群智感知中采用节点社会属性的亲密度量化方法

发布时间：2018-03-13 11:29

本文选题：群智感知　切入点：社会关系量化　出处：《西安交通大学学报》2017年06期 　论文类型：期刊论文

【摘要】：针对群智感知中节点社会关系计算存在的层次关系划分不清、关联因子描述粗糙等问题,提出一种采用社会属性的亲密度量化方法。该方法通过分析影响节点社会关系的多维要素,将节点属性因子划分为静态和动态两个维度。通过构造多维语义分级树和空间索引编码,对节点静态属性进行挖掘和形式化表示。同时,引入交互信息熵,对社会关系的不对称性进行分析和比较,以提高亲密度量化方法的客观性。基于层次分析法实现节点动态属性的二级评判和有效聚合,并通过节点融合度对亲密关系进行二次修正。实验结果表明:与已有方法相比,采用社会属性的亲密度量化方法预测准确率提高了14.67%,该方法能够有效降低群智感知中移动节点的误判概率,提高网络社团识别准确率,为候选服务节点集的选择提供有效依据。
[Abstract]:In order to solve the problems such as unclear classification of hierarchical relations and rough description of correlation factors in the computation of social relations between nodes in the perception of group intelligence, and so on, In this paper, a quantitative method of affinity using social attributes is proposed, which is based on the analysis of the multi-dimensional factors that affect the social relations of nodes. The node attribute factor is divided into static and dynamic dimensions. By constructing multi-dimensional semantic hierarchical tree and spatial index coding, the static attributes of nodes are mined and formally represented. At the same time, interactive information entropy is introduced. The asymmetry of social relations is analyzed and compared in order to improve the objectivity of the method of affinity quantification. Based on the analytic hierarchy process (AHP), the two-level evaluation and effective aggregation of node dynamic attributes are realized. The experiment results show that: compared with the existing methods, The prediction accuracy rate of social attribute based affinity quantization method is improved by 14.67. this method can effectively reduce the misjudgment probability of mobile nodes in group intelligence perception, improve the accuracy of network community identification, and provide an effective basis for the selection of candidate service node sets.
【作者单位】：西安交通大学电子与信息工程学院;新疆大学信息科学与工程学院;西安交通大学陕西省计算机网络重点实验室;
【基金】：国家自然科学基金资助项目(61502380) 陕西省自然科学基金资助项目(2014JQ8322) 陕西省科技统筹创新工程资助项目(2013SZS16-Z01/P01/K01) 中央高校基本科研业务费专项资金资助项目(XJJ2014049)
【分类号】：C912.3

【相似文献】