Перекласти текст EN-UA

Потрібно перекласти невеликий кусочок тексту:

Knowledge of the statistical properties of strings can be used to infer words only to the extent that the device that generated the strings in the first place used knowledge of words, and only knowledge of words, to generate the string in the first place; and, in actual fact, the systems that generate our natural language strings employ systems at several levels: it is not words, but morphemes that consist of relatively arbitrary sequences of letters, and words are the result of a system responsible for the linear placement of morphemes. In addition, there is a system responsible for the sequen-tial placement of words – we call it syntax – and it too has a great impact on the statistics of letter placement. A system that tries to learn the structure of language on the basis of a model that is far poorer than the real structure of language will necessarily fail – we may be impressed by how well it does at first, but failure is inevitable, unless and until we endow the learning algorithm with the freedom of thought to consider models that take into consideration the structure that indeed lies behind and within language. In the next section, we turn to the task of learning morphological structure.

Хороший відгук при оперативній та якісній роботі ГАРАНТОВАНО!