A Framework for the Development of Digital Spatial Weights using Social Media Data: Methodological Foundation and Initial Analysis

Abstract. The rapid growth of internet users has significantly impacted human activities, prompting a re-examination of spatial definitions in digital contexts. This study provides a framework for representing space digitally using social media data (Twitter, now X) and compares digital space representations with traditional physical space representations. Spatial weight matrices were developed to represent both physical and digital spaces in selected cities in Metro Manila and Quezon City (QC) barangays. Four digital spatial weight schemas were created using Twitter data to facilitate this comparison. Digital and physical spatial weights were then compared through matrix and weight maps. The results show that digital and physical spaces represent distinct spatial realities. Although exploratory in nature, the study highlights the temporal and spatial effects of using digital spatial weights compared to physical spatial weights. A key advantage of digital space representation is that it bypasses the disconnected spatial units issue present in most common physical representations, and temporal variations can be incorporated. This allows for more subtle relationships to be captured in spatial and temporal analysis. Since most phenomena occur in both physical and digital spaces, a combined representation of both may provide a more accurate model. Overall, the comparisons reveal the differences between digital and physical spaces in terms of matrix structures and weight maps. It is recommended to further explore spatial weights in the context of modeling implementations and the integration of both spatial representations.