Under the background of the park city, there is a new trend in the development of park green space. Citizens carry out night recreation activities in the park green space, which changes the traditional recreation mode and stimulates the city’s vitality at night. Based on mobile phone APP and POI multi-source big data, this paper preliminarily reveals the differences in daytime and nighttime recreational activity levels and the structure of dominant influencing factors. Taking 12 typical park green spaces in Shanghai as examples, the monitoring and comparative analysis of recreation heat value was carried out, and the quantitative analysis index “day and night activity” of recreation activity level was proposed. The results show that the proportion of pure day active park green space is only 16.67%. Thus the park green space recreation mode has changed from the traditional day mode to day and night multi-mode. This paper further analyzes the leading influencing factors that affect the night activity level of the park green space. Through the correlation analysis of the night recreation heat value (H) of the park green space and the internal and external influencing factors, the influence ranking and correlation structure diagram of the significant related factors are drawn. The research results optimizing the supporting facilities of park green space services in high-density urban areas and improving the level of refined management of park green space, thus having a positive promoting effect on the planning and design of park cities.