NEON as a navigator of energy open innovation ecosystems

Abstract

Global energy challenges have highlighted the importance of open innovation, which facilitates the development of efficient energy solutions through collaboration and knowledge sharing. However, owing to the increasing diversity of the energy open innovation ecosystem and the lack of standardized datasets, navigating the ecosystem and identifying potential opportunities for open innovation remain challenging. As a solution, we present a large-scale network modeling and embedding-based method called the Navigator of Energy Open Innovation Networks (NEON). First, scientific literature in the energy field is collected, and its metadata and textual information are systematically preprocessed. Specific topics in the energy field are identified using energy-specific pre-trained language models and density-based clustering. Then, a large-scale network is constructed to represent the energy open innovation ecosystem, incorporating papers, organizations, nations, topics, and their interrelations. Research collaborations are represented in the triplet form. Finally, network embedding techniques are applied to identify potential collaboration partners and research topics. Our case study, involving 149 nations, 6104 organizations, 88,727 papers, and 63 topics, confirms the effectiveness of the proposed approach in exploring complex energy open innovation ecosystems.