A new statistical method of telemetry
Автор: Zenkov A.V.
Журнал: Международный журнал гуманитарных и естественных наук @intjournal
Рубрика: Физико-математические науки
Статья в выпуске: 3-1 (6), 2017 года.
Бесплатный доступ
A new method of statistical analysis of textual criticism. Investigated the distribution of different frequencies of first significant digits in numerals connected copyright texts in Russian language. It is shown that these frequencies approximately correspond to the Benford's law with a sharp predominance of units. Deviations from Benford's law are statistically stable copyright features available under certain conditions to explore the question of authorship, in particular, to distinguish texts of different authors. The distribution of the numbers across row {1, 2,..., 9} are subject to strong fluctuations and unrepresentative. The proposed approach is illustrated and conclusions are supported by examples of computer analysis of works of M. Ageev, Vladimir Nabokov, Mikhail Sholokhov, N. Nekrasov, etc. the Results are justified based on non-parametric U-test of Mann-Whitney and Kruskal-Wallis test.
Benford''s law first significant digit, telemetry, attribution of texts, word processing, mann-whitney test, the criterion of the kruskal-wallis test
Короткий адрес: https://sciup.org/170190322
IDR: 170190322