The Statistical Parameters of Ivan Franko’s Authorial Style Determined by the Chi-square Test
Iryna Khomytska, Iryna Bazylevych, Vasyl Teslyuk · 2022 IEEE 17th International Conference on Computer Sciences and Information Technologies (CSIT) · 2022
The novelty of the research is the method of a combination of an analysis of the absolute and relative frequencies of occurrence of letters of Ukrainian and the chi-square test for identification of Ivan Franko’s authorial style in the literary works ”Na loni pryrody’ and “Malyy Myron”. According to the results, typical of Ivan Franko’s authorial style are the letters of Ukrainian, the values of absolute frequencies of occurrence of which differ in less than 10 units (б, з, т, у, ф, х, ч, ц, ш, щ, ю). The same percentage has been obtained for the relative frequencies of occurrence of the following letters: б, в, г, д, ж, з, к, у, ф, ч, ц, ш, ь, ю, я. The applied statistical method – the chi-square test has shown that the differences between the literary works ”Na loni pryrody’ and “Malyy Myron” are not statistically significant and consequently, the works are written by the same author. The results have proved that the chi-square test is efficient for solving the authorship attribution task on the level of letters. The test validity is 95%. The used software is Python and R. The practical application of the research is identification of the authorial style on the level of letters.