Research On QoS History Records Based Service Composition Recommendation In Big Data Environment

With the rapid spread of mobile devices such as smart phones, tablet PCs and development of the wireless networks and related technologies such as Web2.0, the number of mobile Internet users is growing. And it also spawned a diverse and large-scale development of services. A large number of service description, transaction records and monitoring information are produced in every corner of the world all the time and "Big Data for Service" gradually forms. How to deal with service composition problem based on big data applications has become a problem.Meanwhile, the environment of Service changes all the time and the QoS(Quality of Service) may alters as the internal or external changes of the services. It would be a serious obstacle to the user’s selection especially when some of service providers advertised unreliable or inaccurate QoS description. Using the history records of actual service transaction as a standard metrics for QoS, and processing these history records to evaluate the quality of services and recommending the best combination plan is often a good choice.In view of the above problems, this paper proposes two service composition optimization methods which can overcome the slowly processing of massive data and the unstable environment of service based on large-scale history records of services’ QoS from different perspectives. Experiments show that our approaches have higher operating efficiency while maintaining a high degree of accuracy compared with the traditional global method. What’s more, as the amount of service data increases, the efficiency improving of our approaches is more apparent. Meanwhile, this paper also discusses the services composition recommendation in the big data environment applications, and designs a services composition recommendation system architecture in big data environment. It can effectively collect and organize massive service data, and automatically recommend the ideal plans of service combination according to the individual needs of the users. Case study shows that system architecture has good feasibility and efficiency.