[论文解读] Penetrating a Social Network: The Follow-back Problem
本文提出了“回粉问题”——一种通过优化关注者获取策略来最大化代理在社交网络中影响力的策略。该研究基于目标度数和相互网络重叠提出了一种概率模型,为任意图结构制定了整数规划模型,并通过在Twitter数据上的模拟实验表明,该方法相比基线策略显著提升了预期关注数。
Modern threats have emerged from the prevalence of social networks. Hostile actors, such as extremist groups or foreign governments, utilize these networks to run propaganda campaigns with different aims. For extremists, these campaigns are designed for recruiting new members or inciting violence. For foreign governments, the aim may be to create instability in rival nations. Proper social network counter-measures are needed to combat these threats. Here we present one important counter-measure: penetrating social networks. This means making target users connect with or follow agents deployed in the social network. Once such connections are established with the targets, the agents can influence them by sharing content which counters the influence campaign. In this work we study how to penetrate a social network, which we call the follow-back problem. The goal here is to find a policy that maximizes the number of targets that follow the agent. We conduct an empirical study to understand what behavioral and network features affect the probability of a target following an agent. We find that the degree of the target and the size of the mutual neighborhood of the agent and target in the network affect this probability. Based on our empirical findings, we then propose a model for targets following an agent. Using this model, we solve the follow-back problem exactly on directed acyclic graphs and derive a closed form expression for the expected number of follows an agent receives under the optimal policy. We then formulate the follow-back problem on an arbitrary graph as an integer program. To evaluate our integer program based policies, we conduct simulations on real social network topologies in Twitter. We find that our polices result in more effective network penetration, with significant increases in the expected number of targets that follow the agent.
研究动机与目标
- 开发一种通过最大化目标对象关注部署代理的数量来渗透社交网络的战略方法。
- 识别影响目标对象关注代理可能性的行为与网络特征。
- 基于网络拓扑结构和代理与目标之间的重叠,建立目标对象关注代理的概率模型。
- 将最优渗透策略形式化为适用于任意图结构的整数规划问题。
- 通过在真实Twitter网络拓扑结构上的模拟,评估并验证所提出的策略。
提出的方法
- 通过实证分析发现,目标对象的度数以及代理与目标之间的共同邻域大小显著影响关注概率。
- 提出一种概率模型,其中关注概率取决于代理与目标在网络中的重叠程度。
- 在有向无环图上,通过递归期望分解推导出最优策略下预期关注数的闭式表达式。
- 使用系数 $ a^k_i $、函数 $ f(k) $ 和 $ g(F_v) $ 递归建模关注概率,其中 $ f(k) $ 表示在重叠程度为 $ k $ 时的基准关注概率。
- 将一般图上的回粉问题转化为整数规划模型,以寻找最优代理部署方案。
- 通过在真实Twitter拓扑结构上的模拟,评估策略性能并与基线策略进行对比。
实验结果
研究问题
- RQ1哪些网络与行为特征最能预测社交网络中目标对象是否会关注代理?
- RQ2在代理部署受限的条件下,如何最大化预期关注数?
- RQ3在有向无环图中,为最大化关注获取,代理部署的最优策略是什么?
- RQ4如何高效地将回粉问题在任意图结构上进行建模与求解?
- RQ5在真实社交网络数据上,所提出的策略与基线策略相比,在预期关注获取方面表现如何?
主要发现
- 目标对象关注代理的概率随着目标对象的度数以及代理与目标之间共同邻域大小的增加而提高。
- 在有向无环图上,最优策略可导出预期关注数的闭式表达式,该表达式通过递归期望分解推导得出。
- 整数规划模型能有效捕捉任意图结构上的最优代理部署策略,支持可扩展的部署。
- 在真实Twitter拓扑结构上的模拟结果表明,所提出的策略相比基线策略显著提升了预期关注数。
- 通过实证分析验证了模型的预测准确性,确认共同网络重叠程度与目标对象的度数是决定关注行为的关键因素。
更好的研究,从现在开始
从阅读论文到最终审阅,大幅缩短您的研究时间。
无需绑定信用卡
本解读由 AI 生成,并经人工编辑审核。