A Data-driven Approach for Constructing Multilayer Network-based Service Ecosystem Models

Services are flourishing drastically both on the Internet and in the real world. Additionally, services have become much more interconnected to facilitate transboundary business collaboration to create and deliver distinct new values to customers. Various service ecosystems have become a focus in both research and practice. However, due to the lack of widely recognized service ecosystem models and sufficient data for constructing such models, existing studies on service ecosystems are limited to very narrow scope and cannot effectively guide the design, optimization, and evolution of service ecosystems. We propose a Multilayer network-based Service Ecosystem Model, which covers a variety of service-related elements, including stakeholders, channels, functional and nonfunctional features, and domains, and especially, structural and evolutionary relations between them. "Events" are introduced to describe the triggers of service ecosystem evolution. We propose a data-driven approach for constructing MSEM from public media news and external data sources. Qualitative comparison with state-of-the-art models shows that MSEM has a higher coverage degree of fine-grained elements/relations in service ecosystems and richer semantics for higher interpretability. Experiments conducted on real news corpora show that compared with other approaches, our approach can construct large-scale models for real-world service ecosystems with lower cost and higher efficiency.