\u5730\u70b9\uff1a<\/b><\/strong>\u6e56\u5317\u7701\u6b66\u6c49\u5e02\u6b66\u660c\u73de\u73c8\u5c71\u6b66\u6c49\u5927\u5b66\u8ba1\u7b97\u673a\u5b66\u96628\u697c<\/b><\/strong>\u62a5\u544a\u5385<\/b><\/strong><\/h3>\n\u4e3b\u529e\u5355\u4f4d\uff1a<\/b><\/strong>\u7f51\u5b89\u56fd\u9645\u5b66\u672f<\/b><\/strong>\u8bba\u575b<\/b><\/strong>InForSec<\/b><\/strong><\/h3>\n\u627f\u529e\u5355\u4f4d\uff1a\u6b66\u6c49\u5927\u5b66\u8ba1\u7b97\u673a\u5b66\u9662<\/strong><\/h3>\n![\"a2\"](\"https:\/\/www.inforsec.org\/wp\/wp-content\/uploads\/2016\/11\/a2-1.png\")
<\/p>\n
\u5609\u5bbe\u4ecb\u7ecd\uff1a<\/strong>
\n1\u3001\u7eaa\u5b88\u9886<\/strong>
\n \u6f14\u8bb2\u4e3b\u9898\uff1a SmartWalk: Enhancing Social Network Security via Adaptive Random Walks<\/strong>
\n![\"%e7%ba%aa%e5%ae%88%e9%a2%86\"](\"https:\/\/www.inforsec.org\/wp\/wp-content\/uploads\/2016\/11\/\u7eaa\u5b88\u9886-200x300.jpg\")
\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>
\n\u6d59\u6c5f\u5927\u5b66\u201c\u767e\u4eba\u8ba1\u5212\u201d\u7814\u7a76\u5458\u3001\u4f50\u6cbb\u4e9a\u7406\u5de5\u5b66\u9662Research Faculty\u3001\u535a\u58eb\u751f\u5bfc\u5e08\u30022013\u5e74\u83b7\u4f50\u6cbb\u4e9a\u5dde\u7acb\u5927\u5b66\u8ba1\u7b97\u673a\u79d1\u5b66\u535a\u58eb\u5b66\u4f4d\uff0c2015\u5e74\u83b7\u4f50\u6cbb\u4e9a\u7406\u5de5\u7535\u5b50\u4e0e\u8ba1\u7b97\u673a\u5de5\u7a0b\u535a\u58eb\u5b66\u4f4d\u3002\u7814\u7a76\u65b9\u5411\u4e3b\u8981\u4e3a\uff1a\u5927\u6570\u636e\u5b89\u5168\u9690\u79c1\u3001\u6570\u636e\u9a71\u52a8\u5b89\u5168\u3001\u7f51\u7edc\u5b89\u5168\u3001\u667a\u80fd\u8bbe\u5907\u5b89\u5168\u3001\u8ba1\u7b97\u673a\u7f51\u7edc\u7b49\u3002\u53d1\u8868\u8bba\u658770\u4f59\u7bc7\u3002
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\nRandom walks form a critical foundation in many social network-based security systems and applications. Currently, the design of such social security mechanisms is limited to the classical paradigm of using fixed-length random walks for all nodes on a social graph. However, the fixed-length walk paradigm induces a poor trade-off between security and other desirable properties. In this paper, we propose SmartWalk, a security enhancing system which incorporates adaptive random walks in social network security applications. We utilize a set of supervised machine learning techniques to predict the necessary random walk length based on the structural characteristics of a social graph. Using experiments on multiple real world topologies, we show that the desired walk length starting from a specific node can be well predicted given the local features of the node, and limited knowledge for a small set of training nodes. We describe node-adaptive and path-adaptive random walk usage models, where the walk length adaptively changes based on the starting node and the intermediate nodes on the path, respectively. We experimentally demonstrate the applicability of adaptive random walks on a number of social network based security and privacy systems, including Sybil defenses, anonymous communication and link privacy preserving systems, and show up to two orders of magnitude improvement in performance.<\/p>\n2\u3001\u9648\u5efa\u519b<\/strong>
\n\u6f14\u8bb2\u4e3b\u9898\uff1a\u5982\u4f55\u201c\u9ed1\u201d\u6389\u7f8e\u56fd\u56fd\u5bb6\u5b89\u5168\u5c40(NSA)\u7f51\u7ad9\u2014\u2014HTTP\u534f\u8bae\u5934\u7684\u4e8c\u4e49\u6027<\/strong>
\n![\"%e9%99%88%e5%bb%ba%e5%86%9b\"](\"https:\/\/www.inforsec.org\/wp\/wp-content\/uploads\/2016\/09\/\u9648\u5efa\u519b.png\")
\n\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>
\n\u9648\u5efa\u519b\uff0c\u6e05\u534e\u5927\u5b66\u535a\u58eb\u751f\uff0c\u5bfc\u5e08\u6bb5\u6d77\u65b0\u6559\u6388\u3002\u00a02013\u5e74\u5728\u6b66\u6c49\u5927\u5b66\u5b8c\u6210\u672c\u79d1\u5b66\u4e1a\u540e\uff0c\u88ab\u4fdd\u9001\u5230\u6e05\u534e\uff0c2015\u5e74\u5230\u7f8e\u56fdUC Berkeley\u5728Vern Paxson\u6307\u5bfc\u4e0b\u8fdb\u884c\u8bbf\u95ee\u7814\u7a76\u3002\u76ee\u524d\u53d1\u8868\u4e86\u4e09\u7bc7\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u8bba\u6587\uff08NDSS, CCS, IEEE S&P\uff09\uff0c\u5176\u4e2d\u5173\u4e8eCDN\u8f6c\u53d1\u5faa\u73af\u653b\u51fb\u7684NDSS\u8bba\u6587\u4ee5\u7b2c\u4e00\u4f5c\u8005\u8eab\u4efd\u83b7\u5f97\u4f1a\u8bae\u00a0“\u6770\u51fa\u8bba\u6587\u5956”\uff0c\u8fd9\u662f\u4e2d\u56fd\u5b66\u8005\u9996\u6b21\u83b7\u5f97\u7684\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u7684\u6700\u4f73\u8bba\u6587\u5956\u3002\u4ed6\u7684\u7814\u7a76\u6210\u679c\u4e0d\u4ec5\u53d7\u5230\u5b89\u5168\u5b66\u672f\u754c\u8ba4\u53ef\uff0c\u8fd8\u534f\u52a9\u4f17\u591a\u56fd\u5185\u5916\u5382\u5546\u548c\u5f00\u6e90\u8f6f\u4ef6\u4fee\u590d\u591a\u4e2a\u6f0f\u6d1e\u3002
\n\u6f14\u8bb2\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\n\u6211\u4eec\u53d1\u73b0\u4e86\u4e92\u8054\u7f51\u4e2d\u5e94\u7528\u6700\u5e7f\u6cdb\u7684HTTP\u534f\u8bae\u91cd\u5927\u5b89\u5168\u6f0f\u6d1eHost of Toubles\uff0c\u5f71\u54cd\u5927\u91cfHTTP\u8f6f\u4ef6\uff0c\u53ef\u4ee5\u9020\u6210\u7f13\u5b58\u6c61\u67d3\uff0c\u9632\u706b\u5899\u7ed5\u8fc7\u7b49\u653b\u51fb\uff0c\u8fd9\u4e00\u6210\u679c\u53d1\u8868\u5728\u56fd\u9645\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS 2016\u4e0a\u3002\u8be5\u6f0f\u6d1e\u662f\u7531\u4e8e\u4e0d\u540cHTTP\u5b9e\u73b0\u89e3\u6790\u548c\u5904\u7406HTTP\u8bf7\u6c42\u4e2d\u7684Host\u5934\u4e0d\u4e00\u81f4\u9020\u6210\u7684\u3002\u901a\u8fc7\u5229\u7528\u8fd9\u79cd\u4e0d\u4e00\u81f4\u6027\uff0c\u653b\u51fb\u8005\u53ea\u9700\u8981\u53d1\u9001\u4e00\u4e2a\u7cbe\u5fc3\u6784\u9020\u7684HTTP\u8bf7\u6c42\uff0c\u4fbf\u53ef\u4ee5\u5b9e\u73b0\u6c61\u67d3ISP\u7f13\u5b58\uff08\u5305\u62ec\u6bd2\u9c7f\u9c7c\u653b\u51fb\uff09\uff0c \u6c61\u67d3CDN\u7f13\u5b58\uff0c\u7ed5\u8fc7\u9632\u706b\u5899\uff0c\u7ed5\u8fc7WAF\u7b49\u653b\u51fb\u3002\u8fd9\u4e9b\u653b\u51fb\uff0c\u7279\u522b\u662f\u6c61\u67d3ISP\u7f13\u5b58\u653b\u51fb\uff0c\u5bf9\u4e92\u8054\u7f51\u5b89\u5168\u6709\u91cd\u5927\u5f71\u54cd\uff0c\u6211\u4eec\u7684\u5927\u89c4\u6a21\u6d4b\u91cf\u7ed3\u679c\u663e\u793a97%\u7684ISP\u7f13\u5b58\u7528\u6237\u90fd\u53ef\u4ee5\u88ab\u6211\u4eec\u53d1\u73b0\u7684\u653b\u51fb\u5f71\u54cd\u3002<\/p>\n3\u3001\u6f14\u8bb2\u4eba\uff1a\u5b59\u8317\u73c5<\/strong>
\n\u6f14\u8bb2\u4e3b\u9898\uff1aTaintART: \u57fa\u4e8e ART (Android RunTime) \u7684\u591a\u7ea7\u6c61\u70b9\u5206\u6790\u7cfb\u7edf<\/strong>
\n![\"%e5%ad%99%e8%8c%97%e7%8f%85\"](\"https:\/\/www.inforsec.org\/wp\/wp-content\/uploads\/2016\/11\/\u5b59\u8317\u73c5-225x300.png\")
\n\u6f14\u8bb2\u4eba\u4ecb\u7ecd\uff1a<\/b><\/strong>\u5b59\u8317\u73c5\u535a\u58eb\u6bd5\u4e1a\u4e8e\u9999\u6e2f\u4e2d\u6587\u5927\u5b66 (The Chinese University of Hong Kong) \u8ba1\u7b97\u673a\u79d1\u5b66\u4e0e\u6280\u672f\u7cfb\uff0c\u7f51\u7edc\u7cfb\u7edf\u7814\u7a76\u5b9e\u9a8c\u5ba4 (Advanced Networking and System Research Laboratory)\uff0c\u5e08\u4ece John C.S. Lui \u6559\u6388\u3002\u535a\u58eb\u671f\u95f4\uff0c\u4ed6\u591a\u6b21\u5728\u56fd\u9645\u4f1a\u8bae (CCS, RAID, ACSAC \u7b49) \u4e0a\u53d1\u8868\u8bba\u6587\u4ee5\u53ca\u6f14\u8bb2\uff0c\u5e76\u79ef\u6781\u53c2\u4e0e\u4f1a\u8bae\u3001\u671f\u520a\u7684\u5ba1\u7a3f\u3002\u4e3b\u8981\u7684\u7814\u7a76\u65b9\u5411\u662f\u79fb\u52a8\u7cfb\u7edf\u5b89\u5168\u3001\u6076\u610f\u8f6f\u4ef6\u5206\u6790\u3002\u4e3b\u9875\uff1ahttp:\/\/www.cse.cuhk.edu.hk\/~mssun<\/a>
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981<\/b><\/strong>\uff1a<\/b><\/strong>\u8fd1\u671f\uff0c Android \u4f7f\u7528\u4e86 AOT (ahead-of-time) \u7f16\u8bd1\u5668\u66ff\u4ee3\u4e86\u539f\u6709\u7684 Dalvik JIT \u7f16\u8bd1\u5668\u3002\u8fd9\u4e2a AOT \u7f16\u8bd1\u5668\u53ef\u4ee5\u5728\u5b89\u88c5\u9636\u6bb5\u628a\u5b57\u8282\u7801 (bytecode) \u8f6c\u6362\u6210\u539f\u751f\u7684\u673a\u5668\u7801 (native machine code)\u3002\u5728\u8fd9\u6b21\u6f14\u8bb2\u4e2d\uff0c\u4f5c\u8005\u4ecb\u7ecd\u4e86\u5728\u8fd9\u79cd\u65b0\u7684\u7f16\u8bd1\u7b56\u7565\u4e0b\uff0c\u5982\u4f55\u5229\u7528\u7f16\u8bd1\u5668\u5b9e\u73b0\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\u3002\u4f5c\u8005\u901a\u8fc7\u4fee\u6539 Android \u539f\u6709\u7684 AOT \u7f16\u8bd1\u5668\uff0c\u5229\u7528 CPU \u5bc4\u5b58\u5668\u5feb\u901f\u8bbf\u95ee\u6c61\u70b9\u6807\u8bb0\uff0c\u5b9e\u73b0\u4e86\u4e00\u4e2a\u9ad8\u6548\u7684\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\uff0c\u4f7f\u5f97\u6c61\u70b9\u5206\u6790\u5728 ART \u4e0a\u6210\u4e3a\u53ef\u80fd\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u672c\u6b21InForSec\u5b66\u672f\u8bba\u575b\u4e3b\u9898\u4e3a\u56fd\u9645\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS2016\u62a5\u544a\u5206\u4eab\uff0c\u5c06\u4e8e11\u670820\u65e5\u5728\u6b66\u6c49… Continue reading
\u627f\u529e\u5355\u4f4d\uff1a\u6b66\u6c49\u5927\u5b66\u8ba1\u7b97\u673a\u5b66\u9662<\/strong><\/h3>\n<\/p>\n
\u5609\u5bbe\u4ecb\u7ecd\uff1a<\/strong>
\n1\u3001\u7eaa\u5b88\u9886<\/strong>
\n \u6f14\u8bb2\u4e3b\u9898\uff1a SmartWalk: Enhancing Social Network Security via Adaptive Random Walks<\/strong>
\n\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>
\n\u6d59\u6c5f\u5927\u5b66\u201c\u767e\u4eba\u8ba1\u5212\u201d\u7814\u7a76\u5458\u3001\u4f50\u6cbb\u4e9a\u7406\u5de5\u5b66\u9662Research Faculty\u3001\u535a\u58eb\u751f\u5bfc\u5e08\u30022013\u5e74\u83b7\u4f50\u6cbb\u4e9a\u5dde\u7acb\u5927\u5b66\u8ba1\u7b97\u673a\u79d1\u5b66\u535a\u58eb\u5b66\u4f4d\uff0c2015\u5e74\u83b7\u4f50\u6cbb\u4e9a\u7406\u5de5\u7535\u5b50\u4e0e\u8ba1\u7b97\u673a\u5de5\u7a0b\u535a\u58eb\u5b66\u4f4d\u3002\u7814\u7a76\u65b9\u5411\u4e3b\u8981\u4e3a\uff1a\u5927\u6570\u636e\u5b89\u5168\u9690\u79c1\u3001\u6570\u636e\u9a71\u52a8\u5b89\u5168\u3001\u7f51\u7edc\u5b89\u5168\u3001\u667a\u80fd\u8bbe\u5907\u5b89\u5168\u3001\u8ba1\u7b97\u673a\u7f51\u7edc\u7b49\u3002\u53d1\u8868\u8bba\u658770\u4f59\u7bc7\u3002
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\nRandom walks form a critical foundation in many social network-based security systems and applications. Currently, the design of such social security mechanisms is limited to the classical paradigm of using fixed-length random walks for all nodes on a social graph. However, the fixed-length walk paradigm induces a poor trade-off between security and other desirable properties. In this paper, we propose SmartWalk, a security enhancing system which incorporates adaptive random walks in social network security applications. We utilize a set of supervised machine learning techniques to predict the necessary random walk length based on the structural characteristics of a social graph. Using experiments on multiple real world topologies, we show that the desired walk length starting from a specific node can be well predicted given the local features of the node, and limited knowledge for a small set of training nodes. We describe node-adaptive and path-adaptive random walk usage models, where the walk length adaptively changes based on the starting node and the intermediate nodes on the path, respectively. We experimentally demonstrate the applicability of adaptive random walks on a number of social network based security and privacy systems, including Sybil defenses, anonymous communication and link privacy preserving systems, and show up to two orders of magnitude improvement in performance.<\/p>\n2\u3001\u9648\u5efa\u519b<\/strong>
\n\u6f14\u8bb2\u4e3b\u9898\uff1a\u5982\u4f55\u201c\u9ed1\u201d\u6389\u7f8e\u56fd\u56fd\u5bb6\u5b89\u5168\u5c40(NSA)\u7f51\u7ad9\u2014\u2014HTTP\u534f\u8bae\u5934\u7684\u4e8c\u4e49\u6027<\/strong>
\n
\n\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>
\n\u9648\u5efa\u519b\uff0c\u6e05\u534e\u5927\u5b66\u535a\u58eb\u751f\uff0c\u5bfc\u5e08\u6bb5\u6d77\u65b0\u6559\u6388\u3002\u00a02013\u5e74\u5728\u6b66\u6c49\u5927\u5b66\u5b8c\u6210\u672c\u79d1\u5b66\u4e1a\u540e\uff0c\u88ab\u4fdd\u9001\u5230\u6e05\u534e\uff0c2015\u5e74\u5230\u7f8e\u56fdUC Berkeley\u5728Vern Paxson\u6307\u5bfc\u4e0b\u8fdb\u884c\u8bbf\u95ee\u7814\u7a76\u3002\u76ee\u524d\u53d1\u8868\u4e86\u4e09\u7bc7\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u8bba\u6587\uff08NDSS, CCS, IEEE S&P\uff09\uff0c\u5176\u4e2d\u5173\u4e8eCDN\u8f6c\u53d1\u5faa\u73af\u653b\u51fb\u7684NDSS\u8bba\u6587\u4ee5\u7b2c\u4e00\u4f5c\u8005\u8eab\u4efd\u83b7\u5f97\u4f1a\u8bae\u00a0“\u6770\u51fa\u8bba\u6587\u5956”\uff0c\u8fd9\u662f\u4e2d\u56fd\u5b66\u8005\u9996\u6b21\u83b7\u5f97\u7684\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u7684\u6700\u4f73\u8bba\u6587\u5956\u3002\u4ed6\u7684\u7814\u7a76\u6210\u679c\u4e0d\u4ec5\u53d7\u5230\u5b89\u5168\u5b66\u672f\u754c\u8ba4\u53ef\uff0c\u8fd8\u534f\u52a9\u4f17\u591a\u56fd\u5185\u5916\u5382\u5546\u548c\u5f00\u6e90\u8f6f\u4ef6\u4fee\u590d\u591a\u4e2a\u6f0f\u6d1e\u3002
\n\u6f14\u8bb2\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\n\u6211\u4eec\u53d1\u73b0\u4e86\u4e92\u8054\u7f51\u4e2d\u5e94\u7528\u6700\u5e7f\u6cdb\u7684HTTP\u534f\u8bae\u91cd\u5927\u5b89\u5168\u6f0f\u6d1eHost of Toubles\uff0c\u5f71\u54cd\u5927\u91cfHTTP\u8f6f\u4ef6\uff0c\u53ef\u4ee5\u9020\u6210\u7f13\u5b58\u6c61\u67d3\uff0c\u9632\u706b\u5899\u7ed5\u8fc7\u7b49\u653b\u51fb\uff0c\u8fd9\u4e00\u6210\u679c\u53d1\u8868\u5728\u56fd\u9645\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS 2016\u4e0a\u3002\u8be5\u6f0f\u6d1e\u662f\u7531\u4e8e\u4e0d\u540cHTTP\u5b9e\u73b0\u89e3\u6790\u548c\u5904\u7406HTTP\u8bf7\u6c42\u4e2d\u7684Host\u5934\u4e0d\u4e00\u81f4\u9020\u6210\u7684\u3002\u901a\u8fc7\u5229\u7528\u8fd9\u79cd\u4e0d\u4e00\u81f4\u6027\uff0c\u653b\u51fb\u8005\u53ea\u9700\u8981\u53d1\u9001\u4e00\u4e2a\u7cbe\u5fc3\u6784\u9020\u7684HTTP\u8bf7\u6c42\uff0c\u4fbf\u53ef\u4ee5\u5b9e\u73b0\u6c61\u67d3ISP\u7f13\u5b58\uff08\u5305\u62ec\u6bd2\u9c7f\u9c7c\u653b\u51fb\uff09\uff0c \u6c61\u67d3CDN\u7f13\u5b58\uff0c\u7ed5\u8fc7\u9632\u706b\u5899\uff0c\u7ed5\u8fc7WAF\u7b49\u653b\u51fb\u3002\u8fd9\u4e9b\u653b\u51fb\uff0c\u7279\u522b\u662f\u6c61\u67d3ISP\u7f13\u5b58\u653b\u51fb\uff0c\u5bf9\u4e92\u8054\u7f51\u5b89\u5168\u6709\u91cd\u5927\u5f71\u54cd\uff0c\u6211\u4eec\u7684\u5927\u89c4\u6a21\u6d4b\u91cf\u7ed3\u679c\u663e\u793a97%\u7684ISP\u7f13\u5b58\u7528\u6237\u90fd\u53ef\u4ee5\u88ab\u6211\u4eec\u53d1\u73b0\u7684\u653b\u51fb\u5f71\u54cd\u3002<\/p>\n3\u3001\u6f14\u8bb2\u4eba\uff1a\u5b59\u8317\u73c5<\/strong>
\n\u6f14\u8bb2\u4e3b\u9898\uff1aTaintART: \u57fa\u4e8e ART (Android RunTime) \u7684\u591a\u7ea7\u6c61\u70b9\u5206\u6790\u7cfb\u7edf<\/strong>
\n
\n\u6f14\u8bb2\u4eba\u4ecb\u7ecd\uff1a<\/b><\/strong>\u5b59\u8317\u73c5\u535a\u58eb\u6bd5\u4e1a\u4e8e\u9999\u6e2f\u4e2d\u6587\u5927\u5b66 (The Chinese University of Hong Kong) \u8ba1\u7b97\u673a\u79d1\u5b66\u4e0e\u6280\u672f\u7cfb\uff0c\u7f51\u7edc\u7cfb\u7edf\u7814\u7a76\u5b9e\u9a8c\u5ba4 (Advanced Networking and System Research Laboratory)\uff0c\u5e08\u4ece John C.S. Lui \u6559\u6388\u3002\u535a\u58eb\u671f\u95f4\uff0c\u4ed6\u591a\u6b21\u5728\u56fd\u9645\u4f1a\u8bae (CCS, RAID, ACSAC \u7b49) \u4e0a\u53d1\u8868\u8bba\u6587\u4ee5\u53ca\u6f14\u8bb2\uff0c\u5e76\u79ef\u6781\u53c2\u4e0e\u4f1a\u8bae\u3001\u671f\u520a\u7684\u5ba1\u7a3f\u3002\u4e3b\u8981\u7684\u7814\u7a76\u65b9\u5411\u662f\u79fb\u52a8\u7cfb\u7edf\u5b89\u5168\u3001\u6076\u610f\u8f6f\u4ef6\u5206\u6790\u3002\u4e3b\u9875\uff1ahttp:\/\/www.cse.cuhk.edu.hk\/~mssun<\/a>
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981<\/b><\/strong>\uff1a<\/b><\/strong>\u8fd1\u671f\uff0c Android \u4f7f\u7528\u4e86 AOT (ahead-of-time) \u7f16\u8bd1\u5668\u66ff\u4ee3\u4e86\u539f\u6709\u7684 Dalvik JIT \u7f16\u8bd1\u5668\u3002\u8fd9\u4e2a AOT \u7f16\u8bd1\u5668\u53ef\u4ee5\u5728\u5b89\u88c5\u9636\u6bb5\u628a\u5b57\u8282\u7801 (bytecode) \u8f6c\u6362\u6210\u539f\u751f\u7684\u673a\u5668\u7801 (native machine code)\u3002\u5728\u8fd9\u6b21\u6f14\u8bb2\u4e2d\uff0c\u4f5c\u8005\u4ecb\u7ecd\u4e86\u5728\u8fd9\u79cd\u65b0\u7684\u7f16\u8bd1\u7b56\u7565\u4e0b\uff0c\u5982\u4f55\u5229\u7528\u7f16\u8bd1\u5668\u5b9e\u73b0\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\u3002\u4f5c\u8005\u901a\u8fc7\u4fee\u6539 Android \u539f\u6709\u7684 AOT \u7f16\u8bd1\u5668\uff0c\u5229\u7528 CPU \u5bc4\u5b58\u5668\u5feb\u901f\u8bbf\u95ee\u6c61\u70b9\u6807\u8bb0\uff0c\u5b9e\u73b0\u4e86\u4e00\u4e2a\u9ad8\u6548\u7684\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\uff0c\u4f7f\u5f97\u6c61\u70b9\u5206\u6790\u5728 ART \u4e0a\u6210\u4e3a\u53ef\u80fd\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u672c\u6b21InForSec\u5b66\u672f\u8bba\u575b\u4e3b\u9898\u4e3a\u56fd\u9645\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS2016\u62a5\u544a\u5206\u4eab\uff0c\u5c06\u4e8e11\u670820\u65e5\u5728\u6b66\u6c49… Continue reading
<\/p>\n\n1\u3001\u7eaa\u5b88\u9886<\/strong>
\n \u6f14\u8bb2\u4e3b\u9898\uff1a SmartWalk: Enhancing Social Network Security via Adaptive Random Walks<\/strong>
\n
\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>\n\u6d59\u6c5f\u5927\u5b66\u201c\u767e\u4eba\u8ba1\u5212\u201d\u7814\u7a76\u5458\u3001\u4f50\u6cbb\u4e9a\u7406\u5de5\u5b66\u9662Research Faculty\u3001\u535a\u58eb\u751f\u5bfc\u5e08\u30022013\u5e74\u83b7\u4f50\u6cbb\u4e9a\u5dde\u7acb\u5927\u5b66\u8ba1\u7b97\u673a\u79d1\u5b66\u535a\u58eb\u5b66\u4f4d\uff0c2015\u5e74\u83b7\u4f50\u6cbb\u4e9a\u7406\u5de5\u7535\u5b50\u4e0e\u8ba1\u7b97\u673a\u5de5\u7a0b\u535a\u58eb\u5b66\u4f4d\u3002\u7814\u7a76\u65b9\u5411\u4e3b\u8981\u4e3a\uff1a\u5927\u6570\u636e\u5b89\u5168\u9690\u79c1\u3001\u6570\u636e\u9a71\u52a8\u5b89\u5168\u3001\u7f51\u7edc\u5b89\u5168\u3001\u667a\u80fd\u8bbe\u5907\u5b89\u5168\u3001\u8ba1\u7b97\u673a\u7f51\u7edc\u7b49\u3002\u53d1\u8868\u8bba\u658770\u4f59\u7bc7\u3002
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\nRandom walks form a critical foundation in many social network-based security systems and applications. Currently, the design of such social security mechanisms is limited to the classical paradigm of using fixed-length random walks for all nodes on a social graph. However, the fixed-length walk paradigm induces a poor trade-off between security and other desirable properties. In this paper, we propose SmartWalk, a security enhancing system which incorporates adaptive random walks in social network security applications. We utilize a set of supervised machine learning techniques to predict the necessary random walk length based on the structural characteristics of a social graph. Using experiments on multiple real world topologies, we show that the desired walk length starting from a specific node can be well predicted given the local features of the node, and limited knowledge for a small set of training nodes. We describe node-adaptive and path-adaptive random walk usage models, where the walk length adaptively changes based on the starting node and the intermediate nodes on the path, respectively. We experimentally demonstrate the applicability of adaptive random walks on a number of social network based security and privacy systems, including Sybil defenses, anonymous communication and link privacy preserving systems, and show up to two orders of magnitude improvement in performance.<\/p>\n
2\u3001\u9648\u5efa\u519b<\/strong> 3\u3001\u6f14\u8bb2\u4eba\uff1a\u5b59\u8317\u73c5<\/strong> \u672c\u6b21InForSec\u5b66\u672f\u8bba\u575b\u4e3b\u9898\u4e3a\u56fd\u9645\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS2016\u62a5\u544a\u5206\u4eab\uff0c\u5c06\u4e8e11\u670820\u65e5\u5728\u6b66\u6c49… Continue reading
\n\u6f14\u8bb2\u4e3b\u9898\uff1a\u5982\u4f55\u201c\u9ed1\u201d\u6389\u7f8e\u56fd\u56fd\u5bb6\u5b89\u5168\u5c40(NSA)\u7f51\u7ad9\u2014\u2014HTTP\u534f\u8bae\u5934\u7684\u4e8c\u4e49\u6027<\/strong>
\n
\n\u6f14\u8bb2<\/b><\/strong>\u4eba\u7b80\u4ecb<\/b><\/strong>\uff1a<\/b><\/strong>
\n\u9648\u5efa\u519b\uff0c\u6e05\u534e\u5927\u5b66\u535a\u58eb\u751f\uff0c\u5bfc\u5e08\u6bb5\u6d77\u65b0\u6559\u6388\u3002\u00a02013\u5e74\u5728\u6b66\u6c49\u5927\u5b66\u5b8c\u6210\u672c\u79d1\u5b66\u4e1a\u540e\uff0c\u88ab\u4fdd\u9001\u5230\u6e05\u534e\uff0c2015\u5e74\u5230\u7f8e\u56fdUC Berkeley\u5728Vern Paxson\u6307\u5bfc\u4e0b\u8fdb\u884c\u8bbf\u95ee\u7814\u7a76\u3002\u76ee\u524d\u53d1\u8868\u4e86\u4e09\u7bc7\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u8bba\u6587\uff08NDSS, CCS, IEEE S&P\uff09\uff0c\u5176\u4e2d\u5173\u4e8eCDN\u8f6c\u53d1\u5faa\u73af\u653b\u51fb\u7684NDSS\u8bba\u6587\u4ee5\u7b2c\u4e00\u4f5c\u8005\u8eab\u4efd\u83b7\u5f97\u4f1a\u8bae\u00a0“\u6770\u51fa\u8bba\u6587\u5956”\uff0c\u8fd9\u662f\u4e2d\u56fd\u5b66\u8005\u9996\u6b21\u83b7\u5f97\u7684\u7f51\u7edc\u5b89\u5168\u9876\u7ea7\u4f1a\u8bae\u7684\u6700\u4f73\u8bba\u6587\u5956\u3002\u4ed6\u7684\u7814\u7a76\u6210\u679c\u4e0d\u4ec5\u53d7\u5230\u5b89\u5168\u5b66\u672f\u754c\u8ba4\u53ef\uff0c\u8fd8\u534f\u52a9\u4f17\u591a\u56fd\u5185\u5916\u5382\u5546\u548c\u5f00\u6e90\u8f6f\u4ef6\u4fee\u590d\u591a\u4e2a\u6f0f\u6d1e\u3002
\n\u6f14\u8bb2\u5185\u5bb9\u6458\u8981\uff1a<\/b><\/strong>
\n\u6211\u4eec\u53d1\u73b0\u4e86\u4e92\u8054\u7f51\u4e2d\u5e94\u7528\u6700\u5e7f\u6cdb\u7684HTTP\u534f\u8bae\u91cd\u5927\u5b89\u5168\u6f0f\u6d1eHost of Toubles\uff0c\u5f71\u54cd\u5927\u91cfHTTP\u8f6f\u4ef6\uff0c\u53ef\u4ee5\u9020\u6210\u7f13\u5b58\u6c61\u67d3\uff0c\u9632\u706b\u5899\u7ed5\u8fc7\u7b49\u653b\u51fb\uff0c\u8fd9\u4e00\u6210\u679c\u53d1\u8868\u5728\u56fd\u9645\u5b89\u5168\u9876\u7ea7\u4f1a\u8baeCCS 2016\u4e0a\u3002\u8be5\u6f0f\u6d1e\u662f\u7531\u4e8e\u4e0d\u540cHTTP\u5b9e\u73b0\u89e3\u6790\u548c\u5904\u7406HTTP\u8bf7\u6c42\u4e2d\u7684Host\u5934\u4e0d\u4e00\u81f4\u9020\u6210\u7684\u3002\u901a\u8fc7\u5229\u7528\u8fd9\u79cd\u4e0d\u4e00\u81f4\u6027\uff0c\u653b\u51fb\u8005\u53ea\u9700\u8981\u53d1\u9001\u4e00\u4e2a\u7cbe\u5fc3\u6784\u9020\u7684HTTP\u8bf7\u6c42\uff0c\u4fbf\u53ef\u4ee5\u5b9e\u73b0\u6c61\u67d3ISP\u7f13\u5b58\uff08\u5305\u62ec\u6bd2\u9c7f\u9c7c\u653b\u51fb\uff09\uff0c \u6c61\u67d3CDN\u7f13\u5b58\uff0c\u7ed5\u8fc7\u9632\u706b\u5899\uff0c\u7ed5\u8fc7WAF\u7b49\u653b\u51fb\u3002\u8fd9\u4e9b\u653b\u51fb\uff0c\u7279\u522b\u662f\u6c61\u67d3ISP\u7f13\u5b58\u653b\u51fb\uff0c\u5bf9\u4e92\u8054\u7f51\u5b89\u5168\u6709\u91cd\u5927\u5f71\u54cd\uff0c\u6211\u4eec\u7684\u5927\u89c4\u6a21\u6d4b\u91cf\u7ed3\u679c\u663e\u793a97%\u7684ISP\u7f13\u5b58\u7528\u6237\u90fd\u53ef\u4ee5\u88ab\u6211\u4eec\u53d1\u73b0\u7684\u653b\u51fb\u5f71\u54cd\u3002<\/p>\n
\n\u6f14\u8bb2\u4e3b\u9898\uff1aTaintART: \u57fa\u4e8e ART (Android RunTime) \u7684\u591a\u7ea7\u6c61\u70b9\u5206\u6790\u7cfb\u7edf<\/strong>
\n
\n\u6f14\u8bb2\u4eba\u4ecb\u7ecd\uff1a<\/b><\/strong>\u5b59\u8317\u73c5\u535a\u58eb\u6bd5\u4e1a\u4e8e\u9999\u6e2f\u4e2d\u6587\u5927\u5b66 (The Chinese University of Hong Kong) \u8ba1\u7b97\u673a\u79d1\u5b66\u4e0e\u6280\u672f\u7cfb\uff0c\u7f51\u7edc\u7cfb\u7edf\u7814\u7a76\u5b9e\u9a8c\u5ba4 (Advanced Networking and System Research Laboratory)\uff0c\u5e08\u4ece John C.S. Lui \u6559\u6388\u3002\u535a\u58eb\u671f\u95f4\uff0c\u4ed6\u591a\u6b21\u5728\u56fd\u9645\u4f1a\u8bae (CCS, RAID, ACSAC \u7b49) \u4e0a\u53d1\u8868\u8bba\u6587\u4ee5\u53ca\u6f14\u8bb2\uff0c\u5e76\u79ef\u6781\u53c2\u4e0e\u4f1a\u8bae\u3001\u671f\u520a\u7684\u5ba1\u7a3f\u3002\u4e3b\u8981\u7684\u7814\u7a76\u65b9\u5411\u662f\u79fb\u52a8\u7cfb\u7edf\u5b89\u5168\u3001\u6076\u610f\u8f6f\u4ef6\u5206\u6790\u3002\u4e3b\u9875\uff1ahttp:\/\/www.cse.cuhk.edu.hk\/~mssun<\/a>
\n\u6f14\u8bb2<\/b><\/strong>\u5185\u5bb9\u6458\u8981<\/b><\/strong>\uff1a<\/b><\/strong>\u8fd1\u671f\uff0c Android \u4f7f\u7528\u4e86 AOT (ahead-of-time) \u7f16\u8bd1\u5668\u66ff\u4ee3\u4e86\u539f\u6709\u7684 Dalvik JIT \u7f16\u8bd1\u5668\u3002\u8fd9\u4e2a AOT \u7f16\u8bd1\u5668\u53ef\u4ee5\u5728\u5b89\u88c5\u9636\u6bb5\u628a\u5b57\u8282\u7801 (bytecode) \u8f6c\u6362\u6210\u539f\u751f\u7684\u673a\u5668\u7801 (native machine code)\u3002\u5728\u8fd9\u6b21\u6f14\u8bb2\u4e2d\uff0c\u4f5c\u8005\u4ecb\u7ecd\u4e86\u5728\u8fd9\u79cd\u65b0\u7684\u7f16\u8bd1\u7b56\u7565\u4e0b\uff0c\u5982\u4f55\u5229\u7528\u7f16\u8bd1\u5668\u5b9e\u73b0\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\u3002\u4f5c\u8005\u901a\u8fc7\u4fee\u6539 Android \u539f\u6709\u7684 AOT \u7f16\u8bd1\u5668\uff0c\u5229\u7528 CPU \u5bc4\u5b58\u5668\u5feb\u901f\u8bbf\u95ee\u6c61\u70b9\u6807\u8bb0\uff0c\u5b9e\u73b0\u4e86\u4e00\u4e2a\u9ad8\u6548\u7684\u6c61\u70b9\u5206\u6790\u7cfb\u7edf\uff0c\u4f7f\u5f97\u6c61\u70b9\u5206\u6790\u5728 ART \u4e0a\u6210\u4e3a\u53ef\u80fd\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"