Recent years have witnessed a revolution in Spatial Crowdsourcing (SC), in which people with mobile connectivity can perform spatio-temporal tasks that involve traveling to specified locations. In this paper, we identify and study in depth a new multi-center-based task allocation problem in the context of SC, where multiple allocation centers exist. In particular, we aim to maximize the total number of the allocated tasks while minimizing the allocated task number difference. To solve the problem, we propose a two-phase framework, called Task Allocation with Geographic Partition, consisting of a geographic partition and a task allocation phase. The first phase divides the whole study area based on the allocation centers by using both a basic Voronoi diagram-based algorithm and an adaptive weighted Voronoi diagram-based algorithm. In the allocation phase, we utilize a Reinforcement Learning method to achieve the task allocation, where a graph neural network with the attention mechanism is used to learn the embeddings of allocation centers, delivery points, and workers. To further improve the efficiency, we propose an early stopping optimization strategy for the adaptive weighted Voronoi diagram-based algorithm in the geographic partition phase and give a distance-constrained graph pruning strategy for the Reinforcement Learning method in the task allocation phase. Extensive experiments give insight into the effectiveness and efficiency of the proposed solutions.

Join order selection plays an important role in DBMS query optimizers. The problem aims to find the optimal join order with the minimum cost, and usually becomes an NP-hard problem due to the exponentially increasing search space. Recent advanced studies attempt to use deep reinforcement learning (DRL) to generate better join plans than the ones provided by conventional query optimizers. However, DRL-based methods require time-consuming training, which is not suitable for online applications that need frequent periodic re-training. In this paper, we propose a novel framework, namely efficient Join Order selection learninG with Graph-basEd Representation (JOGGER). We firstly construct a schema graph based on the primary-foreign key relationships, from which table representations are well learned to capture the correlations between tables. The second component is the state representation, where a graph convolutional network is utilized to encode the query graph and a tailored-tree-based attention module is designed to encode the join plan. To speed up the convergence of DRL training process, we exploit the idea of curriculum learning, in which queries are incrementally added into the training set according to the level of difficulties. We conduct extensive experiments on JOB and TPC-H datasets, which demonstrate the effectiveness and efficiency of the proposed solutions.

The ongoing COVID-19 pandemic has dramatically changed people's daily lives. A robust forecasting model for COVID-19 infections is essential for governments and institutions to plan timely and perform accurate interventions. Mainstream solutions for COVID-19 prediction fit reported data only by considering observed cases. However, the neglected facts that positive samples are incomplete and many facts of the novel disease are unknown may be prone to cause severe error accumulation, especially in long-term predictions. To fully understand the spreading patterns of the virus, we propose an encoder-decoder framework: (i) in the encoder we embed historical case data into multiple expose-infection ranges and learn message passing between time slices and across ranges with coarse-grained human mobility data incorporated; (ii) in the decoder, we decode the embedded features based on reported cases as well as deaths to jointly consider the effect of both observed and hidden data. We model the spreading of disease in over 60 counties of California and New York, which are two of the most metropolitan areas in the US. The proposed framework significantly outperforms state-of-the-art baselines on JHU COVID-19 dataset on both weekly prediction and daily prediction tasks. We design detailed ablation studies to verify the effectiveness of each key module and find the model not only works with the assistance of mobility data but also with purely cases and deaths, which implies its broad application scenarios.

Recent years have witnessed a revolution in Spatial Crowdsourcing (SC), in which people with mobile connectivity can perform spatio-temporal tasks that involve travel to specified locations. In this paper, we identify and study in depth a new multi-center-based task allocation problem in the context of SC, where multiple allocation centers exist. In particular, we aim to maximize the total number of the allocated tasks while minimizing the average allocated task number difference. To solve the problem, we propose a two-phase framework, called Task Allocation with Geographic Partition, consisting of a geographic partition phase and a task allocation phase. The first phase is to divide the whole study area based on the allocation centers by using both a basic Voronoi diagram-based algorithm and an adaptive weighted Voronoi diagram-based algorithm. In the allocation phase, we utilize a Reinforcement Learning method to achieve the task allocation, where a graph neural network with the attention mechanism is used to learn the embeddings of allocation centers, delivery points and workers. Extensive experiments give insight into the effectiveness and efficiency of the proposed solutions.