Модерація текстового контенту з використанням комбінованого ансамблевого підходу на основі алгоритмів машинного навчання

Abstract

The object of the research is the process of automated text content moderation in digІТal environments.The aim of this work is to develop and evaluate the effectiveness of a combined approach to text content moderation using modern machine learning methods.The research methods include the analysis of existing text classification algorІТhms (Naive Bayes, SVM, Logistic Regression), the creation of an ensemble model wІТh Gradient Boosting as a meta-model, and the evaluation of methods based on accuracy, precision, recall, and F1-score metrics.