This study examines street-level patterns of mixed non-residential uses in Type II General Residential Zones in Seoul using POI-basedclustering and multinomial logistic regression models. Although non-residential activities are commonly found within r...
This study examines street-level patterns of mixed non-residential uses in Type II General Residential Zones in Seoul using POI-basedclustering and multinomial logistic regression models. Although non-residential activities are commonly found within residential areas,area-based zoning designations remain too broad to capture the functional diversity that emerges at the street level. POI data were collectedthrough the KakaoMap API, and point-based land-use indicators were constructed and linked to individual street segments. Clustering analysiswas then applied to classify street-level non-residential use configurations, followed by regression analysis to identify the factors associatedwith their formation. The results reveal clear functional differentiation among streets within the same residential zoning category, indicatingthe coexistence of diverse spatial structures. While morphological characteristics such as street width and linearity partially explain thesepatterns, non-morphological factors, including retail demand and business competition, also play important roles. The interaction betweenphysical and socioeconomic conditions produces complex, multi-layered spatial structures. These findings highlight the need for more flexibleand fine-grained land-use controls that can respond to the diverse street-level land-use patterns observed within residential areas.