Predictive Modeling of the Emergence and Development of Scientific Fields

点击翻译按钮获取中文摘要

The explosion of scientific publishing in recent decades has created an embarrassment of riches. More information is available about more subjects in ways that are more accessible than ever before. Searches on PubMed or Web of Science routinely return thousands of hits. The easy accessibility of gigantic numbers of publications creates substantial problems for researchers and granting agencies alike: scientists cannot possibly read everything published in their own fields, and policymakers have few clearbases upon which to form judgments of quality. How can reasonable decisions be made about which fields are likely to blossom into transformative research, if the problem of separating wheat from chaff grows exponentially with each passing year? Given the ever-increasing outpouring, it is more important than ever to develop some quantitative means of describing -- and, in turn, predicting -- the growth and development of scientific research. What factors might explain the changes over time of numbers of publications on a given topic, or numbers of researchers working in a given field? What interactions between researchers might account for a scientific community's growth and change? And how might various interventions -- policy initiatives, market forces, or other factors -- alter the dynamics of how scientific fields emerge and develop?Intellectual Merit: The research develops new measures to model and predict processes of scientific research and innovation. The measures are based on mathematical models that capture various aspects of the research enterprise. These tools are sharpened with application to a suite of examples from the physical and biomedical sciences. These examples permit the investigation of different modes of supporting and steering the course of scientific research in six key areas: the growth and development of research publications on string theory, molecular electronics, and nanotechnology from the physical sciences; and research publications on H1N1 influenza, coronary angioplasty, and interventional cardiology from the biomedical sciences. These cases traverse a wide range of scales, from research topics within well-defined subfields to large, discipline-spanning endeavors.Broader Impact: The dissemination of the results includes scientific publications, workshops with expert researchers, an informal bi-weekly seminar series for scholars across a wide range of disciplines - from history and sociology of science, to business and economics, to library science and computer science. In addition, new courses are offered at the undergraduate and graduate levels. Finally, the structure, materials, and resources for these courses are freely disseminated via the internet via MIT's OpenCourseWare initiative.