Cross-layer diffusion in patent networks: Forecasting innovation through enterprise–technology synergy

Firms face serious challenges in predicting the direction of innovation and identifying strategic partners in rapidly evolving technological fields, as traditional single-layer network models tend to ignore the synergies between inter-firm relationships and cross-field technology diffusion. To address this limitation, we combine the firm perspective with bi-layer patent networks to analyze synergistic technology diffusion links and predict the direction of innovation frontiers. We use a bi-layer diffusion-based patent network model that dynamically combines firm-level technology similarity networks with technology topic co-occurrence networks. By employing diffusion-based network analysis, the model quantifies the propagation of technological elements across organizational and disciplinary boundaries and identifies emerging technological directions and collaboration opportunities. The model is empirically validated on a dataset in the field of "Blockchain + Audit" to effectively predict future research directions and recommend potential technology development fields and partners for firms. This study provides valuable insights into the diffusion process of technology elements in bi-layer networks, and provides strategic insights to guide firms’ innovation decisions in dynamic technological environments.