Approaches to automation processing of user requests in a multi-level support service using featured models

Abstract

The article is devoted to the development of the functional subsystem for automatic classification of incoming requests from users of Methodological, Technical and Consulting Support of the Ministry of Education and Science of Russia. Text mining tools are used to classify incoming requests. The paper considers two text representation models-Bag of words and Word2Vec, and compares different classification methods, such as logistic regression, random forest, and gradient boosting. To improve the results, optimization approaches are used-class system restructuration and introducing the technology of two-level classification.