Model of Large Sparse Datasets Processing Efficiency in IIOT

The paper studies the peculiarities of processing big data in industrial systems. Existing methods of optimizing user information and the effectiveness of recommender systems have been considered. The Funk SVD algorithm was analyzed, and the practicality of its application to large sparse data was determined. A modification of the Funk SVD algorithm was proposed, which employs a smaller amount of user data to form recommendations. It was also suggested to use additional parameters that better characterize the services for more accurate work of recommender systems. Simulations were carried out, and a decrease in the duration and an increase in the accuracy of calculations using the proposed methods were determined.