Although urbanisation and urban land limitations are prevalent, residential complexes are seldom discussed as a means of compartmentalising social interactions and contributing to residents’ social interactions. Due to the spatial orientation of social interactions, semi-public spaces in these buildings are unable to host residents’ interactions due to a lack of appropriate arrangements/establishment of tangible and visible objects. The influential components, however, have rarely been identified in residential complexes. To fill this gap, using the theory of human territoriality, the current study explores the influential physical components of human territorialities in semi-public spaces in four residential complexes and investigates whether a significant relationship exists between human territorialities’ physical components and social interactions. This study collected data from 264 residents of four residential complexes with different spatial configuration layouts. Pearson correlation and Spearman’s rank correlation are used to evaluate linear and monotonic relationships between study variables. The correlation between spatial configurations of physical components and residents’ interactions confirms that spatial configurations influence residents’ use of semi-public spaces. Social interaction can, therefore, be improved through green space, brightness, accessibility, and furniture in common areas. The findings prove that residential complexes with clustered arrangements have not performed well in creating social interaction due to the lack of defined spaces and territories for people, but multi-core, mixed, and linear complexes that define several open and semi-open spaces have been more successful in the amount of social interaction of residents. This study is one of the first to identify the influential components using the integration of residents’ perceptions and spatial configuration.