PurposeThe risks associated with digital innovation increasingly challenge value co-creation among stakeholders within the innovation ecosystem. Stakeholder collaboration is helpful in preventing risk occurrence. This study intends to explore the effects of different stakeholder collaboration strategies on risk prevention performance in a digital innovation ecosystem context.Design/methodology/approachA systematic literature analysis was first conducted to identify risk factors of digital innovation based on the technology–organization–environment (TOE) framework. Then, a bidimensional network model was constructed to visualize the collaborative relationships among stakeholders and the identified risks by focusing on a digital innovation case. The social network analysis method was applied to design stakeholder collaboration strategies from the ego and global network perspectives, and a simulation approach was conducted to evaluate the effects of the strategies on risk prevention performance.FindingsThe results validate the positive effect of stakeholder collaboration on risk prevention performance and reveal the important role of network reachability in formulating collaboration strategies. The strategy of strong–strong collaboration strategy can best enhance risk prevention performance like a “Matthew effect” in the digital innovation ecosystem.Originality/valueFirst, risk identification based on the TOE framework provides a systematic list of risk factors for future digital innovation risk management research. Second, this study designs stakeholder collaboration strategies from a network perspective to enhance the understanding of the network status of each stakeholder and the network structure of the digital innovation ecosystem. Third, the simulation results reveal the effects of different collaboration strategies on risk prevention performance.

The infectious disease COVID-19 has swept across the world in 2020, and it continues to cause massive losses of life and severe economic problems in all countries. Providing emergency supplies such as protective medical equipment and materials required to secure people’s livelihood is thus currently prioritized by governments. Establishing a reliable emergency logistics system is critical in this regard. This paper used the Delphi method to design a formal decision structure to assess emergency logistics system reliability (ELSR) by obtaining a consensus from a panel of experts. Assessing ELSR is a typical multiple-attribute decision making (MADM) problem, and the related MADM methods are usually on the basis of symmetry principles. A hybrid MADM model, called the Decision Making Trial and Evaluation Laboratory (DEMATEL)-based Analytical Network Process (D-ANP), was developed to identify the critical factors influencing ELSR. An analysis of empirical evidence showed that the emergency logistics command and coordination system and the emergency material supply system play important roles in ELSR, while the emergency logistics transportation and distribution system and the emergency information system are not so important. This conclusion is different from previous research about traditional disaster emergency logistics. Moreover, the cause–effect relationships among the key factors indicated that the system of command and coordination for emergency logistics and the supply system for emergency materials should be improved. Accordingly, effective suggestions for emergency logistics services for epidemic prevention are provided in this paper. The main contributions of this paper are (1) establishing a comprehensive and systematic evaluating index of ELSR for epidemic prevention; (2) employing a kind of structured, namely D-ANP, to identify the critical factors with non-commensurable and conflicting (competing) characteristics; and (3) comparing the differences of reliable criteria between the emergency logistics of epidemic prevention and the traditional disaster emergency logistics.

The forecast of carbon dioxide (CO2) emissions has played a significant role in drawing up energy development policies for individual countries. Since data about CO2 emissions are often limited and do not conform to the usual statistical assumptions, this study attempts to develop a novel multivariate grey prediction model (MGPM) for CO2 emissions. Compared with other MGPMs, the proposed model has several distinctive features. First, both feature selection and residual modification are considered to improve prediction accuracy. For the former, grey relational analysis is used to filter out the irrelevant features that have weaker relevance with CO2 emissions. For the latter, predicted values obtained from the proposed MGPM are further adjusted by establishing a neural-network-based residual model. Prediction accuracies of the proposed MGPM were verified using real CO2 emission cases. Experimental results demonstrated that the proposed MGPM performed well compared with other MGPMs considered.