如何测度和解释社会空间分异一直以来都是城市地理学和社会地理学研究的重要课题。近年来国际学术界越来越关注社会空间的多元性和动态性，社会空间分异研究逐渐从单纯居住空间尺度向整体活动空间扩展。然而已有研究侧重于活动空间分异模式刻画，分异特征综合评估和影响因素研究不足，难以解释居民活动空间分异的时空格局和发生机制。本研究整合多源时空数据，在构建和完善活动空间分异研究框架的基础上，探索多元与动态度量指标体系，分析居民活动空间分异的多维特征；通过建立小数据与大数据的关系映射，在多个尺度研究日常活动参与、群体特征和城市空间对活动空间分异的作用机制与内在机理，识别高活动空间隔离风险的活动-移动模式、社会群体和空间环境；并探讨降低活动空间隔离风险和提升社会活力的城市时空资源配置方案，为城市空间规划与社会管理提供建议；并通过理论探讨和实践研究完善活动空间分异研究体系，以丰富社会空间分异理论。

How to measure and explain the socio-spatial differentiation has always been an important topic in urban geography and social geography. In recent years, the international academic community has paid more and more attention to the diversity and dynamics of social space. The research on socio-spatial differentiation has gradually expanded from residential area to the activity space, to evaluate the overall exposure to different social environment in individual’s daily life, which called activity-space-based differentiation. However, the existing activity space researches focus on the spatial pattern of activity-space-based differentiation, but the comprehensive evaluation of the multi-dimensional differentiation and influencing mechanism are insufficient. Thus it is difficult to explain the spatial and temporal pattern and mechanism of activity-space-based segregation among different social groups. .This study seeks to construct a comprehensive framework of activity-space-based differentiation study, hoping to expand the content of socio-spatial differentiation theory. First, by integrating multi-source spatial and temporal behavior data, this study measures the multi-dimensional pattern of activity-space-based differentiation, including spatial structure, social interaction and social exposure. Second, the multi-dimensional characteristics and influential mechanism of activity-space-based differentiation are studied using a case study in Shanghai. By combining activity diary data and big data, how different pattern of daily activity participation, vulnerable groups and urban space contribute to the pattern of activity-space-based differentiation, and even segregation, is studied using multi-level models. This study intends to identify the activity-moving mode, social group and space environment of high activity space isolation risk. Third, this study seeks to explore the urban space-time resource allocation plan to reduce the risk of activity space isolation, providing suggestions for spatial planning and social management. At last, this study intends to improve the research of activity-space-based differentiation through theoretical exploration and practical research, to enrich the theory of socio-spatial differentiation.