异构符号网络中正负社交关系的分类预测研究
|
摘要
符号社会网络正负关系分类是社会网络分析与挖掘领域的重要研究分支,在朋友关系预测,广告推荐和社团发现等方向具有重要的理论和应用价值。但是现有的分类模型所提取的特征均基于单一的节点属性和同质的链接结构,且依赖于同构网络,具有较大的局限性。针对以上问题,提出了一种新颖的基于异构网络特征的关系分类模型,特征提取主要通过引入隐朴素贝叶斯模型度量相邻异构关系的影响和结合社会化平衡理论形成的三角关系构建基于链接获得,并采用SVM等三类经典的有监督模型进行分类,验证特征的有效性。对2个大规模符号社会网络的实验表明,本文提出的模型在Precision,Recall,F1-Measure等指标均有较优的分类效果,同时也为异构社会网络关系的特征发现提供一种新的思路。
Positive and negative social relation classification is an important research branch of signed social network analy-sis and mining areas, which having important theoretical and practical value at the area of friend's relationship prediction,advertising recommendations and community detection. However, the existing models rely on the feature which basing on asingle node attributes and homogeneous link structure and depending on the homogeneous network, with great limitations.To solve the above problems, a relationship classification model based on heterogeneous network is proposed; it measuresthe impact of neighboring relations through Hidden Na?ve Bayesian model and constructed the balanced and unbalanced tri-angular relationship by integrating social theory, then uses three categories of classical supervision model such as SVM forclassification. Experimental results show that the proposed model in Precision, Recall, F1-Measure have optimum effect,and it also provides a new way of thinking for social relation feature extracting.
Positive and negative social relation classification is an important research branch of signed social network analy-sis and mining areas, which having important theoretical and practical value at the area of friend's relationship prediction,advertising recommendations and community detection. However, the existing models rely on the feature which basing on asingle node attributes and homogeneous link structure and depending on the homogeneous network, with great limitations.To solve the above problems, a relationship classification model based on heterogeneous network is proposed; it measuresthe impact of neighboring relations through Hidden Na?ve Bayesian model and constructed the balanced and unbalanced tri-angular relationship by integrating social theory, then uses three categories of classical supervision model such as SVM forclassification. Experimental results show that the proposed model in Precision, Recall, F1-Measure have optimum effect,and it also provides a new way of thinking for social relation feature extracting.
引文
1 Kaplan A M,Haenlein M.Users of the world,unite!The challenges and opportunities of Social Media[J].Business horizons,2010,53(1):59-68.
2 刘建国,周涛,汪秉宏.个性化推荐系统的研究进展[J].自然科学进展,2009,19(1):1-15.
3 Han J,Faloutsos C.Link Mining:Models,Algorithms,and Applications[M].Verlin:Springer,2010:87.
4 黄创光,印鉴,汪静.不确定近邻的协同过滤推荐算法[J].计算机学报,2010,33(8):1369-1377.
5 赵亚楠,董晶,董佳梁.基于社会化标注的博客标签推荐方法[J].计算机工程与设计,2012,33(12):4609-4613.
6 林友芳,王天宇,唐锐.一种有效的社会网络社区发现模型和算法[J].计算机研究与发展,2012,49(2):337-345.
7 Al Hasan M,Chaoji V,Salem S,et al.Link prediction using supervised learning[C]//SDM’06:Workshop on Link Analysis,Counter-terrorism and Security,2006.
8 Popescul,Alexandrin,and Lyle H.Ungar.Statistical relational learning for link prediction[C]//IJCAI workshop on learning statistical models from relational data,2003.
9 Sarukkai R R.Link prediction and path analysis using Markov chains[J].Computer Networks,2000,33(1):377-386.
10 Zhu J.Max-margin nonparametric latent feature models for link prediction[J].ar Xiv preprint ar Xiv:1206.4659,2012.
11 Liben‐Nowell D,Kleinberg J.The link‐prediction problem for social networks[J].Journal of the American society for information science and technology,2007,58(7):1019-1031.
12 Popescul A,Ungar L H.Statistical relational learning for link prediction[C]//IJCAI workshop on learning statistical models from relational data,2003.
13 Sun Y,Barber R,Gupta M,et al.Co-author relationship prediction in heterogeneous bibliographic networks[C]//ASONAM,2011 International Conference on.IEEE,2011:121-128.
14 Liu Z,Zhang Q M,LüL,et al.Link prediction in complex networks:A local na?ve Bayes model[J].EPL(Europhysics Letters),2011,96(4):48007.
15 Chiang,K.Y.,Natarajan,N.,Tewari,A.,&Dhillon,I.S..Exploiting longer cycles for link prediction in signed networks[C]//Proceedings of the 20th ACM CIKM,ACM,2011.
2 刘建国,周涛,汪秉宏.个性化推荐系统的研究进展[J].自然科学进展,2009,19(1):1-15.
3 Han J,Faloutsos C.Link Mining:Models,Algorithms,and Applications[M].Verlin:Springer,2010:87.
4 黄创光,印鉴,汪静.不确定近邻的协同过滤推荐算法[J].计算机学报,2010,33(8):1369-1377.
5 赵亚楠,董晶,董佳梁.基于社会化标注的博客标签推荐方法[J].计算机工程与设计,2012,33(12):4609-4613.
6 林友芳,王天宇,唐锐.一种有效的社会网络社区发现模型和算法[J].计算机研究与发展,2012,49(2):337-345.
7 Al Hasan M,Chaoji V,Salem S,et al.Link prediction using supervised learning[C]//SDM’06:Workshop on Link Analysis,Counter-terrorism and Security,2006.
8 Popescul,Alexandrin,and Lyle H.Ungar.Statistical relational learning for link prediction[C]//IJCAI workshop on learning statistical models from relational data,2003.
9 Sarukkai R R.Link prediction and path analysis using Markov chains[J].Computer Networks,2000,33(1):377-386.
10 Zhu J.Max-margin nonparametric latent feature models for link prediction[J].ar Xiv preprint ar Xiv:1206.4659,2012.
11 Liben‐Nowell D,Kleinberg J.The link‐prediction problem for social networks[J].Journal of the American society for information science and technology,2007,58(7):1019-1031.
12 Popescul A,Ungar L H.Statistical relational learning for link prediction[C]//IJCAI workshop on learning statistical models from relational data,2003.
13 Sun Y,Barber R,Gupta M,et al.Co-author relationship prediction in heterogeneous bibliographic networks[C]//ASONAM,2011 International Conference on.IEEE,2011:121-128.
14 Liu Z,Zhang Q M,LüL,et al.Link prediction in complex networks:A local na?ve Bayes model[J].EPL(Europhysics Letters),2011,96(4):48007.
15 Chiang,K.Y.,Natarajan,N.,Tewari,A.,&Dhillon,I.S..Exploiting longer cycles for link prediction in signed networks[C]//Proceedings of the 20th ACM CIKM,ACM,2011.