Application of machine learning methods for analyzing data from the nomenclature directory of the enterprise resource planning system

This article is a part of research and development of software for a corporate directory of materials in the UMMC CIS. The developed product should allow to automat a number of functions that are currently performed with considerable laboriousness or require a long data processing time. The data set was prepared and analyzed. The analysis of data consisted in multiclass classification. The following methods were used: random forest, naive Bayes and XGBoosting.