Understanding IT Innovations Through Computational Analysis of Discourse

How do Information Technology (IT) innovation concepts emerge, coexist, evolve, and relate to each other? To address this question, we theorize that innovation concepts are interrelated in an idea network, where they can be likened to species in a competitive and symbiotic resource space. Communities of organizations and people interested in the innovations produce discourse that both reflects and enables the flows of attention among innovations. From this ecological perspective, we apply discourse analysis to innovation research and propose computational approach to scale up the analysis. Specifically, we employed Kullback-Leibler divergence to compare the linguistic patterns of 48 IT innovations reported in InformationWeek and Computerworld over a decade. Using multidimensional scaling, we found that similar innovations demonstrated similar discourses. The results demonstrate the validity, scalability, and utility of computational discourse analysis for practitioners and scholars to understand the socio-technical dynamics in the IT innovation ecosystem