Understanding investment-construction processes and their dynamics is crucial for effective urban planning, but these processes are often complex and fragmented. While advances in GeoAI offer powerful new tools for analysis, conventional approaches to studying urban development face two key limitations. First, they frequently neglect to represent the processual characteristics of development, regarding complex, multi-phase projects as a succession of isolated incidents. Second, the analysis of what is being built is usually limited to broad, formal categories that hide the differences between projects. This results in a methodological gap, distinguishing the analysis of project intentions (semantics) from their structure (quantitative attributes). To address these limitations, this study introduces and validates a novel GeoAI-driven methodology that uses building permit data as a high-resolution proxy to conduct an integrated analysis of investment-construction processes. By combining formal (quantitative) and semantic (textual) dimensions, our approach provides a comprehensive understanding of these dynamics, demonstrated through an extensive analysis of over 60 000 building permits issued in Wroclaw, Poland (2006–2023). The core methodology involves three primary contributions. First, after using graph analysis to aggregate individual permits into cohesive investment components, we introduce the Investment Complexity Index (ICI), a novel synthetic metric designed to quantify the diversity, scale, and duration of these processes. Second, we employ semantic clustering (using BERT) to analyze textual permit descriptions, creating a nuanced typology of what is being built. Third, we apply quantitative clustering (using AHC) to a comprehensive set of attributes to identify how these investments are structured in terms of their spatial, temporal, and sequential features. This study shows that the proposed integrated analytical approach, which combines quantitative, semantic, and synthetic measures of complexity, provides a much richer and more nuanced picture of investment processes than one-dimensional methods. This has important implications for urban research and planning practices and offers tools to better understand and manage the dynamics of urban development. • An integrated methodology to analyze investment-construction processes is proposed. • Graph analysis transforms discrete building permits into coherent development processes. • A novel Investment Complexity Index (ICI) quantifies process intensity. • The methodology provides a data-driven tool to monitor urban planning policies. • Integrates textual descriptions (semantics) with spatiotemporal attributes (structure).
Kaczmarek et al. (Sat,) studied this question.