This article aims to study the design evaluation mechanism of Shanghai’s historic districts’ revitalization and renewal based on multi-source data. It will also select four distinctive historic districts in Shanghai and excellent cases of district regeneration in China as comparative references by using intelligent and informative design evaluation methods. Multi-source data technology is used to explore Shanghai’s renewal elements in four dimensions: spatial pattern, architectural features, alley space, and landscape greening, and apply the hierarchical analysis method to analyze the spatial environment of the historic districts quantitatively. Secondly, based on the geographic information data and field research information, the spatial syntax method is used to make a comparative analysis of the spatial accessibility of street accessibility. Structured networks and Python are used to obtain the data of the government platform, street scene pictures, Baidu search word frequency, POI industry distribution, and other open data on the network. Deep learning, a full convolutional network, is collected to semantically segment the street scene pictures and make a perceptual assessment of the content of the urban imagery and spatial quality evaluation based on collected street scene pictures; the perceptual assessment of the content is made. Meanwhile, the fuzzy evaluation hierarchical analysis method is used to make a fuzzy comprehensive judgment of the indicators. Finally, the enhancement of the evaluation index data of road accessibility, functional diversity, environmental quality, and district vitality in the research results makes this multi-method cross-validation improve the scientific and practical aspects of the research results and provide more accurate decision support for the subsequent design practice to reduce the renewal cost of the historic district.