{"id":5845,"date":"2018-08-02T18:44:56","date_gmt":"2018-08-02T10:44:56","guid":{"rendered":"https:\/\/www.icnalb.cn\/?p=5845"},"modified":"2018-12-12T04:20:55","modified_gmt":"2018-12-11T20:20:55","slug":"17%e7%ba%a7%e8%a6%83%e5%ad%9f%e5%90%8c%e5%ad%a6%e4%b8%a4%e7%af%87%e8%ae%ba%e6%96%87%e5%88%86%e5%88%ab%e4%b8%ad%e7%a8%bfacm-sigcomm-2018%e3%80%81wledge-based-systems-%e6%9c%9f%e5%88%8a","status":"publish","type":"post","link":"https:\/\/www.icnlab.cn\/?p=5845","title":{"rendered":"17\u7ea7\u8983\u5b5f\u540c\u5b66\u5206\u522b\u4e2d\u7a3fSIGCOMM’18 Workshop\u548cKnowledge-Based Systems\u671f\u520a"},"content":{"rendered":"

ACM SIGCOMM 2018 Workshop on Network Meets AI & ML (NetAIM 2018)<\/strong><\/p>\n

ACM SIGCOMM 2018(https:\/\/conferences.sigcomm.org\/sigcomm\/2018\/<\/a>) (CCF A)\u5c06\u4e8e2018\u5e748\u670820\u65e5\u81f325\u65e5\u5728\u5308\u7259\u5229\u5e03\u8fbe\u4f69\u65af\u4e3e\u884c\uff0cSIGCOMM\u662f\u8ba1\u7b97\u673a\u7f51\u7edc\u901a\u4fe1\u9886\u57df\u7684\u91cd\u8981\u5b66\u672f\u4f1a\u8bae\uff0c\u4e14\u4eca\u5e74\u4e3aNetAI Workshop(https:\/\/conferences.sigcomm.org\/sigcomm\/2018\/workshop-netaim.html<\/a>)\u5728SIGCOMM\u4f1a\u8bae\u4e3e\u529e\u7684\u7b2c\u4e00\u5c4a\uff0c\u8be5workshop\u4e3b\u8981\u5173\u6ce8\u5982\u4f55\u5229\u7528\u4eba\u5de5\u667a\u80fd\u6280\u672f\u6709\u6548\u5e94\u5bf9\u73b0\u6709\u7f51\u7edc\u7cfb\u7edf\u7684\u9762\u4e34\u7684\u6311\u6218\u3002<\/p>\n

17\u7ea7\u540c\u5b66\u8983\u5b5f\uff0c\u5728\u5b9e\u9a8c\u5ba4\u96f7\u8001\u5e08\u7684\u6307\u5bfc\u4e0b\uff0c\u5b8c\u6210\u4e00\u7bc7\u957f\u6587\uff0c\u5e76\u5df2\u88abSIGCOMM 2018 NetAI Workshop\u5f55\u7528\uff0c\u8bba\u6587\u7b80\u4ecb\u5982\u4e0b\uff1a<\/p>\n

\u6807\u9898:\u00a0Adaptive Multiple Non-negative Matrix Factorization for Temporal Link Prediction in Dynamic Networks<\/strong><\/p>\n

\u4f5c\u8005: Kai Lei, Meng Qin, Bo Bai*, Gong Zhang.<\/strong><\/p>\n

\u82f1\u6587\u6458\u8981:<\/strong>\u00a0The prediction of mobility, topology and traffic is an effective technique to improve the performance of various network systems, which can be generally represented as the temporal link prediction problem. In this paper, we propose a novel adaptive multiple nonnegative matrix factorization (AM-NMF) method from the view of network embedding to cope with such problem. Under the framework of non-negative matrix factorization (NMF), the proposed method embeds the dynamic network into a low-dimensional hidden space, where the characteristics of different network snapshots are comprehensively preserved. Especially, our new method can effectively incorporate the hidden information of different time slices, because we introduce a novel adaptive parameter to automatically adjust the relative contribution of different terms in the uniform model. Accordingly, the prediction result of future network topology can be generated by conducting the inverse process of NMF form the shared hidden space. Moreover, we also derive the corresponding solving strategy whose convergence can be ensured. As an illustration, the new model will be applied to various network datasets such as human mobility networks, vehicle mobility networks, wireless mesh networks and data center networks. Experimental results show that our method outperforms state-of-the-art methods for the temporal link prediction of both unweighted and weighted networks.<\/p>\n

\u4e2d\u6587\u7b80\u4ecb\uff1a<\/strong>\u5bf9\u4e8e\u7528\u6237\u79fb\u52a8\u6027\u3001\u7f51\u7edc\u52a8\u6001\u62d3\u6251\u548c\u7f51\u7edc\u6d41\u91cf\u7684\u9884\u6d4b\u662f\u6539\u5584\u5404\u79cd\u7f51\u7edc\u7cfb\u7edf\u6027\u80fd\u7684\u6709\u6548\u624b\u6bb5\uff0c\u800c\u76f8\u5173\u7684\u7f51\u7edc\u7cfb\u7edf\u52a8\u6001\u6027\u9884\u6d4b\u95ee\u9898\u80fd\u591f\u4ee5\u590d\u6742\u7f51\u7edc\u5206\u6790\u7684\u89c2\u70b9\u4e00\u822c\u6027\u5730\u62bd\u8c61\u4e3a\u65f6\u5e8f\u94fe\u8def\u9884\u6d4b(temporal link prediction)\u95ee\u9898\u3002\u4ece\u7f51\u7edc\u8868\u5f81(network embedding)\u7684\u89c2\u70b9\u51fa\u53d1\uff0c\u63d0\u51fa\u4e00\u79cd\u81ea\u9002\u5e94\u591a\u91cd\u975e\u8d1f\u77e9\u9635\u5206\u89e3(adaptive nonnegative matrix factorization, AM-NMF)\u6a21\u578b\u89e3\u51b3\u4e0a\u8ff0\u95ee\u9898\u3002\u5728\u975e\u8d1f\u77e9\u9635\u5206\u89e3(nonnegative matrix factorization, NMF)\u6846\u67b6\u4e0b\uff0c\u8be5\u6a21\u578b\u5c06\u52a8\u6001\u7f51\u7edc\u5d4c\u5165\u5230\u4e00\u4e2a\u4fdd\u7559\u4e86\u4e0d\u540c\u7f51\u7edc\u5feb\u7167\u52a8\u6001\u53d8\u5316\u7279\u5f81\u7684\u4f4e\u7ef4\u9690\u542b\u7a7a\u95f4\u3002\u7279\u522b\u5730\uff0c\u7531\u4e8e\u5f15\u5165\u81ea\u9002\u5e94\u53c2\u6570\u81ea\u52a8\u5730\u8c03\u8282\u6df7\u5408\u6a21\u578b\u4e2d\u4e0d\u540c\u5206\u91cf\u7684\u76f8\u5bf9\u91cd\u8981\u6027\uff0c\u8be5\u6a21\u578b\u8fd8\u80fd\u6709\u6548\u5730\u7ed3\u5408\u4e0d\u540c\u65f6\u95f4\u7247\u4e0b\u7684\u9690\u542b\u4fe1\u606f\uff0c\u5e76\u8003\u8651\u5355\u4e2a\u65f6\u95f4\u7247\u4e0e\u52a8\u6001\u7f51\u7edc\u6574\u4f53\u7684\u5185\u5728\u5173\u8054\u6027\u3002\u8fdb\u4e00\u6b65\u5730\uff0c\u5173\u4e8e\u4e0b\u4e00\u4e2a\u65f6\u95f4\u7247\u7f51\u7edc\u5feb\u7167\u7684\u9884\u6d4b\u7ed3\u679c\u80fd\u591f\u901a\u8fc7\u6267\u884cNMF\u7684\u9006\u8fc7\u7a0b\u751f\u6210\u3002\u4f5c\u4e3a\u4e00\u4e2a\u5e94\u7528\u793a\u4f8b\uff0c\u8be5\u6a21\u578b\u4e5f\u88ab\u5e94\u7528\u4e8e\u5404\u79cd\u7f51\u7edc\u7cfb\u7edf\u76f8\u5173\u7684\u6570\u636e\u96c6\uff0c\u5305\u62ec\u4eba\u79fb\u52a8\u7f51\u7edc\u3001\u8f66\u8f86\u79fb\u52a8\u7f51\u7edc\u3001\u65e0\u7ebf\u7f51\u683c\u7f51\u7edc\u548c\u6570\u636e\u4e2d\u5fc3\u7f51\u7edc\u3002\u76f8\u5173\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u8be5\u65b9\u6cd5\u5728\u65e0\u6743\u7f51\u7edc\u548c\u5e26\u6743\u7f51\u7edc\u7684\u65f6\u5e8f\u94fe\u8def\u9884\u6d4b\u4efb\u52a1\u4e0a\u7684\u6027\u80fd\u8d85\u8fc7\u73b0\u6709\u7684\u65b9\u6cd5\u3002<\/p>\n

 <\/p>\n

Knowledge-Based Systems \u671f\u520a<\/strong><\/p>\n

Knowledge-Based Systems(https:\/\/www.journals.elsevier.com\/knowledge-based-systems\/<\/a>) \u662f\u4eba\u5de5\u667a\u80fd\u9886\u57df\u8de8\u5b66\u79d1\u3001\u9762\u5411\u5e94\u7528\u7684\u5b66\u672f\u671f\u520a\uff0c\u6700\u65b0\u7684\u5f71\u54cd\u56e0\u5b50(IF)\u4e3a4.396\u3002<\/p>\n

17\u7ea7\u540c\u5b66\u8983\u5b5f\uff0c\u5728\u5b9e\u9a8c\u5ba4\u96f7\u8001\u5e08\u7684\u6307\u5bfc\u4e0b\uff0c\u5b8c\u6210\u4e00\u7bc7\u8bba\u6587\uff0c\u5e76\u5df2\u786e\u8ba4\u88abKnowledge-Based Systems\u671f\u520a\u5f55\u7528\uff0c\u8bba\u6587\u5177\u4f53\u7b80\u4ecb\u5982\u4e0b:<\/p>\n

\u6807\u9898: Adaptive Community Detection Incorporating Topology and Content in Social Networks<\/strong><\/p>\n

\u4f5c\u8005: Meng Qin, Di Jin*, Kai Lei*, Bogdan Gabrys, Katarzyna Musial<\/strong><\/p>\n

\u5f15\u7528\u683c\u5f0f:\u00a0<\/strong><\/p>\n

@article{Qin2018Adaptive,
\ntitle={Adaptive community detection incorporating topology and content in social networks},
\nauthor={Qin, Meng and Jin, Di and Lei, Kai and Gabrys, Bogdan and Musialgabrys, Katarzyna},
\njournal={Knowledge Based Systems},
\nyear={2018}
\n}<\/p>\n

\u8bba\u6587\u4e0b\u8f7d\u94fe\u63a5:<\/strong> https:\/\/authors.elsevier.com\/c\/1X~J83OAb8tDQD<\/a><\/p>\n

\u90e8\u5206\u6838\u5fc3\u6e90\u4ee3\u7801\u4e0b\u8f7d\u94fe\u63a5:<\/strong> https:\/\/github.com\/KuroginQin\/ASCD<\/a><\/p>\n

\u82f1\u6587\u6458\u8981:\u00a0<\/strong>In social network analysis, community detection is a basic step to understand the structure and function of networks. Some conventional community detection methods may have limited performance because they merely focus on the networks\u2019 topological structure. Besides topology, content information is another significant aspect of social networks. Although some state-of-the-art methods started to combine these two aspects of information for the sake of the improvement of community partitioning, they often assume that topology and content carry similar information. In fact, for some examples of social networks, the hidden characteristics of content may unexpectedly mismatch with topology. To better cope with such situations, we introduce a novel community detection method under the framework of non- negative matrix factorization (NMF). Our proposed method integrates topology as well as content of networks and has an adaptive parameter (with two variations) to effectively control the contribution of content with respect to the identified mismatch degree. Based on the disjoint community partition result, we also introduce an additional overlapping community discovery algorithm, so that our new method can meet the application requirements of both disjoint and overlapping community detection. The case study using real social networks shows that our new method can simultaneously obtain the community structures and their corresponding semantic description, which is helpful to understand the semantics of communities. Related performance evaluations on both artificial and real networks further indicate that our method outperforms some state-of-the-art methods while exhibiting more robust behavior when the mismatch between topology and content is observed.<\/p>\n

\u4e2d\u6587\u7b80\u4ecb:<\/strong> \u5728\u793e\u4ea4\u7f51\u7edc\u5206\u6790\u4e2d\uff0c\u793e\u56e2\u53d1\u73b0(community detection)\u662f\u7406\u89e3\u7f51\u7edc\u7ed3\u6784\u548c\u529f\u80fd\u7684\u57fa\u672c\u6b65\u9aa4\u3002\u4e00\u4e9b\u4f20\u7edf\u7684\u793e\u56e2\u53d1\u73b0\u65b9\u6cd5\u7531\u4e8e\u53ea\u8003\u8651\u7f51\u7edc\u7684\u62d3\u6251\u7ed3\u6784\uff0c\u793e\u56e2\u5212\u5206\u7684\u6027\u80fd\u53d7\u5230\u9650\u5236\u3002\u9664\u4e86\u7f51\u7edc\u62d3\u6251\uff0c\u5185\u5bb9\u4fe1\u606f\u662f\u793e\u4ea4\u7f51\u7edc\u53e6\u4e00\u4e2a\u91cd\u8981\u65b9\u9762\u3002\u4e3a\u8fdb\u4e00\u6b65\u63d0\u5347\u793e\u56e2\u5212\u5206\u7684\u51c6\u786e\u6027\uff0c\u73b0\u6709\u7684\u4e00\u4e9b\u65b9\u6cd5\u4e5f\u5f00\u59cb\u7ed3\u5408\u7f51\u7edc\u62d3\u6251\u548c\u5185\u5bb9\uff0c\u7136\u800c\u8fd9\u4e9b\u65b9\u6cd5\u5747\u5047\u8bbe\u62d3\u6251\u548c\u5185\u5bb9\u5177\u6709\u76f8\u4f3c\u7684\u9690\u542b\u7279\u5f81\u3002\u5b9e\u9645\u4e0a\uff0c\u5bf9\u4e8e\u4e00\u4e9b\u771f\u5b9e\u793e\u4ea4\u7f51\u7edc\uff0c\u5185\u5bb9\u4fe1\u606f\u7684\u9690\u542b\u7279\u5f81\u53ef\u80fd\u4e0e\u62d3\u6251\u7ed3\u6784\u4e0d\u5339\u914d(mismatch)\u3002\u4e3a\u66f4\u597d\u5730\u5e94\u5bf9\u4e0a\u8ff0\u95ee\u9898\uff0c\u57fa\u4e8e\u975e\u8d1f\u77e9\u9635\u5206\u89e3(non-negative matrix factorization, NMF)\u6846\u67b6\u63d0\u51fa\u4e00\u79cd\u65b0\u7684\u793e\u56e2\u53d1\u73b0\u65b9\u6cd5\u3002\u8be5\u65b9\u6cd5\u7ed3\u5408\u4e86\u7f51\u7edc\u7684\u62d3\u6251\u7ed3\u6784\u548c\u5185\u5bb9\u4fe1\u606f\u4e24\u4e2a\u65b9\u9762\uff0c\u5e76\u5f15\u5165\u4e00\u4e2a\u57fa\u4e8e\u4e0d\u5339\u914d\u7a0b\u5ea6(mismatch degree)\u7684\u81ea\u9002\u5e94\u53c2\u6570\u6765\u63a7\u5236\u5185\u5bb9\u4fe1\u606f\u5728\u6df7\u5408\u6a21\u578b\u4e2d\u7684\u76f8\u5bf9\u4f5c\u7528\u3002\u5728\u975e\u91cd\u53e0\u793e\u56e2\u5212\u5206(disjoint community detection)\u7684\u57fa\u7840\u4e0a\uff0c\u8fd8\u63d0\u51fa\u4e86\u4e00\u79cd\u91cd\u53e0\u793e\u56e2\u53d1\u73b0(overlapping community detection)\u7b97\u6cd5\uff0c\u56e0\u6b64\u8be5\u6a21\u578b\u80fd\u591f\u540c\u65f6\u5e94\u5bf9\u975e\u91cd\u53e0\u548c\u91cd\u53e0\u793e\u56e2\u53d1\u73b0\u4e24\u79cd\u5e94\u7528\u573a\u666f\u3002\u5728\u771f\u5b9e\u793e\u4ea4\u7f51\u7edc\u6570\u636e\u96c6\u4e0a\u7684\u4e00\u4e2a\u5b9e\u4f8b\u5206\u6790\u4e5f\u8868\u660e\uff0c\u8be5\u65b9\u6cd5\u5177\u6709\u540c\u65f6\u5b8c\u6210\u7f51\u7edc\u793e\u56e2\u5212\u5206\u5e76\u751f\u6210\u793e\u56e2\u8bed\u4e49\u63cf\u8ff0\u7684\u80fd\u529b\uff0c\u800c\u751f\u6210\u7684\u8bed\u4e49\u63cf\u8ff0\u80fd\u591f\u6709\u6548\u5730\u5e2e\u52a9\u7406\u89e3\u7f51\u7edc\u793e\u56e2\u7ed3\u6784\u7684\u8bed\u4e49\u3002\u5728\u4eba\u5de5\u7f51\u7edc\u548c\u771f\u5b9e\u793e\u4ea4\u7f51\u7edc\u4e0a\u7684\u76f8\u5173\u6027\u80fd\u5206\u6790\u4e5f\u8868\u660e\uff0c\u8be5\u65b9\u6cd5\u76f8\u5bf9\u4e8e\u73b0\u6709\u65b9\u6cd5\u5177\u6709\u66f4\u597d\u7684\u793e\u56e2\u53d1\u73b0\u6027\u80fd\uff1b\u5e76\u4e14\u5728\u7f51\u7edc\u62d3\u6251\u4e0e\u5185\u5bb9\u4e0d\u5339\u914d\u8f83\u4e25\u91cd\u7684\u60c5\u51b5\u4e0b\uff0c\u8868\u73b0\u51fa\u4e86\u66f4\u5f3a\u7684\u9c81\u68d2\u6027\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

ACM SIGCOMM 2018 Workshop on Network Meets AI & ML […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[31,4],"tags":[],"_links":{"self":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/5845"}],"collection":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5845"}],"version-history":[{"count":5,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/5845\/revisions"}],"predecessor-version":[{"id":6199,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/5845\/revisions\/6199"}],"wp:attachment":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5845"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5845"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5845"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}