GC-GTM: A generative-convergent extension of generative topographic mapping for discovering and interpreting technology opportunities

Technological innovation is increasingly driven by the convergence of diverse domains and industries. However, existing GTM-based vacancy maps predominantly focus on single technological fields, often overlooking cross-domain convergence and the associated market value. Furthermore, these approaches lack interpretive mechanisms to translate vacancies into potential innovation directions. To address these limitations, we propose a novel framework called GC-GTM (Generative-Convergent GTM) which is a modified GTM framework that integrates technology–market convergence in the mapping process and leverages generative AI to interpret latent vacancies. This dual enhancement addresses the limitations of conventional GTM in both discovery and interpretation. The framework conceptualizes the current technology–market configuration as a current state, and the unexplored convergence areas as a latent opportunity space. We explore keyword-level technology and market convergence. Then, GTM is employed to generate a two-dimensional technology–market vacancy map in the latent space, where each vacancy is inverse-mapped to its latent keyword distribution and translated into textual descriptions via generative modeling. A case study on-device AI demonstrates the framework’s effectiveness.