A New Paradigm of Software Service Engineering in the Era of Big Data and Big Service
Servitization is one of the most significant trends that reshapes the information world and society in recent years. The requirement of collecting,storing, processing, and sharing of the Big Data has led to massive software resources being developed and made accessible as web-based services to facilitate such process. These services that handle the Big Data come from various domains and heterogeneous networks, and converge into a huge complicated service network (or ecosystem), called the Big Service.The key issue facing the big data and big service ecosystem is how to optimally configure and operate the related service resources to serve the specific requirements of possible applications, i.e., how to reuse the existing service resources effectively and efficiently to develop the new applications or software services, to meet the massive individualized requirements of end-users.Based on analyzing the big service ecosystem, we present in this paper a new paradigm for software service engineering, RE2SEP (Requirement-Engineering Two-Phase of Service Engineering Paradigm), which includes three components: service-oriented requirement engineering, domain-oriented service engineering, and software service development approach. RE2SEP enables the rapid design and implementation of service solutions to match the requirement propositions of massive individualized customers in the Big Service ecosystem. A case study on people's mobility service in a smart city environment is given to demonstrate the application of RE2SEP.RE2SEP can potentially revolutionize the traditional life-cycle oriented software engineering, leading to a new approach to software service engineering.