Evidence has suggested that built environments are significantly associated with residents’ health and the conditions of built environments vary between neighborhoods. Recently, there have been remarkable technological advancements in using deep learning to detect built environments on fine spatial scale remotely sensed images. However, integrating the extracted built environment information by deep learning with geographic information systems (GIS) is still rare in existing literature. This method paper presents how we harnessed deep leaning techniques to extract built environments and then further utilized the extracted information as input data for analysis and visualization in a GIS environment. Informative guidelines on data collection with an unmanned aerial vehicle (UAV), greenspace extraction using a deep learning model (specifically U-Net for image segmentation), and mapping spatial distributions of greenspace and sidewalks in a GIS environment are offered. The novelty of this paper lies in the integration of deep learning into the GIS decision-making system to identify the spatial distribution of built environments at the neighborhood scale.
The aim of this study was to assess the correlation of depression and anxiety with time spent at home among students at two universities—one urban and the other suburban—during the COVID-19 pandemic. Methods: Geolocation data from the smartphones of 124 participants were collected between February 2021 and May 2021. The level of depression was estimated by the PHQ-9 and PHQ-2 screening tools, and anxiety scores were estimated by the GAD-2 and GAD-7 screening tools. Results: 51% of participants in the PHQ-9 surveys indicated mild to severe depression. Participants spent on average 75% of their time at home during COVID. Time spent at home had a positive correlation with the mental health of urban students but a negative correlation with suburban students. The relation between the time at home with mental health was stronger among female participants than among male participants. Correlations between female depression, anxiety, and time at home were significant. Conclusions: Lockdown and distance learning contributed to the high levels of depression in university students. This research highlights the importance of time spent at home for mental health being during the pandemic and the importance of distinguishing between urban and suburban settings when formulating public health recommendations. Quality of time spent at home versus time spent outside differentiated the mental well-being of students located in different environments. Staying at home may be recommended for students without access to safe outdoor places as it is associated with lower levels of depression.
While recommender systems have significantly benefited from implicit feedback, they have often missed the nuances of multi-behavior interactions between users and items. Historically, these systems either amalgamated all behaviors, such as \textit{impression} (formerly \textit{view}), \textit{add-to-cart}, and \textit{buy}, under a singular 'interaction' label, or prioritized only the target behavior, often the \textit{buy} action, discarding valuable auxiliary signals. Although recent advancements tried addressing this simplification, they primarily gravitated towards optimizing the target behavior alone, battling with data scarcity. Additionally, they tended to bypass the nuanced hierarchy intrinsic to behaviors. To bridge these gaps, we introduce the \textbf{H}ierarchical \textbf{M}ulti-behavior \textbf{G}raph Attention \textbf{N}etwork (HMGN). This pioneering framework leverages attention mechanisms to discern information from both inter and intra-behaviors while employing a multi-task Hierarchical Bayesian Personalized Ranking (HBPR) for optimization. Recognizing the need for scalability, our approach integrates a specialized multi-behavior sub-graph sampling technique. Moreover, the adaptability of HMGN allows for the seamless inclusion of knowledge metadata and time-series data. Empirical results attest to our model's prowess, registering a notable performance boost of up to 64\% in NDCG@100 metrics over conventional graph neural network methods.
Similar product recommendation is one of the most common scenes in e-commerce. Many recommendation algorithms such as item-to-item Collaborative Filtering are working on measuring item similarities. In this paper, we introduce our real-time personalized algorithm to model product similarity and real-time user interests. We also introduce several other baseline algorithms including an image-similarity-based method, item-to-item collaborative filtering, and item2vec, and compare them on our large-scale real-world e-commerce dataset. The algorithms which achieve good offline results are also tested on the online e-commerce website. Our personalized method achieves a 10% improvement on the add-cart number in the real-world e-commerce scenario.
Web-based interactions can be frequently represented by an attributed graph, and node clustering in such graphs has received much attention lately. Multiple efforts have successfully applied Graph Convolutional Networks (GCN), though with some limits on accuracy as GCNs have been shown to suffer from over-smoothing issues. Though other methods (particularly those based on Laplacian Smoothing) have reported better accuracy, a fundamental limitation of all the work is a lack of scalability. This paper addresses this open problem by relating the Laplacian smoothing to the Generalized PageRank and applying a random-walk based algorithm as a scalable graph filter. This forms the basis for our scalable deep clustering algorithm, RwSL, where through a self-supervised mini-batch training mechanism, we simultaneously optimize a deep neural network for sample-cluster assignment distribution and an autoencoder for a clustering-oriented embedding. Using 6 real-world datasets and 6 clustering metrics, we show that RwSL achieved improved results over several recent baselines. Most notably, we show that RwSL, unlike all other deep clustering frameworks, can continue to scale beyond graphs with more than one million nodes, i.e., handle web-scale. We also demonstrate how RwSL could perform node clustering on a graph with 1.8 billion edges using only a single GPU.