Abstract: As a continuous, large-scale, and multi-attribute network system, streets can be accurately controlled by finely characterizing their network measurement system to plan intervention directions in stock updates. At the same time, the network type classification mode can also guide the refinement design strategy of streets. Taking residents' mobility, interaction, and ecological behavior as the core, using the concept of complete streets as the medium, and multi-attribute street network elements as the guarantee, establish a multi-attribute street network measurement system covering transportation, society, and nature. Selecting Dalian City as the research area, principal component analysis, case comparison, and cluster analysis were used to evaluate the characteristics of street networks with different attributes. Research has found that convenient transportation networks have relatively high levels of density, form, and proximity to bus stops, which can ensure the safety needs of mobile behavior and are in line with the goal of safe streets; Social service-oriented networks have a relatively high level of location advantages, functional complexity, and interface quality, which can to some extent meet the requirements Rich demands for social interaction behavior, reaching consensus with the goals of dynamic streets; The blue-green level and accessibility of natural dominant networks are relatively high, which can guide ecological behavior and better match the goals of green streets; However, the deficient network is seriously inadequate in terms of traffic safety, social vitality, and environmental quality. It is urgent to improve the quality of the street network through measures such as improving station proximity, functional support, and green quality. Finally, suggestions and references are provided for the refinement of street design and the revitalization of existing street networks in the context of urban renewal.