Approximate Entropy as a function of the size of m-grams in 14 European languages.

The Approximate Entropy (ApEn) as a function of m-grams for different texts in 14 European languages and Esperanto. The values show a consistent statistical behaviour for the languages and independence of the books. The ApEn method works taking into account the structure of the language at different levels. The randomized versions of all the texts are similar, showing a thin dependence on the difference in the number of letters of the alphabet. Esperanto shows lower values of ApEn for the same m-grams, and the random versions high values at short m-grams and an abrupt down about m= 5. Vertical bars indicate the standard deviation. The lower right corner values from 0 to 2.5 only aplies to the random texts.