IMPROVING THE FALSE INFORMATION DETECTION METHOD USING THE BAYESIAN CLASSIFIER

Abstract

The article provides an analysis of solutions to applied problems, which are solved by applying methods of cluster analysis. These are classic Data Mining tasks: clustering, classification, and tasks typical only for text documents: automatic annotation, extraction of key concepts, etc. An abbreviated analysis of information clustering methods was conducted. The postulate that clustering involves dividing a set of elements into clusters, the number of which is determined by the localization of the elements of a given set in the vicinity of some natural centers of clusters, is confirmed. An analysis of the application of the Bayesian classifier was carried out. it is proven that the Bayesian classifier, in the presence of a priori probabilities, works with high accuracy in determining false information. However, this method does not give an answer about obtaining this probability. Due to the use of a naive Bayesian classifier to detect false information, the method of detecting false information has been improved. This method allows you to solve the problem of the uncertainty of the a priori probability. The proposed naive Bayesian classifier for text processing turned out to be quite effective. The effectiveness of the algorithm of the proposed method of developing a classifier for determining the veracity of information was evaluated. On the basis of the primary data obtained from the Internet, the numerical values of the evaluation of the algorithm of the improved method of determining false information were calculated. The obtained metric values are: Recall = 0.853; Precision = 0.869; F-measure = 0.861; Accuracy = 0.855. The obtained results prove that the improved method (without additional training) immediately has good results. This proves the adequacy of the developed method and provides an effective scientific method for detecting false information. The improved method of detecting false information is especially relevant at the present time, in the conditions of an information war.