Smart contract is a trusted service provided on the blockchain, while cloud service is a traditional service mode with a large number of resources. The combination of the blockchain and cloud service is of great significance to the trusted expansion of services and the access to services inside and outside the blockchain. In this paper, we study the smart contract extension service in blockchain-cloud collaborative computing. The service module decoupling method of smart contract is proposed, and the parallel execution algorithm of smart contract service is designed, which improves the execution efficiency of smart contract service. Finally, this paper designs a secure data interaction method of smart contract and cloud service, which helps to maintain the data consistency between cloud computing and blockchain. The experimental results show that the proposed method can save at most 42.13% of the running time, and it can promote the data consistency between the cloud service and the blockchain.

City express delivery services (a.k.a.~last-mile delivery) have become more prominent in recent years. Many logistics giants, such as Amazon, JD, and Cainiao, have deployed intelligent express delivery systems to deal with the growing demand for parcel delivery. Existing works adopt queuing or batch processing approaches to assign parcels to couriers. However, these approaches do not fully consider the distribution of parcels and couriers, leading to poor quality of task assignment. In this paper, we investigate a problem of delivery matching based on revenue maximization in real-time city express delivery services. Given a set of couriers and a stream of parcel collection tasks, our problem aims to assign each collection task to a suitable courier to maximize the overall revenue of the platform. The problem is shown to be NP-hard. To tackle the problem efficiently, we present a time-aware batch matching algorithm to offer high-quality courier-task matching in each sliding window. We further theoretically analyze the matching approximation bound. In addition, we propose an efficient deep reinforcement learning-based approach to adaptively determine the sliding window size for better matching results. Finally, extensive experiments demonstrate that our proposed algorithms can achieve desirable effectiveness and efficiency under a wide range of parameter settings.

As a convenient, economical and low carbon transport mode, ride hitching has gained tremendously prevalent in the globe. This paper presents PRID that offers an efficient ride hitching service with sufficient results analysis. In general, PRID is implemented by a publish/subscribe service, which provides a user-friendly web browser client that allows users (e.g., company managers and data analysts) to perform the performance analysis for ride hitching through simple interactive operations. Moreover, the server side of the PRID system consists of three modules including a top-k initialization module, an order dispatch module, and a top-k maintenance module, which can process the matching between drivers and riders efficiently. PRID system aims at continuously maintaining the update-to-date k best matched ride orders for each driver with less time cost, and monitoring their matching efficiency and effectiveness meanwhile. Finally, we demonstrate the PRID system on a real-world dataset, showing its good performance. The source code of the demonstration has been shared on https://github.com/wuqingshun/PRID.

The wide adoption of GPS-enabled smart devices has greatly promoted spatial crowdsourcing, where the core issue is how to assign tasks to workers efficiently and with high quality. In this paper, we build a novel visual analysis system for spatial crowdsourcing, namely AMRAS, which can not only intuitively present the task allocation for workers under different time window scales to users (e.g., data analysts and managers) in real-time, but also help users analyze task assignment decision model and its learning process. AMRAS has the following novel features. First, AMRAS provides two user-friendly interfaces that allow users to employ simple and easy-to-use console to perform statistical analysis. Secondly, AMRAS provides three powerful visualization tools, such as the visualization of assignment results, assignment process, and assignment decision model, which not only allow users to intuitively analyze the whole process of task assignment, but also help users discover the computational bottleneck of their task assignment solution. Finally, AMRAS enables online access to real-time data, providing users with instant assignment and instant analysis. We have implemented and deployed AMRAS on Alibaba Cloud and demonstrated its usability and efficiency in real-world datasets. The demonstration video of AMRAS has been uploaded to Google Drive.