A machine-learning approach for research model creation in theory building research

We propose and illustrate an automated process for research model creation and hypothesis generation. To do so we argue that the early phases of theory building research are algorithmic and can be automated, given that a corpus of knowledge from past research has been appropriately codified. A pilot implementation is illustrated, together with an argument for the feasibility of knowledge extraction from past research. When fully implemented, the approach could significantly disrupt the research process, increasing process speed and enhancing creativity.