News Text Classification Based on Naive Bayesian Method of Factor Analysis
News text classification is an effective way to manage network news.Naive Bayes is a common method for news text classification.When the feature dimension is large,the conditional independence assumption is difficult to be estab-lished,which will lead to poor classification effect.This paper proposes a naive Bayesian classification model of factor analy-sis,and classifies news texts in the text corpus of Fudan University.The classification results are compared with naive Bayes,multinomial logistic regression model and decision tree method.The experimental results show that the classification accuracy of naive Bayes classification model for factor analysis is 84％,and the effect is optimal.
naive Bayesianfactor analysisnews text classificationmachine learning
万方数据
维普
