Sentiment and Conflict Prediction in Urban Development: Data-Driven Approach

The article presents a methodology for detecting and analyzing social tension in a metropolis using neural network and mathematical models built on time series. It considers the problem of assessing and predicting the development of the situation in real time, based on the content generated by users and their digital footprints, as illustrated by the implementation of a transport project. The integration of neural network and mathematical models made it possible to identify semantic negative accents, determine the features of project positioning in the media space, identify segments of the greatest informational attention, the level of social tension around the construction project, and also predict the development of the situation. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.