Resumen
The article presents the possibilities of semantic text analysis in studying the discourse of neighboring online communities. The discourse generated online by such a community in a social media group is a semantic field that contains narratives about everyday life, habitat, interaction between neighbors, as well as conceptualizes the distinctive features of the urban environment. With the help of an automated semantic analysis supplemented by a qualitative analysis of the text, the key intensively discussed concepts of the «Pyat? Uglov (Five Corners)» neighboring community located in the central part of St. Petersburg were identified. The text corpus of the community online discussions was collected with Scrapy and Selenium libraries of Python programming language. The semantic fields of the key concepts were studied with the method of bigrams (stable combinations) of words that are in direct connection with each other in the text. The proposed approach is demonstrated with analysis of the key concept ?street? and its bigrams. Based on the results of the analysis, four dimensions, associated with the community? mental representation of the street, were identified: a) streets - toponyms, or street names pronounced in the discourse of the community, which reflect the actually existing fragments of the urban environment; b) problematic issues related to the organization of street life; c) value characteristics of the street; d) the mental representation of the spatial structure of the street.