Tools for automated analysis of fiction work translation

The paper describes the features of the creation and approbation of tools for automated analysis of the original fiction text and its translation to assess the adequacy of the transfer of meanings for the philosophical category of quality. The material of the study is a text of the novel by F. Kafka “The Castle” in German and its translation into Russian. A method of automated text analysis is applied, which combines objective programmatic and subjective speculative analysis, as well as elements of the corpus approach, for which the spaCy natural language processing Python library is used, which has various functions for analyzing unmarked-up text arrays. The stages of creating a method for automated analysis of the qualitative characteristics of the original fiction text and its extrapolation to the translated text are considered in detail, taking into account the formal differences in the selected languages. The developed method allows the authors to provide for the normalization, tokenization and frequency analysis of textual material, to visualize and interpret the obtained empirical data, among which the deviation coefficients seem to be the most relevant. The authors come to the conclusion that the use of modern software tools for processing unmarked-up texts can be effective in identifying differences in the perception of a qualitative characteristic in the original text and the translated text. The established differences may be the result of the choice of inaccurate translation options. As a promising direction for improving the tools, it is proposed to work on improving the accuracy of the output data by involving the latest software solutions, as well as approbation of the tools on the material of fiction works in other foreign languages and their translations within the philosophical categories of quality, space and time.