Development of recommendations on the safety of urban improvement facilities based on automated analysis of data from geoinformation services

The study examines the possibilities of automated assessment of the safety of urban infrastructure for the creation of recommendations to improve the socio-psychological well-being of city dwellers (using the example of St. Petersburg). Using the Outscraper service, anonymous user reviews of green spaces were collected from the object cards on the Google Maps platform. An automated analysis of the content and sentiment of reviews of facilities located in 18 city districts was conducted. The final dataset included 291 facilities (360,000 reviews). Based on this data, the main areas for improvement were identified: accessibility, safety, environmental friendliness, comfort, aesthetics, training, and mobility. The developed recommendations are highly significant in the field of urban planning and development, and can help systematically identify potential hazards and risks associated with landscape features, as well as promote more efficient use of urban resources.