Comparison of statistical methods and application of Grabbs's criterion for determining and eliminating a contradiction

Abstract

The use of conflict detection methods is not limited to the search and correction of errors in the data, they can also be used to find distorted data. Text analysis algorithms are used to detect inconsistencies and identify artificially created copies of original texts.This article discusses the effectiveness of statistical methods used to identify and eliminate inconsistencies and anomalies in large data. Compliance with the law of normal distribution of data was determined. The study is based on the comparison of statistical criteria with the Grubbs criterion, which gave high results in identifying conflicting data. A block diagram of the general algorithm of the Grubbs criterion was created and the results of iterations obtained in the Python programming environment were presented.