The Evolution of Niche Fitness in China's Regional Digital Innovation Ecosystems: An Evaluation and Prediction Framework

Abstract

With the rapid expansion of digital innovation ecosystems (DIEs), the quantitative assessment of system adaptability and collaborative evolutionary capacity, as well as the forward-looking prediction of future development trajectories, has become increasingly important. Nevertheless, existing studies still lack an integrated analytical framework capable of simultaneously capturing multidimensional ecological elements and supporting reliable trend forecasting. To address this limitation, this study develops an ecology-inspired evaluation and prediction framework specifically designed for DIEs. The framework conceptualizes niche structure across four dimensions, including digital innovation communities, resources, environment, and demand, and applies the method based on the removal effects of criteria (MEREC) for objective weighting. This approach is combined with an improved measurement of alternatives and ranking according to compromise solution (MARCOS) method using the Heronian mean (HM) operator to derive composite measures of niche fitness. In addition, enhanced grey prediction models are introduced to strengthen the framework’s capacity to forecast evolutionary trends under weak-information conditions. The framework is empirically applied to provincial-level DIEs data from China covering the period from 2015 to 2021. The results indicate that overall niche fitness exhibits a fluctuating upward trend, accompanied by pronounced regional differentiation. Persistent imbalances are observed across niche dimensions, with resources, communities, and demand niches emerging as the primary constraints on the evolutionary process of DIEs. Forecasting results further suggest that niche fitness is likely to continue improving in the coming years, although structural misalignments may remain difficult to fully resolve in the short term. Overall, the proposed evaluation and prediction framework enables the systematic identification of structural constraints and latent development potential within DIEs, while also facilitating the analysis of their future evolutionary trajectories. As such, it provides a generalizable and scalable methodological tool for comparative analysis, dynamic monitoring, and cross-regional research on DIEs.