REVIEW ON THE FINAL QUALIFYING WORKS a bachelor degree students Krutchenko Olga Vitalevna Direction 230700 "Applied Informatics" SAINT PETERSBURG STATE UNIVERSITY Theme of final qualifying work: «DEPENDENCE OF PROCEDURES OF CONTEXTUAL PREDICTABILITY ON A TEXT GENRE» Work Krutchenko O.V. devoted to the research selection methods related wordy structures in the texts in Russian. The task is an actual, as the results are used in areas such as the allocation of the terms of specialized texts, selection of proper names, key phrases and word combinations, and in a number of other tasks. The exact allocation of wordy constructions can improve the quality of decisions listed algorithms, but so far, this problem has not been solved completely. For research was used a number of known measures determining the proximity of words in a phrase, for example, t-score, MI, Dice, surprisal and others. For the processing of selected corpuses used such software and libraries as NLTK and PyMorphy, and the final calculation of statistical parameters was carried out using the Python programming language. These results demonstrate the possession of a high-level tools in the field of programming skills, as well as a high level of theoretical training. Checking the results was carried out using a cloze-tests methodology. On the one hand, the results suggest that the technique has shown its practical applicability and relevance to the general lexicon. On the other hand, the selection of informants should be carried out among experts familiar with domain terminology. In this case, the accuracy of the guesses of special terms could be higher. In addition, there are a number of errors and inaccuracies in the diploma text. Thus, for example, is not justified by the form of the entropy formula, which differs in various publications, depending on the tasks. The work contains a certain number of spelling and syntax errors (for example, on pp. 11 and those present and others), some of which were in the names of the sections (see. p. 1.2.6). Although this work is written in a lively and understandable language, and the errors do not affect the perceptual quality of the results. I believe that the final work Krutchenko O.V. is very interesting. It corresponds not only to the level of bachelor work, but also the level of master's thesis, it deserves an "excellent" (5) mark, and Krutchenko O.V. - awarding her Bachelor's degree. E.S. Klyshinsky