This article presents a model-based approach that integrates mathematical methods and neural network semantic algorithms to analyze well-being, identify the presence or absence of social tension, and provide predictive analytics for the development of events related to a major urban planning project. The database included relevant materials, comprising a total of 33,700,700 tokens. Data collection was conducted from January 1, 2021, at 23:59:59 to December 19, 2021, at 23:59:59. All algorithms produced identical results, indicating a positive attitude of residents towards the project’s implementation and the absence of social tension. The analysis and interpretation of the data enabled the construction of a forecast for the conflict-free development of the situation surrounding the project, which was subsequently confirmed by the actual course of events.