Group‐in‐a‐Box Meta‐Layouts for Topological Clusters and Attribute‐Based Groups: Space‐Efficient Visualizations of Network Communities and Their Ties

An important part of network analysis is understanding community structures like topological clusters and attribute‐based groups. Standard approaches for showing communities using colour, shape, rectangular bounding boxes, convex hulls or force‐directed layout algorithms remain valuable, however our Group‐in‐a‐Box meta‐layouts add a fresh strategy for presenting community membership, internal structure and inter‐cluster relationships. This paper extends the basic Group‐in‐a‐Box meta‐layout, which uses a Treemap substrate of rectangular regions whose size is proportional to community size. When there are numerous inter‐community relationships, the proposed extensions help users view them more clearly: (1) the Croissant–Doughnut meta‐layout applies empirically determined rules for box arrangement to improve space utilization while still showing inter‐community relationships, and (2) the Force‐Directed layout arranges community boxes based on their aggregate ties at the cost of additional space. Our free and open source reference implementation in NodeXL includes heuristics to choose what we have found to be the preferable Group‐in‐a‐Box meta‐layout to show networks with varying numbers or sizes of communities. Case study examples, a pilot comparative user preference study (nine participants), and a readability measure‐based evaluation of 309 Twitter networks demonstrate the utility of the proposed meta‐layouts.     

一种基于Normal矩阵的时间序列聚类方法*

提出了一种基于Normal矩阵的时间序列聚类方法。该算法首先对时间序列数据进行向量形式转换,计算出各个时间序列间的相似度并构建复杂网络,然后利用基于Normal矩阵的方法进行复杂网络社团划分,同一类的时间序列被划分到一个社团,即实现对时间序列数据的聚类。为了验证该方法的可行性和有效性,将其应用于股票时间序列数据聚类分析中,并在两个实际的数据集上与其他方法相比较,取得了较好的实验结果。     

Temporal evolution of contacts and communities in networks of face-to-face human interactions

Temporal dynamics of social interaction networks as well as the analysis of communities are key aspects to gain a better understanding of the involved processes, important influence factors, their effects, and their structural implications. In this article, we analyze temporal dynamics of contacts and the evolution of communities in networks of face-to-face proximity. As our application context, we consider four scientific coil- ferences. On a structural level, we focus on static and dynamic properties of the contact graphs. Also, we analyze the resulting community structure using state-of-the-art automatic community detection algorithms. Specifically, we analyze tile evolution of contacts and communities over time to consider the stability of the respective conmmnities. Furthermore, we assess different factors which have an influence on the quality of com- munity prediction. Overall, we provide first important insights into the evolution of contacts and communities in face-to-face contact networks.     

基于迁移学习的卷积神经网络通道剪枝

卷积神经网络在计算机视觉等多个领域应用广泛,然而其模型参数量众多、计算开销庞大,导致许多边缘设备无法满足其存储与计算资源要求。针对其边缘部署困难,提出使用迁移学习策略改进基于BN层缩放因子通道剪枝方法的稀疏化过程。本文对比不同层级迁移方案对稀疏化效果与通道剪枝选取容限的影响;并基于网络结构搜索观点设计实验,探究其精度保持极限与迭代结构的收敛性。实验结果表明,对比原模型,采用迁移学习的通道剪枝算法,在精度损失不超过0.10的前提下,参数量减少89.1%,模型存储大小压缩89.3%;对比原剪枝方法,将剪枝阈值从0.85提升到0.97,进一步减少参数42.6%。实验证明,引入迁移策略更易实现充分的稀疏化,提高通道剪枝阈值选取容限,实现更高压缩率;并在迭代剪枝的网络结构搜索过程中,提供更高效的搜索起点,利于快速迭代趋近至搜索空间的一个网络结构局部最优解。     

民生视阈下提升高职院校社区服务能力的研究

服务社区是高职院校服务社会、实现自身发展的有效途径。通过对高职院校服务社区过程中存在的问题和原因分析,给出提升高职院校社区服务能力的对策及可行性建议,为今后的相关工作提供参考。     

Group identification as a mediator of the effect of players’ anonymity on cheating in online games

This study aims to add to the discussion about the applicability of the classical deindividuation theory and social identity model of deindividuation effects (SIDE) in explaining online behaviours. It explores the effect of anonymity in facilitating social influence of group identity in online game cheating. A nationally representative survey was conducted face to face. Results from the survey administered in Singapore confirm predictions derived from the SIDE and challenge the classical deindividuation theory. Specifically, it was concluded that the frequency of gaming with online strangers (anonymous gaming) significantly predicted the frequency of cheating in online games. The effect of anonymity on game cheating was found to be significantly mediated by the group identification with online gaming communities/groups. Gender differences were found. Male gamers cheated more frequently than female gamers. Female gamers are more likely to cheat as a consequence of group identification than male gamers. Implications and future research are discussed.     

基于支持向量机的炒作微博识别方法

微博是舆论传播的中心和渠道,同时参与舆论的形成、发展与引导过程,其自媒体发布、意见领袖参与等因素在一定程度上造成了微博谣言、虚假炒作、社会动员等现象。针对炒作微博的传播特点,分析其群体的隐蔽策划现象,挖掘出普通微博和炒作微博在传播网络结构、转发增量统计等方面的差异。通过社交网站的应用程序接口对目标微博的所有评论、转发和点赞用户进行信息获取,构建该微博的传播网络,利用社团模块度、平均最短路径和网络直径这3个属性度量该网络的紧密程度,基于支持向量机对所抽取的微博进行分类,进而识别出炒作微博。实验结果表明,该方法对微博传播用户的属性信息依赖小以及传播网络结构特征敏感,并且具有较高的炒作微博识别准确率。     

针对有向图的局部扩展的重叠社区发现算法

当前社区发现算法主要是针对无向图研究社区结构,但在实际复杂网络中,链接关系时常表现出非对称性或方向性,比如Twitter的用户关注关系,文献网络的引 用关系,网页之间的超链接关系等应用网络。因此,本文依据信息在复杂网络中的传播规律和流动方向性,提出了k-Path共社区邻近相似性概念及计算方法,用于衡量结点在同一社区的相似性程度,并给出了把有向图转换为带方向权值的无向图的方法。基于带权无向图提出了一种从局部扩展来探测社区的重叠社区发现算法(Local and wave-like extension algorithm of detecting overlapping community, LWS-OCD)。在真实数据集上的实验表明,共社区邻近相似性概念实现了有向到无向的合理转换,而且提高了社区结点的聚集效果,LWS-OCD算法能够有效地发现带权无向图中的重叠社区。     

社团挖掘和话题监控的互动模型研究

社团挖掘是Web信息挖掘领域的重要应用,而话题监控是文本信息研究领域的重要应用,目前这两种技术是各自独立的。为更好地应用于互联网形成的复杂社会网络,将这两种技术结合起来研究,发现了社团和话题之间的关系,创建了社团挖掘和话题监控的静态和动态互动模型,设计了社团挖掘、话题识别以及社团跟踪算法。     

Detecting Overlapping Community Structures in Networks

Community structure has been recognized as an important statistical feature of networked systems over the past decade. A lot of work has been done to discover isolated communities from a network, and the focus was on developing of algorithms with high quality and good performance. However, there is less work done on the discovery of overlapping community structure, even though it could better capture the nature of network in some real-world applications. For example, people are always provided with varying characteristics and interests, and are able to join very different communities in their social network. In this context, we present a novel overlapping community structures detecting algorithm which first finds the seed sets by the spectral partition and then extends them with a special random walks technique. At every expansion step, the modularity function Q is chosen to measure the expansion structures. The function has become one of the popular standards in community detecting and is defined in Newman and Girvan (Phys. Rev. 69:026113, 2004). We also give a theoretic analysis to the whole expansion process and prove that our algorithm gets the best community structures greedily. Extensive experiments are conducted in real-world networks with various sizes. The results show that overlapping is important to find the complete community structures and our method outperforms the C-means in quality.