22-04-2022, 12:57 PM
I should note that I don't have programming skills, so I work in more primitive ways, with the help of available tools (Word, WordPad, etc.). However, by showing these examples and comparisons, I wanted to clearly demonstrate the essence of my question. If the frequency of occurrences in the text of a group of bigrams in the amount of more than ten exceeds the allowable norms for ordinary languages, how can this be changed? Do you (everyone who tests with programming) mean that two (or even three) simple glyphs from a frequently repeated n-gram can stand for a single letter of plain text? What approaches could you take to reduce the frequencies of such bigrams or increase the rare bigrams? Testing and numbers are good, but I would like to see any visual logical explanation with specific examples. Or the task to check the data will appear just after machine will give suitable results?