Data Optimization for Industrial IoT-Based Recommendation Systems

The most common problems that arise when working with big data for intelligent production are analyzed in the article. work of recommendation systems finding relevant user information was considered. features singular-value decomposition (SVD) and Funk SVD algorithms reducing dimensionality providing quick recommendations were determined. An improvement algorithm using a smaller required amount analysis proposed. According to results experiments, proposed modification improves speed processing on average by 50–70% depending number users allows spending fewer computing resources. As follows, provided shorter period more relevant. faster calculation modified exchange optimal parameters between nodes It determined execution time can be reduced 75% ten exchanging parameter compared one. Using Spark technology 20% Hadoop architecture IIoT system proposed, which uses optimize edge devices monitors effectiveness control centers cloud