Abstract: As urban expansion slows, the transformation of industrial heritage in cities has become a new research direction and the rational reuse of existing industrial heritage building space is a major point in transformation projects. Industrial buildings have special functions and they are usually designed for the production, storage, or processing of goods. They have huge internal space and the spatial layout often follows an order that reflects efficiency and production logic. Therefore, three objective problems need to be solved during the transformation of industrial buildings into public buildings: functional layout, spatial accessibility, and diversity of spatial sequences.
　　The social network analysis method greatly enriches the simple linear approach of traditional spatial planning. It enables the analysis of more complex intersections of spatial units and their three-dimensional and dynamic associations by using data association, integration analysis, text mining, and other digital analysis methods, thereby enabling the observation of many relationships and performances that are difficult to observe using a linear mindset. In this study, a case study based on the Renovation Project of Shanghai Sanlinqiao Community Centre was performed. The spatial structural rationality of the industrial heritage building park was discussed based on social network analysis and its spatial network relation model was built using UCINET. Spatial units in the park were abstracted into nodes and behavior correlations among nodes were abstracted into lines, thus building a spatial structural network and enabling a quantitative analysis of relationships among nodes. The spatial characteristics of the park could be explored by analyzing the cohesion, centrality, and association of its spatial network and spatial nodes, which assisted in the optimization of the functional layout, spatial accessibility, and diversity of spatial sequences. Five specific optimization strategies were proposed according to the calculation results of the social network analysis. (1) Determine spaces with the same function according to the calculation of cohesive subgroups and place the corresponding functions according to the specific positions of groups on the site to improve the overall balance of vitality among groups. (2) Create gray space, increase the overlapping area and spatial layers among architectural monomorphisms, enhance the direct connection among nodes, and improve the accessibility of spaces. (3) Improve the spatial hierarchy and the relationship between marginal space and the whole green area, increase the direct correlation between marginal space and other spaces, and enhance the accessibility of the overall spatial system. (4) Create outdoor space, increase outdoor public activity space reasonably, connect internal green area spaces with outdoor high-activity spaces, increase connections among spatial units, and increase the sequence options for users in the space system. (5) Improve the tightness of the spatial connections within a group, increase the complexity and diversity of public space adaptation, strengthen the interactions among nodes, and increase the combination of different spatial sequences. 
　　After applying these optimization strategies, a new spatial network model was constructed and analyzed, demonstrating a significant optimization effect. The reconstructed area had reasonable functional zoning, a stronger vitality balance of groups, higher spatial vitality, and greater spatial network cohesion. The betweenness centrality among nodes decreased, the spatial power of nodes was balanced, and the accessibility was improved. The spatial network density was generally increased, with increased direct and indirect connections among spaces, richer combinations, and a higher diversity of spatial sequences. This social network analysis provides a spatial strategy for the transformation of industrial heritage buildings into public buildings and provides a reference for similar projects. 

Key words: industrial heritage, spatial transformation, social network analysis, quantitative evaluation, transformation strategy, Sanlinqiao Community Center

摘要： 随着城市扩张速度的减慢，城市工业遗产的改造成为了一个新的研究方向，对工业遗产建筑现有空间进行合理再利用是改造项目中的一大要点。工业建筑改造过程中存在功能布局、空间可达性、空间序列多样性三个客观问题亟需解决，以上海三邻桥社区中心改造项目为例，使用社会网络分析法对三邻桥厂区空间的相关指标进行分析，提出针对性的优化策略。通过对改造后的空间重新构建空间网络模型再次分析验证了厂区的空间结构得到优化。利用社会网络分析法为将工业遗产建筑改造为公共建筑提出了空间方面的策略，可以为类似的项目提供参考。

关键词: 工业遗产, 空间改造, 社会网络分析, 量化评价, 改造策略, 三邻桥社区中心