A Text Mining Approach to Measuring Educational Scholarly Impact

• In Event: Hilton Roundtable Session 6
  In Roundtable Session: Machine Learning and Data Sciences

Abstract

Identifying influential articles in a corpus is an important and challenging problem in educational research. However, these methods do not pay enough attention to the context of articles. Continuous improvements in text mining techniques present new opportunities for measuring scholarly impact beyond citation-analysis. But few studies have been done applying these methods in educational research with real world data. In this paper, we introduce an innovative approach to measuring educational scholarly impact based on text mining. About 2000 articles from the Teachers College Record from 1995 to 2016 have been analyzed. We also discuss the potential and the challenges of applying text mining in bibliometrics.

Authors

• Yi Chen, Columbia University

• Xiang Liu, Educational Testing Service

• Hui Soo Chae, Teachers College, Columbia University

• Gary J. Natriello, Teachers College, Columbia University