Understanding public-access cyberlearning projects using text mining and topic analysis

The federal government has encouraged open access to publicly funded federal science research results, but it is unclear what knowledge can be gleaned from them and how the knowledge can be used to improve scientific research and shape federal research policies. In this article, we present the results of a preliminary study of cyberlearning projects funded by the National Science Foundation (NSF) that address these issues. Our work demonstrates that text-mining tools can be used to partially automate the process of finding NSFʹs cyberlearning awards and characterizing the fine-grained topics implicit in award abstracts. The methodology we have established to assess NSFʹs cyberlearning investments should generalize to other areas of research and other repositories of public-access documents.