Modeling knowledge diffusion in the disciplinary citation network based on differential dynamics

Knowledge diffusion based on disciplinary citation resembles disease propagation through actual contact. Inspired by the epidemic spread model, the study classifies disciplines from the viewpoint of knowledge diffusion into five states: knowledge recipient disciplines (S), potential knowledge diffusion disciplines (E), knowledge diffusion disciplines (I), knowledge skeptic disciplines (Z), and knowledge immune disciplines (R). The classifications of disciplines can change from one state to another at a rate of α, β, ω, γ, θ or μ. As a result, evolution rules for knowledge diffusion in the disciplinary citation network are created, and the knowledge diffusion SEIZRS model of differential dynamics in the disciplinary citation of a non-uniform network is formed, followed by a comparative analysis between the SEIZRS model and the classic SIR model. Next, the evolution of knowledge diffusion and the influence of state transition parameters on it are discussed to reveal the dynamic mechanism of knowledge diffusion in the disciplinary citation network. Research has shown that the latent mechanism, skeptical mechanism, and feedback mechanism of knowledge introduced in this study can effectively reveal the dynamic mechanism of knowledge diffusion in the disciplinary citation network. The knowledge diffusion state evolution of disciplines in the disciplinary citation network is affected by both the knowledge diffusion evolution states and the relative citation weight (knowledge contact intensity) of neighboring disciplines. Moreover, changes in state transition parameters have different effects on the evolution of knowledge diffusion.