Literature review on urban landscape and behavioural science: examining past research and proposing probabilistic and optimization frameworks for advancing future methodological approaches

Urban Landscape and Behavioural Science (ULBS) research has significantly advanced with the rise of data-driven methodologies, yet traditional approaches often fall short in addressing dynamic urban complexities. This review presents an innovative framework for ULBS research, integrating advanced techniques like Agent-Based Modeling, Machine Learning, AI, and Big Data Analytics. Through a systematic review of literature from 2010 to February 2024 across Scopus, Science Direct, and Taylor and Francis, we analyzed 212 methodologically influential articles, identified through bibliometric and thematic analyses, to uncover spatiotemporal trends, methodological advancements, and critical research gaps. A key contribution is the introduction of advanced probabilistic and optimization frameworks, extending beyond basic Bayesian models by incorporating weighted priors, information gain functions, and multi-objective optimization for evaluating methodological suitability. These frameworks enable adaptive method selection based on data characteristics, constraints, and objectives. Our findings highlight the need for real-time analytics, predictive modeling, and interactive tools to address data representativeness, methodological limitations, and interdisciplinary challenges. By promoting sophisticated evaluation techniques and integrating ethical considerations, this review aims to enhance the precision of ULBS research, supporting the development of sustainable, resilient, and inclusive urban environments that are better equipped to respond to the complexities of modern urban living.