in international influence networks to novelty scores, and 5) identify the fields and countries where this inequality in influence and the rising or waning novelty is the most and least severe. The project’s guiding hypothesis is that across a range of scientific fields, influence increasingly originates from a smaller sect of highly developed countries. Their unequal influence on science performed in other countries results in a homogenization of what is researched around the world in many fields. As influence in a field is increasingly concentrated and research topics become increasingly similar across borders, the global novelty of ideas declines.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.