ID	原文	译文
4944	其次，提出了一种对二进制函数进行单调编码的算法，基于图形嵌入原理，融合执行流程图与函数代码基本块特征，形成每个函数代码的投影特征；	Secondly, a monotonic encodingalgorithm for the App function was designed, which encoded the function based on the control flow graph structure andbasic block attributes.
4945	最后，提出了一种高效的第三方库函数过滤方法，进一步提升了克隆检测效果。	Finally, a three-party libraries filtering method was proposed by using an efficient clustering algo-rithm based on the function feature.
4946	实验证明，所提方法具有 97.6%的准确率，在实验条件下判定一个 App 克隆对只需要 79 ms，能够支撑云环境下应用市场级别的大规模克隆检测。	Experiments verified the effectiveness of the proposed scheme. The average searchtime is close to 79 ms, and the clone detection accuracy achieves 97.6%.
4947	针对复杂海洋背景中高机动目标的快速检测问题，提出了一种基于稀疏分数阶模糊函数（SFRAF）的雷达目标检测算法，所提算法在瞬时自相关函数计算后引入稀疏傅里叶变换（SFT），结合了 SFT 和分数阶模糊函数（FRAF）的优势，	To solve the problem of fast detection of high maneuvering targets in complex ocean background, a radar tar-get detection algorithm via sparse fractional ambiguity function (SFRAF) was proposed. The sparse Fourier transform(SFT) was introduced after instantaneous autocorrelation function calculation, which combines the advantages of SFTand fractional ambiguity function (FRAF).
4948	因此，SFRAF 对高机动信号具有较好的处理效果，同时能够获得较低的计算复杂度。	Therefore, SFRAF has good processing performance on high maneuveringsignals and can achieve lower computational complexity.
4949	仿真实验和实测数据处理结果表明，相比于经典检测方法，所提方法能够明显改善机动目标的检测性能，而相比于FRAF，所提方法能够实现检测效率的有效提升。	The simulation experiments and the measured data processing results show that compared with classical detection method, the detection performance of maneuvering target can be sig-nificantly improved, and compared with FRAF, the proposed method can achieve higher detection efficiency.
4950	为了解决大数据流式计算平台中存在计算负载波动上升，但集群无法有效应对负载变化的问题，提出了基于流网络的 Flink 平台弹性资源调度策略（FAR-Flink）。	In order to solve the problem that the load of big data stream computing platform is increasing with fluctuationwhile the cluster was not able to rescale efficiently, the Flow-network based auto rescale strategy for Flink was proposed.
4951	该策略首先建立流网络模型并通过构建算法计算每条边的容量值，	Firstly, the flow-network model was set up and the capacity of each edge that was calculated by self-learning algorithm.
4952	其次通过弹性资源调度算法确定集群性能瓶颈并制定动态资源调度计划，	Secondly, the bottleneck of the cluster was acquired by maximum-flow algorithm and the resource rescheduling plan wasdrawn up.
4953	最后通过基于数据分簇和分桶管理的状态数据迁移算法，实施调度计划并完成节点间的高效数据迁移。	Finally, the resource rescheduling plan was executed and the stateful data was migrated efficiently by the datamigration algorithm based on the strategy of data partitioning by bulk and bucket.