基于最大可达路径的异构网络关系预测方法
发布时间：2018-11-09 08:36
【摘要】：社会空间中人与人之间通过不同类型关系连接形成异构关系网络,针对该异构网络的结构特性与异构信息,综合个体属性、关系语义和关系结构3个重要方面,构建一种最大可达路径度量方法,并将此方法运用于预测异构网络中节点间的不同社会关系。在相关实验数据集上的实验结果表明了该算法的有效性。使用能够反映节点属性和节点之间语义关系的图数据库Neo4j存储人物和人物间关系,与使用关系型数据库Mysql的实验运行时间进行比较,比较结果表明了Neo4j运行时间仅仅为Mysql时间的一半。
[Abstract]:In the social space, the heterogeneous relationship network is formed through the connection of different types of relationships between people. Aiming at the structural characteristics and heterogeneous information of the heterogeneous network, it synthesizes individual attributes, relational semantics and relational structure. A maximum reachable path measurement method is constructed and applied to predict the different social relations between nodes in heterogeneous networks. Experimental results on related experimental data sets show the effectiveness of the proposed algorithm. The graph database Neo4j, which can reflect the semantic relationship between nodes and the attributes of nodes, is used to store the relationships between people and people, which is compared with the experimental running time of Mysql, a relational database. The comparison results show that the running time of Neo4j is only half of that of Mysql.
【作者单位】： 太原理工大学计算机科学与技术学院;
【基金】：国家863高技术研究发展计划基金项目(2014AA015204) 山西省自然科学基金项目(2014011022-1)
【分类号】：O157.5;TP301.6
本文编号：2319902
[Abstract]:In the social space, the heterogeneous relationship network is formed through the connection of different types of relationships between people. Aiming at the structural characteristics and heterogeneous information of the heterogeneous network, it synthesizes individual attributes, relational semantics and relational structure. A maximum reachable path measurement method is constructed and applied to predict the different social relations between nodes in heterogeneous networks. Experimental results on related experimental data sets show the effectiveness of the proposed algorithm. The graph database Neo4j, which can reflect the semantic relationship between nodes and the attributes of nodes, is used to store the relationships between people and people, which is compared with the experimental running time of Mysql, a relational database. The comparison results show that the running time of Neo4j is only half of that of Mysql.
【作者单位】： 太原理工大学计算机科学与技术学院;
【基金】：国家863高技术研究发展计划基金项目(2014AA015204) 山西省自然科学基金项目(2014011022-1)
【分类号】：O157.5;TP301.6
【相似文献】
相关期刊论文 前1条
1 周爱平;程光;;博弈论在异构的多接入网络选择策略应用[J];中国科技论文;2012年07期
相关硕士学位论文 前5条
1 常浩;匹配博弈理论在异构无线网络中的应用[D];燕山大学;2016年
2 刘松林;基于博弈论和凸优化的异构网络资源分配方法研究[D];哈尔滨工业大学;2016年
3 祝明睿;多维异构网络上的边和社团的预测与演化的研究[D];哈尔滨工业大学;2014年
4 曾玉洁;符号网络和异构网络中的社区检测方法[D];西安电子科技大学;2014年
5 邓媛;基于博弈论的异构网络功率控制技术研究[D];华南理工大学;2011年
，本文编号：2319902
本文链接：https://www.wllwen.com/kejilunwen/yysx/2319902.html
