The monograph outlines the range of problems associated with the features of automatic analysis of spoken Russian in interactive dialogue systems. Methods of remote recording of speech, taking into account the variability of pronunciation in spoken speech, compact presentation of the dictionary, as well as syntactic and statistical modeling of the language in automatic speech recognition systems are described.