Comparing European and North American is Research Using Concept Frequency Analysis: Methods and Preliminary Findings

Abstract

This study introduces a computational text-mining technique called Concept Frequency Analysis, and demonstrates its usefulness by detecting convergence and divergence trends between North American and European information systems (IS) research. Applying this technique to the corpora of articles published in leading North American and European IS journals (MISQ and ISR vs. EJIS and JIT, respectively) from 1991 through 2013 (2,959 articles), using 10.6 million unique concept labels identified from Wikipedia, and performing approximately 31.4 billion document search operations, we find that while the conceptual density of research affiliated with European and North American research seems to be gradually converging, differences in the concepts and topics being discussed on these two continents have been growing since the mid-1990s. Hence, the world of IS research is not yet flat, and in fact appears to be shifting away from cross-continental convergence.