As the most typical public facility, public toilets reflect the civilized level and management service level of the city and are an important window for building a civilized image of the city. Current research focuses on the accessibility and coverage of public toilets, treats public toilets as spatial points without discrimination, and ignores the heterogeneity of public toilets in different urban functional areas. How to establish a comprehensive and accurate public toilet space evaluation system and analyze the comprehensive service capabilities of public toilets in different regions is obviously insufficient in the current research, which is not conducive to the deployment of public toilets and the advancement of the equalization of basic public services. The emergence of multi-source data provides a new perspective for the research of urban public facilities. Therefore, this paper proposes a rationality evaluation method of public toilet spatial layout based on POI big data from the perspective of urban functional area. We use Term Frequency-inverse Document Frequency (TF-IDF) information weighting technology combined with Point of Interest (POI) frequency density to identify urban functional areas, and integrate OpenStreetMap (OSM) road network density and WorldPop population data to construct a population travel vitality index and evaluate public toilet services in urban functional areas. Finally, the population and spatial coverage rate and the spatial imbalance index are calculated to determine the difference between streets and towns and the rationality of the layout of public toilets in streets and towns. Based on multi-source data, this method quantitatively analyzes the rationality of the allocation of public toilets in different functional areas and explores the differentiating factors of the space allocation of public toilets. This article takes Shanghai, one of the most urbanized cities in China, as an example for calculation. The study finds that: ① The number of toilets in different urban functional areas is different. The number of planned commercial service functional areas is the largest, but the public toilets in the commercial service functional area have the highest qualifications. "Industry-Commercial service" and "Green Space-Commercial service" and other commercial service-related joint functional areas are also at a high level of qualification. This is because a lot of commercial service organizations provide public toilet services to the outside, which has improved the service capacity of public toilets in the area; ② The public functional area has the lowest qualification rate, only 10.27%, which is related to the openness of this type of attached public toilet facilities; ③ The eligibility of public toilets in Shanghai’s streets and towns is generally reasonable, with an average space coverage rate of 67.31% and an average population coverage rate of 70.72%. The imbalance index between public toilet service and travel vitality distribution in streets and towns ranges significantly from 0 to 0.76, of which 147 have an imbalance index less than 0.4, accounting for 69.34%, indicating the comparison of the space allocation of public toilets between these streets and towns is reasonable. The rationality of the allocation of public toilets decreases significantly from the downtown area of Shanghai to the southwest, southeast, and Chongming Island, while there is no obvious attenuation to the northwest, showing a good contiguous service capacity. The results show that this method fully takes into account the issue of the heterogeneity of the functional areas of public toilets and the travel vitality of the population, and the spatial analysis is more accurate and has more practical value.

Night-time economy refers to the related economic activities mainly in the services taking place in urban space and at night, which is an important representation of a city's economic development and consumption level. Currently, researchers at home and abroad mostly rest on the theoretical level, or small-scale refined research based on market research and questionnaire survey, lacking in-depth mining using models and mathematical statistics methods, and rarely intuitively show the specific temporal and spatial distribution of large-scale night-time economy. With the development of information technology, night lighting data and perception big data provide new data sources for quantitative research of night-time economy. This paper provides a new perspective for night-time economy by fusing multi-source data. Compared with the traditional survey data, it is more rapid, efficient, and extensive, which is suitable for large-scale research of night-time economy. Based on taxi OD flow, this paper uses spatial clustering algorithms such as DBSCAN and K-Means ++ to identify hot areas of night-time activities in Xiamen City from the perspective of consumers. Based on the night-time lighting image and POI, this paper analyzes the supply and demand relationship by the method of profit and loss and identifies the distribution area of night service facilities from the perspective of merchants. Then we analyze the temporal and spatial distribution pattern of night-time economy in Xiamen City. The results show that: ① The spatial distribution of night activities in Xiamen City is multi ring and decreases to the surrounding areas. The distribution of hot spots of night activities varies from place to place; ② The existing service facilities in some areas of Xiamen City fail to serve the night economy well, and the existing lighting infrastructure, such as lighting and nightscape, is insufficient; ③ There is a moderate positive correlation between residential population density and night activity density, and the results are valid. The profit and loss value and quantity of night service facilities, night lighting, and night activity density are moderately and weakly correlated, and catering facilities are more dependent on night lighting. Finally, we put forward some suggestions for Xiamen's future night-time economic construction, such as providing different night-time services according to different consumer groups and psychology, strengthening the construction of night light infrastructure and market support. The research conclusions are of positive significance to promote social employment, and enhance the utilization rate of infrastructure. At the same time, they can also provide reference for urban economic development and policy formulation.

Urban hotspots prediction is a basic but significant task for future urban management. Accurate urban hotspots prediction can improve the efficiency of urban planning and security construction. Existing deep learning methods mainly adopt geographic grid maps, provided urban network, or external data to capture spatiotemporal dynamics. However, we observe that mining some latent self-semantics from raw data and fusing them with geospatial based grid images can also improve the performance of spatiotemporal predictions. In this paper, we propose Geographic-Semantic Ensemble Neural Network (GSEN), a novel deep learning approach, to stack geographical prediction neural network and semantical prediction neutral network. GSEN model integrates the structures of Predictive Recurrent Neural Network (PredRNN), Graph Convolutional Predictive Recurrent Neural Network (GC-PredRNN), and Ensemble Layer to capture spatiotemporal dynamics from different views. Furthermore, this model can also be correlated with some latent high-level dynamics in the real-world without any external data. We evaluate our proposed model on three different real-world datasets. The experimental results demonstrate the generalization and effectiveness of GSEN in different urban hotspots spatiotemporal prediction tasks.

It is of profound significance in the fields of multi-scale map expression and digital mapping to recognize the architectural pattern intelligently by integrating spatial relations, geometry, semantics, and other features. Applying graph convolution neural network to intelligent recognition of building pattern can overcome the shortcomings of traditional methods, such as relying on artificial experience to set parameters and making rules too strict. However, this method often has the problem of sample proportion imbalance, which easily leads to the category with a small number of samples can not be properly identified. In this paper, the building centriod is used as the sample to cluster to obtain the building groups. Secondly, the building centriod in the same group is used as the node to construct the graph structure of the building group of Delaunay triangulation. The feature selection of the graph node can describe the area, size, direction, and other related indicators of the geometric characteristics of the building group. Thirdly, the building groups graph structure with small number of samples is enhanced by oversampling the graph structure. Then the data before and after the enhancement of the architectural group graph structure with small number of samples are input into the graph convolution neural network model for training, and the performance of the model is evaluated with several evaluation indexes such as ROC curve. The experimental results show that the recognition accuracy of the model for buildings with lesser number of samples is significantly improved after the structure of buildings with small number of samples is enhanced.