ID	原文	译文
3643	通过卷积神经网络获得多变量的时间模式信息作为模型的注意力机制。	which was the multivariate time pattern information obtained by convolutional neural networks (CNN).
3644	通过对模型训练、学习特征信息，并对相关性高的序列分配较大的学习权重，来实现解码器输出高度相关的有用特征预测目标序列。	The characteristic information was learnedby the model training, and the sequence with higher correlation was assigned a larger learning weight, so that the highlycorrelated features output from the decoder could be used to predict the target sequence.
3645	应用多个停车场数据集对模型进行测试，测试结果及对比分析表明，所提模型在步长达到 36 时对泊位占有率的预测数据能较好地估计真实值，预测精度和稳定性相比 LSTM 均有提高。	Data sets of multiple parking lotwere adopted to test the model. The test results show that the proposed model can estimate the real value well when thestep length of berth occupancy prediction reaches 36. The prediction accuracy and stability of the model are improvedcompared with long short-term memory (LSTM) model.
3646	针对当前用于提取突发事件的方法存在精度低和效率低的问题，提出一种基于词相关性特征的突发事件检测方法，能从社会网络数据流中快速地检测出突发事件，以便相关的决策者可以及时有效地采取措施进行处理，使突发事件的负面影响被尽量降低，维护社会的安定。	For current methods for extracting emergencies had problems of low accuracy and low efficiency, an emer-gency detection method based on the characteristics of word correlation was proposed, which could quickly detect emer-gency events from the social network data stream, so that relevant decision makers could take timely and effective meas-ures to deal with, making the negative impact of emergencies can be reduced as much as possible to maintain social sta-bility.
3647	仿真结果表明，突发事件检测方法在实时博文数据流中具有很好的事件检测效果，	The simulation results show that the emergency event detection method has a better event detection effect in thereal-time blog post data stream.
3648	与已有的方法相比，所提方法可以满足突发事件检测的需求，不仅能检测到子事件的详细信息，而且能准确地检测出事件的相关信息。	Compared with the existing methods, the proposed method can meet the needs of emer-gency detection. Not only the detailed information of the sub-events can be detected, but also the related information of the events can be accurately detected.
3649	为同时提升车辆的计算迁移率和边缘计算服务器的资源利用率，提出了一种基于交通流量预测的车联网边缘计算迁移方案。	With an aim of maximizing the efficiency of edge offloading and the resource utilization of edge computingserver simultaneously, a new flow-of-traffic prediction based edge computing offloading solution was proposed for In-ternet of vehicles (IoV).
3650	首先通过设计考虑任务优先级的计算迁移率和资源利用率效用函数，将问题建模为双目标优化问题；	Firstly, both the efficiency utility function of vehicle and the resource utilization of mobile edgecomputing (MEC) server were established by considering task priority.
3651	然后将问题转化为车辆与边缘计算服务器之间的资源双边拍卖问题，再通过设计交通流量信息感知的报价函数，采用 McAfee 拍卖算法，进而完成边缘计算迁移。	Next, the formulated dual-objective optimizationproblem was transformed into a double auction problem between vehicles and MEC servers. Finally, based on the de-signed flow-of-traffic based pricing function of vehicle and MEC server, a McAfee auction algorithm was adopted tocomplete the edge computing process.
3652	仿真结果表明，通过交通流量预测信息的辅助，所提方案能有效提高车联网的边缘计算迁移率和资源利用率。	Simulation results show that benefiting from the flow-of-traffic prediction infor-mation, the proposed solution can significantly improve both the efficiency of computation offloading and the utilizationof computation resource.