Innovation rarely unfolds as a one-sided process; it arises from the interplay of dissimilar forces whose operational mechanisms often diverge, and manifests in two primary dimensions: technological market supply, integrated from existing knowledge infrastructure, and institutional market demand, articulated through national policy agendas. This duality often creates strategic uncertainty and generates systemic frictions that hinder coordination across scales, making it analytically urgent to diagnose the magnitude and structure of these tensions. However, such efforts remain methodologically underdeveloped. In response, we propose a data-driven analysis framework that integrates complex network modeling and deep-learning techniques to quantitatively measure topic consistency across these structurally decoupled markets. Building on a fully documented end-to end data pipeline, cross-validated deep-learning components, and sensitivity analyses for key modeling choices, the framework achieves methodological rigor and robustness. We validate its effectiveness through an empirical study of the fuel cell vehicle sector. Misalignments such as excess supply of niche technologies and unmet demand for hydrogen infrastructure were effectively identified, highlighting frictions between innovation frontiers and domestic priorities. By quantifying the magnitude of dual-market divergence, this work offers innovative organizations actionable insights to better align investments with both global trends and local policy needs. It also offers a replicable methodology for diagnosing structural frictions in broader innovation ecosystems, with potential applications in global innovation research and sector-specific policy design.