Application of NLP neural network models to evaluate the effectiveness of measures to create a comfortable urban environment

Abstract

The formation of a comfortable urban environment is one of the urgent tasks of modern urban planning. To achieve this goal, in 2018 the President and the Government of the Russian Federation launched the national project “Housing and the Urban Environment”, within the framework of which an exhaustive list of measures to create a comfortable urban environment was determined. The purpose of the work is to form a rating of the effectiveness of such events based on the analysis of the texts of messages from citizens, users of social networks, texts of publications in the media and descriptions of events held in a given territory. The proposed approach is based on an assessment of the semantic similarity of the description of events from a set of datasets for a certain territory. On the basis of the datasets formed for specific municipalities, a discrepancy between the expectations of citizens and the implemented activities in this territory was revealed.