Application of Spatial Clustering Analysis in Identifying Patterns of Affordable Housing Construction: A Developer-Based Study in Melaka

Abstract

Housing affordability remains a pressing challenge in Malaysia, particularly for middle-income groups, requiring more spatially informed and data-driven approaches to guide affordable housing development. Spatial clustering analysis provides critical insights into demand concentrations, spatial disparities, and location suitability; however, its use is limited by inconsistent data quality, a lack of standardisation, and insufficient integration of socio-economic factors. This study investigates the extent of spatial clustering technique use among developers, identifies key strategies to enhance their use, and examines the relationship between these two components. A quantitative correlational design was adopted, with data collected from private developers in Melaka via structured online and in-person questionnaires. The study achieved a strong response rate of 94.4% (n = 34), ensuring robust and representative findings, which were analysed using descriptive statistics and Spearman’s correlation. Results indicate a high overall application level, with DBSCAN, GIS and Local Moran’s I being the most widely used tools for identifying suitable locations, detecting spatial patterns and assessing inequality. Developers prioritise enhancement strategies such as land-use and zoning analysis, advanced clustering algorithms and cross-sector data sharing. Correlation analysis reveals consistently positive but varied relationships, with the strongest relationship found between the use of descriptive spatial statistics and zoning-based strategies for industrial land renewal, indicating that statistical spatial summaries play the most influential role in guiding legally and environmentally aligned redevelopment decisions. Overall, the study provides actionable insights for developers and policymakers, promoting more effective, sustainable and equitable planning of affordable housing in Malaysia through enhanced spatial clustering integration