A Methodology for Data Integration in Decision Support Systems for Critical Infrastructure

Abstract

Tools and methods for the automated collection, processing, and integration of data for scientific analytical systems in critical infrastructure are analyzed. A three-level methodol-ogy is proposed, which includes: collecting information from official, commercial, and unofficial APIs, web resources, and streaming platforms; preprocessing of data—cleaning, normalization, anonymization, and enrichment through cross-source integration; and the creation of a consolidated knowledge base for further use in predictive models and deci-sion support systems. It has been shown that combining web scraping, streaming tech-nologies (Kafka, Flink, Spark Streaming), and intelligent processing allows for working with heterogeneous structured and unstructured data in near real time. Practical examples from the energy, transportation, and cybersecurity sectors demonstrate the effectiveness of this approach. Future development prospects lie in integrating these methods with ma-chine learning and explainable AI to improve the accuracy and transparency of analytical systems.