Public participation in urban regeneration using XAI-driven decision support system

Contemporary methods of community engagement in urban regeneration and, in general, the concept of citizen participation continue to show serious deficiencies in practice, leading to partial or manipulated forms of engagement. The most vulnerable and disadvantaged groups often have little say, while those more familiar with institutional mechanisms manage to influence outcomes to their advantage. This paper attempts to approach this persistent imbalance from a new angle by employing Artificial Intelligence (AI) algorithms and Explainable AI (XAI) tools not merely as analytical instruments but as means of understanding the deeper social and behavioral factors that shape participation and exclusion. The study establishes a base framework for a decision support system using models such as Decision Trees, SVM, and XGBoost, together with interpretability tools including SHAP, LIME, and ELI5, to identify and explain the determinants that influence public engagement within deteriorated urban areas. The purpose is to explore whether these computational methods, when interpreted with care, can illuminate behavioral patterns and barriers among underrepresented communities and help design more inclusive and responsive urban regeneration processes.