THE MODEL OF INTELLIGENT ORCHESTRATION OF WEB SERVICES USING THE EXAMPLE OF STATISTICAL RESEARCH

Abstract

In organizations engaged in analytical research, the production process almost always involves statistical calculations, which serve as statistical functions of the business. The presence of tools for creating a service-oriented architecture raises questions about the development of publicly accessible means to perform such functions by assembling them into a specific plan or algorithm for solving a particular statistical task. To implement such statistical programs, it is worth considering a service-oriented architecture for creating software services that implement the execution of statistical business functions, from which a statistical program can be composed using metadata. One notable feature of such architecture is the ability to distribute tasks according to a specific area of service responsibility, also known as a domain. This architecture employs an orchestrator – a main service to which requests are made directly by users, and from which the request should reach its intended recipient. As noted in [1], orchestration becomes more complex with an increasing number of web services in an application. The main challenge lies in establishing the connection between new services, their tasks, responsibility zones, and the orchestrator. Often, such a change is not feasible without human intervention, particularly from a programming expert who needs to reprogram the main service, including deployment and configuration, which takes a significant amount of additional time. As an alternative to hard-coded interaction between services, [1] explores an approach in which discovery occurs through intelligent orchestration. The core of this approach involves each service providing its semantic description, which the orchestrator analyzes during discovery, and according to which user requests are redirected. This article provides an example of an approach to intelligent orchestration of services for conducting statistical research. Using the issue discussed in [2] as an illustration, the primary objective is to create an application for aggregating statistical data results over a specific time interval, incorporating intelligent service discovery to provide query-based result delivery.