Research On Exploratory Service Composition Environment For Data Analysis Applications

With the rapid development of science and technology and the continuous changes in business requirements,domestic and foreign research scholars in the field of service computing have conducted a lot of exploration and achieved some staged and valuable results.Service science is a discipline technology that integrates related traditional service disciplines.This discipline technology can improve the industry level of services.In the field of distributed computing research,service computing has become an important generic and popular research direction in service science.,And plays an important and irreplaceable role in practical application areas such as urban emergency,medical services,smart grid,etc.,and service composition is the main way to build the functions of these system areas.The existing service composition technology can only solve the situation where the business process can be defined in advance.However,in practical problems,especially in the ever-changing scientific development research,the traditional method of pre-defining complete business processes by professionals has been completely unable to meet the needs of business users.This is because the needs of business users often change constantly with the changes of actual problems.At the same time,it needs professional staff to adjust the business process at the same time to meet this change in demand.However,it is difficult for professionals to implement business processes according to the needs.Synchronous adjustment;while the end user directly builds the service composition process,the end user needs to accurately and strictly define the details.This basic requirement makes it difficult for business users without professional knowledge to implement and complete.In the service computing environment,with the rapid development of the Internet of Things technology,the real-time streaming data generated by various industry fields has exploded.These data contain a lot of information,but these information are not intuitive.We need to carry out in-depth analysis and mining of the generated data,soas to obtain valuable knowledge,and the efficient and sufficient analysis of data has an important role for us to make further decisions.In response to such needs,this paper proposes an exploratory service composition method for data analysis applications.The main work of this paper includes:First,it proposes an exploratory service composition system architecture and exploratory service composition model for data analysis applications.The architecture of the exploratory service composition system includes four modules,namely a communication server module,a service library module,an operation support environment module and an exploratory service composition environment module for streaming data access.The exploratory service composition model contains the data view of the extended time dimension spreadsheet and the process view that expresses the data processing logic.Because the services targeted by this article include streaming data services in the Internet of Things environment,the traditional two-dimensional spreadsheet model can no longer meet the processing requirements of streaming data,so the time dimension is extended on this basis,and the data is completed through time windows and window aggregation functions.Aggregate calculation to complete the analysis and processing of streaming data.Second,an exploratory service composition method is proposed.Because service composition is the most commonly used method for service-oriented computing,and traditional service composition technology is no longer suitable for situations with unclear requirements,it is necessary to provide an exploratory construction service composition process that supports end users.The exploratory service composition method includes a method for constructing a data view from streaming data,an operation of a user on the data view,and an algorithm for converting the corresponding operation into process logic on the process view.Provide users with a method of directly operating the data view,so that the end user can realize the operation of the data view without having professional programming knowledge,and according to the conversion algorithm,the data view is converted into the service composition process structure on the process view,which makes the user more The convenient construction of service composition process meets the needs of users.Third,design and implement an exploratory service composition system.Based on the above service composition model and the conversion rules of the process view,a service composition system that can be directly operated by end users is realized.Experiments show that the exploratory service composition method for data analysis applications proposed in this paper can effectively reduce the complexity of the end user construction process And enhance the participation of end users,thereby improving the efficiency of end users in constructing flow data service service composition processes,and thus improving the efficiency of modeling.