Modeling Firm Search and Innovation Trajectory Using Swarm Intelligence

We developed a swarm intelligence-based model to study firm search across innovation topics. Firm search modeling has primarily been "firm-centric," emphasizing the firm's own prior performance. Fields interested in firm search behavior-strategic management, organization science, and economics-lack a suitable simulation model to incorporate a more robust set of influences, such as the influence of competitors. We developed a swarm intelligence-based simulation model to fill this gap. To demonstrate how to fit the model to real world data, we applied latent Dirichlet allocation to patent abstracts to derive a topic search space and then provide equations to calibrate the model's parameters. We are the first to develop a swarm intelligence-based application to study firm search and innovation. The model and data methodology can be extended to address a number of questions related to firm search and competitive dynamics.