ID 原文 译文
693 为降低计算复杂度来实现网络拓扑的捷变控制，提出了基于连通矩阵弱摄动的动态网络拓扑控制策略。 To reduce the computationalcomplexity and achieve agile control of network topology, a dynamic network topology control strategy based on weak per-turbation of connectivity matrix is proposed.
694 研究结果表明，提出的算法能够通过内点法，可高效地得到次优解，且次优解与全局最优解十分接近。 The results show that the proposed algorithm can obtain suboptimal solutionsvery efficiently through the interior point method, and the suboptimal solution is very close to the global optimal solution.
695 网络欺凌在社交媒体平台的日益泛滥引起了研究者的广泛关注，社会科学和计算机科学研究者从不同的角度对该问题进行了研究与探讨。 Cyberbullying has attracted the increasing attention among researchers. Social and computer science re-searchers have explored cyberbullying from various perspectives.
696 为梳理这些研究，本论文对社会科学领域和计算机领域在网络欺凌方面的研究进行了调查分析。 This paper surveys the existing work on cyberbullying de-tection in social and computer science domains.
697 首先概述了网络欺凌的基本研究内容和网络欺凌特征，重点讨论了各种用于网络欺凌检测的机器学习方法，包括基于监督学习、基于弱监督学习、基于预设规则和深度学习算法， It first introduces the basic research problems and characteristics of cyberbul-lying;second, it discusses a variety of machine learning algorithms for cyberbullying detection, including supervised learn-ing, weakly supervised learning, rule-based and deep learning algorithms;
698 随后总结了 12 个现有的网络欺凌检测数据集和常用的检测性能评价指标， and third, it summarizes 12 existing datasets used incyberbullying detection and the popular metrics for detection performance.
699 最后对基于异构信息网络、融合多种辅助信息和结合心理学特征的欺凌检测方法等进行了展望。 Finally, the paper analyzes the potential researchfrom several aspects, such as cyberbullying detection approaches based on heterogeneous information network, auxiliary in-formation fusion, and psychological characteristics.
700 软件定义网络(Software-Defined Networking，SDN)的“三层两接口”架构使攻击者可以通过分析数据包往返时延分布规律推测网络类型、控制器类型及关键流规则等指纹信息。 The“three-layer two-interface”architecture of software-defined networking (SDN)enables attackers to infer fingerprint information such as network type, controller type, and key flow rules by analyzing the round-trip time distri-bution of packets.
701 目前 SDN 指纹攻击及其防御研究基本处于空白状态，为此，本文系统构建了全要素 SDN 指纹攻击链， Currently SDN fingerprint attack and its defense research are not mature, so this paper constructs a full-factor SDN fingerprint attack chain.
702 并在双重时间维度分别设计了概率加扰和控制器混淆调度防御机制，通过渐变概率加扰与最优混淆调度协同提升 SDN 指纹信息隐藏度。 Then, the probabilistic scrambling mechanism and controller dynamic confusion schedu-ling mechanism are designed in the dual time dimension respectively. More specifically, the gradient probabilistic scramblingand optimal confusion scheduling synergistically promote the information hiding degree of SDN fingerprint.