Data Analysis System based on the Intelligent Agent

This paper considers the multi-agent approach to designing architecture of the information and analytical systems, which implement the complex multi-stage diagnostics of objects. Implementation of the multi-agent approach is demonstrated using the example of a medical information and analytical system with a specialized diagnostics module, which allows us to predict the possibility of surgery for cancer patients. Typical system architecture is presented consisting of four basic modules based on the intelligent agents and of the fifth special multi-agent diagnostics module, which specialization and operation depend on the data analysis purpose and technology. As an example, operation of the basic information collection module was considered, one of its tasks was to determine the incoming data format. The diagnostics module analyzes various types of information and is the most complex. Its implementation is proposed to solve a highly specialized problem of medical diagnostics related to assessing a possibility of a successful operation in the elderly patients with the oncological diseases in the gastrointestinal tract.