After forty years, that have passed since Georgetown experiment, making the system of automatic (automatized) translation is a great problem as before.

The developer of the system Proton-S refuses the idea of full-automatic translation and concentrated on making the interactive system, which assumes the intense dialogue with the user. During the dialogue there is a mutual training: with forming the context-dependent vocabulary in the computer, the user corrects the obtained Russian translation and so began to recognize the syntactic relations in the English sentence.

To increase the quality of productions it is useful to accept as unit of translation not only individual words, but also word-combinations, including whole phrases. To provide this selection of translation unit the original (distant from modern linguistics formalism) technique of working with the "bunches" was developed.

The author of the project sees his task not in creation of big English-Russian dictionary, but in placing at user's disposal necessary set of tools for making this dictionary just for required subject area. The start knowing of English is reduced to ability to find the translation of unknown word in ordinary dictionary. The number of possible users of the system may so be highly great.

The initiator of the task, the developer of the system, and its programmer is the same person - author of the project Mikhailenko Nikolai Nikolaevich. Owing to this triune, there is possibility not only highly to cut down expenses for realization, but retain the major parameters of the system within reasonable limits. They are: rate of translation, requirement of core and free area on the hard disk.

Eight years of working over this project allow author to accumulate highly original experience and create practically working interactive system. However the complicity of the given task requires considerable labour for developing it: improvement of syntactic analyzer, expansion of the structure of article, revision of presentation of grammatical information and detailing of dialogue with user.

This project undoubtedly deserves financial supporting.