{"id":6097,"date":"2018-11-02T19:23:42","date_gmt":"2018-11-02T11:23:42","guid":{"rendered":"https:\/\/www.icnalb.cn\/?p=6097"},"modified":"2018-11-05T15:24:58","modified_gmt":"2018-11-05T07:24:58","slug":"%e8%bf%91%e5%b9%b4%e6%9d%a5%e5%86%85%e5%ae%b9%e4%b8%ad%e5%bf%83%e7%bd%91%e7%bb%9c%e4%b8%8e%e5%8c%ba%e5%9d%97%e9%93%be%e9%87%8d%e7%82%b9%e5%ae%9e%e9%aa%8c%e5%ae%a4%e4%bc%98%e8%b4%a8%e6%96%87%e7%ab%a0","status":"publish","type":"post","link":"https:\/\/www.icnlab.cn\/?p=6097","title":{"rendered":"\u8fd1\u5e74\u6765\u5185\u5bb9\u4e2d\u5fc3\u7f51\u7edc\u4e0e\u533a\u5757\u94fe\u91cd\u70b9\u5b9e\u9a8c\u5ba4\u4f18\u79c0\u6587\u7ae0\u6982\u89c8\uff08\u4e00\uff09"},"content":{"rendered":"<h2>\u5b9e\u9a8c\u5ba4\u7ecf\u8fc7\u8fd9\u51e0\u5e74\u7684\u79ef\u7d2f\uff0c\u53d1\u8868\u4e86\u4e0d\u5c11\u4f18\u79c0\u7684\u6587\u7ae0\uff0c\u73b0\u5728\u5728\u8fd9\u505a\u4e00\u4e2a\u5c55\u793a\u6982\u8981\uff0c\u5305\u62ec\u6587\u7ae0\u7684\u7b80\u4ecb\uff0cpdf\u94fe\u63a5\u4ee5\u53ca\u90e8\u5206\u6587\u7ae0\u7684\u5f00\u6e90\u4ee3\u7801github\u94fe\u63a5\u3002<\/h2>\n<h2><strong>COLING 2018\u56fd\u9645\u4f1a\u8bae<\/strong><\/h2>\n<ul>\n<li><strong>Knowledge as A Bridge: Improving Cross-domain Answer Selection with External Knowledge<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aYang Deng, Ying Shen, Min Yang, Yaliang Li, Nan Du, Wei Fan, Kai Lei<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/aclanthology.coli.uni-saarland.de\/papers\/C18-1279\/c18-1279\">https:\/\/aclanthology.coli.uni-saarland.de\/papers\/C18-1279\/c18-1279<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u7b54\u6848\u9009\u62e9\u662f\u4e00\u9879\u91cd\u8981\u800c\u5177\u6709\u6311\u6218\u6027\u7684\u4efb\u52a1\u3002\u5728\u5927\u91cf\u6807\u8bb0\u7684\u8bad\u7ec3\u6570\u636e\u53ef\u7528\u7684\u9886\u57df\u5df2\u7ecf\u53d6\u5f97\u4e86\u663e\u8457\u7684\u8fdb\u5c55\u3002\u7136\u800c\uff0c\u83b7\u5f97\u4e30\u5bcc\u7684\u6ce8\u91ca\u6570\u636e\u662f\u8017\u65f6\u548c\u6602\u8d35\u7684\u8fc7\u7a0b\uff0c\u5c06\u7b54\u6848\u9009\u62e9\u6a21\u578b\u5e94\u7528\u5230\u5177\u6709\u6709\u9650\u6807\u8bb0\u6570\u636e\u7684\u65b0\u9886\u57df\u5c06\u4f1a\u6709\u5f88\u5927\u7684\u969c\u788d\u3002\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u63d0\u51fa\u4e86\u77e5\u8bc6\u611f\u77e5\u6ce8\u610f\u529b\u7f51\u7edc\uff08KAN\uff09\uff0c\u4e00\u4e2a\u8de8\u9886\u57df\u7b54\u6848\u9009\u62e9\u7684\u8fc1\u79fb\u5b66\u4e60\u6846\u67b6\uff0c\u4f7f\u7528\u77e5\u8bc6\u5e93\u4f5c\u4e3a\u6865\u6881\uff0c\u4f7f\u77e5\u8bc6\u4ece\u6e90\u9886\u57df\u8f6c\u79fb\u5230\u76ee\u6807\u9886\u57df\u3002\u5177\u4f53\u5730\uff0c\u6211\u4eec\u8bbe\u8ba1\u4e86\u4e00\u4e2a\u77e5\u8bc6\u6a21\u5757\uff0c\u5c06\u57fa\u4e8e\u77e5\u8bc6\u7684\u8868\u793a\u5b66\u4e60\u96c6\u6210\u5230\u7b54\u6848\u9009\u62e9\u6a21\u578b\u4e2d\u3002\u6240\u5b66\u7684\u57fa\u4e8e\u77e5\u8bc6\u7684\u5411\u91cf\u8868\u793a\u7531\u6e90\u9886\u57df\u548c\u76ee\u6807\u9886\u57df\u5171\u4eab\uff0c\u8fd9\u4e0d\u4ec5\u5229\u7528\u5927\u91cf\u7684\u8de8\u9886\u57df\u6570\u636e\uff0c\u800c\u4e14\u8fd8\u53d7\u76ca\u4e8e\u6b63\u5219\u5316\u6548\u5e94\uff0c\u4ece\u800c\u5bfc\u81f4\u66f4\u901a\u7528\u7684\u6587\u672c\u8868\u793a\u6765\u5e2e\u52a9\u65b0\u9886\u57df\u4e2d\u7684\u4efb\u52a1\u3002\u4e3a\u4e86\u9a8c\u8bc1\u6211\u4eec\u7684\u6a21\u578b\u7684\u6709\u6548\u6027\uff0c\u6211\u4eec\u4f7f\u7528SQUAD-T\u6570\u636e\u96c6\u4f5c\u4e3a\u6e90\u57df\u6570\u636e\u96c6\u548c\u4e09\u4e2a\u5176\u4ed6\u6570\u636e\u96c6\uff08\u5373yahoo QA\uff0cTREC QA\u548cinsuranceQA\uff09\u4f5c\u4e3a\u76ee\u6807\u57df\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0cKAN\u5177\u6709\u5f88\u5f3a\u7684\u9002\u7528\u6027\u548c\u901a\u7528\u6027\uff0c\u5728\u8de8\u57df\u7b54\u6848\u9009\u62e9\u65b9\u9762\u660e\u663e\u4f18\u4e8e\u76ee\u524d\u6700\u597d\u7684\u6a21\u578b\u7b97\u6cd5\u3002<\/p>\n<p>Abstract<br \/>\nAnswer selection is an important but challenging task. Significant progress has been made in domains where a large amount of labeled training data is available. However, obtaining rich annotated data is a time-consuming and expensive process, creating a substantial barrier for applying answer selection models to a new domain which has limited labeled data. In this paper, we propose Knowledge-aware Attentive Network (KAN), a transfer learning framework for crossdomain answer selection, which uses the knowledge base as a bridge to enable knowledge transfer from the source domain to the target domains. Specifically, we design a knowledge module to integrate the knowledge-based representational learning into answer selection models. The learned knowledge-based representations are shared by source and target domains, which not only leverages large amounts of cross-domain data, but also benefits from a regularization effect that leads to more general representations to help tasks in new domains. To verify the effectiveness of our model, we use SQuAD-T dataset as the source domain and three other datasets (i.e., Yahoo QA, TREC QA and InsuranceQA) as the target domains. The experimental results demonstrate that KAN has remarkable applicability and generality, and consistently outperforms the strong competitors by a noticeable margin for cross-domain answer selection.<\/p>\n<p>&nbsp;<\/p>\n<ul>\n<li><strong>Cooperative Denoising for Distantly Supervised Relation Extraction<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aKai Lei*, Daoyuan Chen*, Yaliang Li, Nan Du, Min Yang, Wei Fan, Ying Shen. (* indicates equal contribution)<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/aclanthology.coli.uni-saarland.de\/papers\/C18-1036\/c18-1036\">https:\/\/aclanthology.coli.uni-saarland.de\/papers\/C18-1036\/c18-1036<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u8fdc\u76d1\u7763\u5173\u7cfb\u62bd\u53d6\u6781\u5927\u5730\u51cf\u5c11\u4e86\u4ece\u975e\u7ed3\u6784\u5316\u6587\u672c\u4e2d\u63d0\u53d6\u5173\u7cfb\u4e8b\u5b9e\u7684\u4eba\u529b\u6210\u672c\u3002\u4f46\u662f\u5b83\u5b58\u5728\u7740\u566a\u58f0\u6807\u7b7e\u7684\u95ee\u9898\uff0c\u8fd9\u4f1a\u6781\u5927\u635f\u5bb3\u62bd\u53d6\u6027\u80fd\u3002\u4e0e\u6b64\u540c\u65f6\uff0c\u77e5\u8bc6\u56fe\u8c31\u4e2d\u6240\u8868\u8fbe\u7684\u6709\u7528\u4fe1\u606f\u4ecd\u672a\u5728\u6700\u5148\u8fdb\u7684\u8fdc\u76d1\u7763\u5173\u7cfb\u63d0\u53d6\u65b9\u6cd5\u4e2d\u5f97\u5230\u5145\u5206\u5229\u7528\u3002\u9488\u5bf9\u8fd9\u4e9b\u6311\u6218\uff0c\u6211\u4eec\u63d0\u51fa\u4e86\u4e00\u79cd\u65b0\u7684\u534f\u540c\u53bb\u566a\u6846\u67b6\uff0c\u8be5\u6846\u67b6\u7531\u4e24\u4e2a\u5206\u522b\u5229\u7528\u6587\u672c\u8bed\u6599\u5e93\u548c\u77e5\u8bc6\u56fe\u8c31\u7684\u57fa\u7840\u7f51\u7edc\u7ec4\u6210\uff0c\u4ee5\u53ca\u4e00\u4e2a\u901a\u8fc7\u81ea\u9002\u5e94\u53cc\u5411\u77e5\u8bc6\u7cbe\u998f\u548c\u4ee5\u52a8\u6001\u96c6\u6210\u5e94\u5bf9\u566a\u58f0\u53d8\u5316\u5b9e\u4f8b\u7684\u534f\u4f5c\u6a21\u5757\u3002\u5728\u771f\u5b9e\u6570\u636e\u96c6\u4e0a\u7684\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u53ef\u4ee5\u6709\u6548\u51cf\u5c11\u566a\u58f0\u6807\u7b7e\uff0c\u5e76\u5728\u6700\u5148\u8fdb\u7684\u65b9\u6cd5\u4e0a\u53d6\u5f97\u5b9e\u8d28\u6027\u7684\u6539\u8fdb\u3002<\/p>\n<p>Abstract<br \/>\nDistantly supervised relation extraction greatly reduces human efforts in extracting relational facts from unstructured texts. However, it suffers from noisy labeling problem, which can degrade its performance. Meanwhile, the useful information expressed in knowledge graph is still underutilized in the state-of-the-art methods for distantly supervised relation extraction. In the light of these challenges, we propose CORD, a novel COopeRative Denoising framework, which consists two base networks leveraging text corpus and knowledge graph respectively, and a cooperative module involving their mutual learning by the adaptive bi-directional knowledge distillation and dynamic ensemble with noisy-varying instances. Experimental results on a real-world dataset demonstrate that the proposed method reduces the noisy labels and achieves substantial improvement over the state-of-the-art methods.<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<h2>ACM SIGIR 2018\u56fd\u9645\u4f1a\u8bae<\/h2>\n<ul>\n<li><strong>Ontology Evaluation with Path-based Text-aware Entropy Computation<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aYing Shen*, Daoyuan Chen*, Min Yang, Yaliang Li, Nan Du, Kai Lei\uff08* indicates equal contribution\uff09<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/dl.acm.org\/citation.cfm?id=3210067\">https:\/\/dl.acm.org\/citation.cfm?id=3210067<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u968f\u7740\u77e5\u8bc6\u4ea4\u6362\u7684\u91cd\u8981\u6027\u65e5\u76ca\u4e0a\u5347\uff0c\u672c\u4f53\u5df2\u6210\u4e3a\u77e5\u8bc6\u4ea4\u6362\u548c\u8bed\u4e49\u96c6\u6210\u7b49\u8bed\u4e49\u9a71\u52a8\u5e94\u7528\u7a0b\u5e8f\u5171\u4eab\u77e5\u8bc6\u6a21\u578b\u5f00\u53d1\u7684\u5173\u952e\u6280\u672f\u3002\u4f7f\u7528\u71b5\u6765\u6d4b\u91cf\u77e5\u8bc6\u78b1\u57fa\u7684\u53ef\u9884\u6d4b\u6027\u548c\u5197\u4f59\u6027\uff0c\u7279\u522b\u662f\u672c\u4f53\u7684\u53ef\u9884\u6d4b\u6027\u548c\u5197\u4f59\uff0c\u5df2\u7ecf\u53d6\u5f97\u4e86\u91cd\u5927\u8fdb\u5c55\u3002\u7136\u800c\uff0c\u76ee\u524d\u7528\u4e8e\u8bc4\u4f30\u672c\u4f53\u7684\u71b5\u5e94\u7528\u53ea\u8003\u8651\u5355\u70b9\u8fde\u63a5\uff0c\u800c\u4e0d\u8003\u8651\u8def\u5f84\u8fde\u63a5\uff0c\u4e3a\u6bcf\u4e2a\u5b9e\u4f53\u548c\u8def\u5f84\u5206\u914d\u76f8\u7b49\u7684\u6743\u91cd\uff0c\u5e76\u5047\u5b9a\u9876\u70b9\u662f\u9759\u6001\u7684\u3002\u9488\u5bf9\u8fd9\u4e9b\u4e0d\u8db3\uff0c\u672c\u6587\u63d0\u51fa\u4e86\u4e00\u79cd\u57fa\u4e8e\u8def\u5f84\u7684\u6587\u672c\u611f\u77e5\u71b5\u8ba1\u7b97\u65b9\u6cd5PTEC\uff0c\u8be5\u65b9\u6cd5\u8003\u8651\u4e0d\u540c\u9876\u70b9\u4e4b\u95f4\u7684\u8def\u5f84\u4fe1\u606f\u548c\u8def\u5f84\u5185\u7684\u6587\u672c\u4fe1\u606f\uff0c\u8ba1\u7b97\u6574\u4e2a\u7f51\u7edc\u7684\u8fde\u63a5\u8def\u5f84\u4ee5\u53ca\u4e0d\u540c\u8282\u70b9\u4e4b\u95f4\u7684\u4e0d\u540c\u6743\u91cd\u3002\u4ece\u57fa\u4e8e\u7ed3\u6784\u7684\u5d4c\u5165\u548c\u57fa\u4e8e\u6587\u672c\u7684\u5d4c\u5165\u83b7\u5f97\u7684\u4fe1\u606f\u4e58\u4ee5\u71b5\u8ba1\u7b97\u7684\u8fde\u901a\u6027\u77e9\u9635\u3002\u57fa\u4e8e\u672c\u4f53\u7edf\u8ba1\u4fe1\u606f(\u6570\u636e\u91cf)\u3001\u71b5\u8bc4\u4f30(\u6570\u636e\u8d28\u91cf)\u548c\u6848\u4f8b\u7814\u7a76(\u672c\u4f53\u7ed3\u6784\u548c\u6587\u672c\u53ef\u89c6\u5316)\uff0c\u5bf9\u4e09\u79cd\u771f\u5b9e\u4e16\u754c\u672c\u4f53\u8fdb\u884c\u4e86\u5b9e\u9a8c\u8bc4\u4ef7\u3002\u8fd9\u4e9b\u65b9\u9762\u76f8\u4e92\u8bc1\u660e\u4e86\u6211\u4eec\u65b9\u6cd5\u7684\u53ef\u9760\u6027\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0cPTEC\u80fd\u591f\u6709\u6548\u5730\u8bc4\u4ef7\u672c\u4f53\uff0c\u7279\u522b\u662f\u5728\u533b\u5b66\u9886\u57df\u3002<\/p>\n<p>ABSTRACT<br \/>\nWith the rising importance of knowledge exchange, ontologies have become a key technology in the development of shared knowledge models for semantic-driven applications, such as knowledge interchange and semantic integration. Significant progress has been made in the use of entropy to measure the predictability and redundancy of knowledge bases, particularly ontologies. However, the current entropy applications used to evaluate ontologies consider only single-point connectivity rather than path connectivity, assign equal weights to each entity and path, and assume that vertices are static. To address these deficiencies, the present study proposes a Path-based Text-aware Entropy Computation method, PTEC, by considering the path information between different vertices and the textual information within the path to calculate the connectivity path of the whole network and the different weights between various nodes. Information obtained from structure-based embedding and text-based embedding is multiplied by the connectivity matrix of the entropy computation. An experimental evaluation of three real-world ontologies is performed based on ontology statistical information (data quantity), entropy evaluation (data quality), and a case study (ontology structure and text visualization). These aspects mutually demonstrate the reliability of our method. Experimental results demonstrate that PTEC can effectively evaluate ontologies, particularly those in the medical field\u3002<\/p>\n<p>&nbsp;<\/p>\n<ul>\n<li><strong>Knowledge-aware Attentive Neural Network for Ranking Question Answer Pairs<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aYing Shen, Yang Deng, Min Yang, Yaliang Li, Nan Du, Wei Fan, Kai Lei<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<a href=\"https:\/\/dl.acm.org\/citation.cfm?id=3210081\">https:\/\/dl.acm.org\/citation.cfm?id=3210081<\/a><\/strong><\/li>\n<li><strong>\u5f00\u6e90\u4ee3\u7801\uff1a<\/strong><a href=\"https:\/\/github.com\/d123y456\/kablstm\">https:\/\/github.com\/d123y456\/kablstm<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u95ee\u7b54\u5bf9\u6392\u5e8f\u7531\u4e8e\u5176\u5e7f\u6cdb\u7684\u5e94\u7528\uff0c\u5982\u4fe1\u606f\u68c0\u7d22\u548c\u95ee\u7b54\uff08QA\uff09\uff0c\u8fd1\u5e74\u6765\u5f15\u8d77\u4e86\u8d8a\u6765\u8d8a\u591a\u7684\u5173\u6ce8\u3002\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u5728\u6b64\u4efb\u52a1\u4e0a\u53d6\u5f97\u4e86\u91cd\u5927\u8fdb\u5c55\u3002\u7136\u800c\uff0c\u5728\u4eba\u7c7b\u6587\u672c\u7406\u89e3\u4e2d\u8d77\u7740\u81f3\u5173\u91cd\u8981\u7684\u4f5c\u7528\u7684\u6587\u672c\u80cc\u666f\u4fe1\u606f\u548c\u9690\u85cf\u5728\u4e0a\u4e0b\u6587\u4e4b\u5916\u7684\u5173\u7cfb\uff0c\u5728\u6700\u8fd1\u53d6\u5f97\u6700\u597d\u7ed3\u679c\u7684\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u4e2d\u51e0\u4e4e\u6ca1\u6709\u88ab\u6df1\u5165\u7814\u7a76\u3002\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u63d0\u51fa\u4e86KABLSTM\uff0c\u4e00\u4e2a\u77e5\u8bc6\u611f\u77e5\u7684\u6ce8\u610f\u53cc\u5411\u957f\u77ed\u8bb0\u5fc6\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\uff0c\u5229\u7528\u5916\u90e8\u77e5\u8bc6\u4ece\u77e5\u8bc6\u56fe\u8c31\uff08KG\uff09\uff0c\u4e30\u5bcc\u4e86QA\u53e5\u5b50\u7684\u5411\u91cf\u8868\u793a\u5b66\u4e60\u3002\u5177\u4f53\u5730\uff0c\u6211\u4eec\u63d0\u51fa\u4e86\u4e0a\u4e0b\u6587\u77e5\u8bc6\u4ea4\u4e92\u5b66\u4e60\u7684\u4f53\u7cfb\u7ed3\u6784\uff0c\u5176\u4e2d\uff0c\u6211\u4eec\u8bbe\u8ba1\u4e86\u4e00\u4e2a\u4e0a\u4e0b\u6587\u4fe1\u606f\u5f15\u5bfc\u7684\u6ce8\u610f\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\uff08CNN\uff09\uff0c\u5c06\u5916\u90e8\u77e5\u8bc6\u5d4c\u5165\u5230\u53e5\u5b50\u8868\u793a\u4e2d\u3002\u6b64\u5916\uff0c\u63d0\u51fa\u4e86\u4e00\u79cd\u77e5\u8bc6\u611f\u77e5\u7684\u6ce8\u610f\u673a\u5236\u6765\u5173\u6ce8QA\u5bf9\u7684\u5404\u4e2a\u90e8\u5206\u4e4b\u95f4\u7684\u91cd\u8981\u4fe1\u606f\u3002KABLSTM\u5728\u4e24\u4e2a\u5e7f\u6cdb\u4f7f\u7528\u7684\u57fa\u51c6QA\u6570\u636e\u96c6\u4e0a\u8bc4\u4f30\uff1aWikiQA\u548cTREC QA\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0cKBLASTM\u5177\u6709\u8f83\u5f3a\u7684\u7ade\u4e89\u4f18\u52bf\u548c\u5f97\u5230\u4e86\u6700\u597d\u7684\u5b9e\u9a8c\u7ed3\u679c\u3002<\/p>\n<p>ABSTRACT<br \/>\nRanking question answer pairs has attracted increasing attention recently due to its broad applications such as information retrieval and question answering (QA). Significant progresses have been made by deep neural networks. However, background information and hidden relations beyond the context, which play crucial roles in human text comprehension, have received little attention in recent deep neural networks that achieve the state of the art in ranking QA pairs. In the paper, we propose KABLSTM, a Knowledge-aware Attentive Bidirectional Long Short-Term Memory, which leverages external knowledge from knowledge graphs (KG) to enrich the representational learning of QA sentences. Specifically, we develop a context-knowledge interactive learning architecture, in which a context-guided attentive convolutional neural network (CNN) is designed to integrate knowledge embeddings into sentence representations. Besides, a knowledge-aware attention mechanism is presented to attend interrelations between each segments of QA pairs. KABLSTM is evaluated on two widely-used benchmark QA datasets: WikiQA and TREC QA. Experiment results demonstrate that KABLSTM has robust superiority over competitors and sets state-of-the-art.<\/p>\n<p>&nbsp;<\/p>\n<h2>IEEE Global Communications Conference 2018\u56fd\u9645\u4f1a\u8bae<\/h2>\n<ul>\n<li><strong>Distributed Information-agnostic Flow Scheduling in Data Centers based on Wait-Time<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aKai Lei, Keke Li, Jie Xing, Bo Jin, Yi Wang<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5<\/strong>\uff1a\u6682\u65e0<\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u6570\u636e\u4e2d\u5fc3\u7f51\u7edc\u73b0\u6709\u7684\u6d41\u91cf\u8c03\u5ea6\u65b9\u6cd5\u4e3b\u8981\u662f\u4e3a\u4e86\u6700\u5c0f\u5316\u77ed\u6d41\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\uff0c\u5e76\u6ca1\u6709\u8003\u8651\u4f18\u5316\u5ef6\u65f6\u654f\u611f\u7684\u957f\u6d41\uff08\u6bd4\u5982VR\u89c6\u9891\u6d41\uff0cAI\u4ea4\u4e92\u5f0f\u95ee\u7b54\u6d41\uff09\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\u3002\u6b64\u5916\uff0c\u5728\u73b0\u6709\u7684\u6d41\u91cf\u8c03\u5ea6\u65b9\u6cd5\u4e2d\uff0c\u4fe1\u606f\u53ef\u77e5\u7684\u65b9\u6848\uff08\u5982L2DCT, D2TCP\uff09\u5728\u5b9e\u9645\u4e2d\u96be\u4ee5\u90e8\u7f72\uff0c\u8fd9\u662f\u56e0\u4e3a\u5b83\u4eec\u9700\u8981\u9884\u5148\u77e5\u9053\u6d41\u7684\u76f8\u5173\u4fe1\u606f\uff08\u5982\u6d41\u7684\u5927\u5c0f\uff09\uff1b\u800c\u4fe1\u606f\u4e0d\u53ef\u77e5\u7684\u8c03\u5ea6\u65b9\u6848\uff08\u5373PIAS\uff09\u867d\u7136\u4e0d\u9700\u8981\u63d0\u524d\u77e5\u9053\u6d41\u7684\u5927\u5c0f\u4fe1\u606f\uff0c\u4f46\u5b83\u9700\u8981\u4e00\u4e2a\u4e2d\u592e\u5316\u7684\u670d\u52a1\u5668\uff0c\u8fd9\u5c31\u5bfc\u81f4\u5728\u7f51\u7edc\u89c4\u6a21\u5f88\u5927\u65f6\uff0cPIAS\u7684\u53ef\u6269\u5c55\u6027\u5f88\u5dee\u3002<\/p>\n<p>\u8003\u8651\u5230\u73b0\u6709\u65b9\u6848\u7684\u5c40\u9650\u6027\uff0c\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u63d0\u51fa\u4e00\u79cd\u5206\u5e03\u5f0f\u4fe1\u606f\u4e0d\u53ef\u77e5\u7684\u6d41\u91cf\u8c03\u5ea6\u65b9\u6cd5\uff08DIAS\uff09\uff0c\u8be5\u65b9\u6cd5\u65e2\u80fd\u4f18\u5316\u77ed\u6d41\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\uff0c\u4e5f\u80fd\u4f18\u5316\u5ef6\u65f6\u654f\u611f\u7684\u957f\u6d41\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\u3002\u5728DIAS\u4e2d\uff0c\u6570\u636e\u5305\u662f\u6839\u636e\u5b83\u4eec\u7684\u4f18\u5148\u7ea7\u8fdb\u884c\u8f6c\u53d1\u7684\uff0c\u800c\u6570\u636e\u5305\u7684\u4f18\u5148\u7ea7\u662f\u6839\u636e\u5b83\u4eec\u5728\u53d1\u9001\u7aef\u7684\u7f13\u51b2\u533a\u5185\u7684\u7b49\u5f85\u65f6\u95f4\u51b3\u5b9a\u7684\uff0c\u6570\u636e\u5305\u7684\u7b49\u5f85\u65f6\u95f4\u8d8a\u4e45\uff0c\u5b83\u7684\u4f18\u5148\u7ea7\u8d8a\u4f4e\u3002\u6b64\u5916\uff0cDIAS\u4e0d\u50cfPIAS\u4e00\u6837\u91c7\u7528\u4e00\u4e2a\u96c6\u4e2d\u5316\u7684\u670d\u52a1\u5668\u6536\u96c6\u6d41\u91cf\u8d1f\u8f7d\u4fe1\u606f\uff0c\u800c\u662f\u91c7\u7528\u6bcf\u4e2a\u4ea4\u6362\u673a\u5c06\u6d41\u91cf\u8d1f\u8f7d\u4fe1\u606f\u9644\u5728ACK\u5305\u4e2d\u8fd4\u56de\u7ed9\u53d1\u9001\u7aef\u7684\u65b9\u5f0f\uff0c\u6d41\u91cf\u8d1f\u8f7d\u4fe1\u606f\u662f\u7528\u6765\u8c03\u6574\u51b3\u5b9a\u6570\u636e\u5305\u4f18\u5148\u7ea7\u7684\u9608\u503c\u7684\u3002ns-3\u6a21\u62df\u5668\u4e2d\u7684\u5b9e\u9a8c\u7ed3\u679c\u663e\u793a\uff0c\u4e0eDCTCP\u3001L2DCT\u76f8\u6bd4\uff0cDIAS\u5206\u522b\u80fd\u591f\u964d\u4f4e54.7%\u548c50.1%\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\uff0c\u6b64\u5916\uff0c\u4e0ePIAS\u76f8\u6bd4\uff0cDIAS\u80fd\u591f\u4fdd\u8bc1\u5ef6\u65f6\u654f\u611f\u7684\u957f\u6d41\u66f4\u77ed\u7684\u6d41\u5b8c\u6210\u65f6\u95f4\uff0c\u56e0\u6b64\u6bd4PIAS\u6027\u80fd\u66f4\u597d\u3002<\/p>\n<p>&nbsp;<\/p>\n<h2>The 2013 International Joint Conference on Neural Networks (IJCNN)<\/h2>\n<ul>\n<li><strong>Massively parallel learning of Bayesian networks with MapReduce for factor relationship analysis<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aChen Wei, Wang Tengjiao, Yang Dongqing, Lei Kai, Liu Yueqin<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/ieeexplore.ieee.org\/document\/6706814\">https:\/\/ieeexplore.ieee.org\/document\/6706814<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u8d1d\u53f6\u65af\u7f51\u7edc\uff08BN\uff09\u662f\u6570\u636e\u6316\u6398\u6280\u672f\u4e2d\u6700\u53d7\u6b22\u8fce\u7684\u6a21\u578b\u4e4b\u4e00\u3002 \u5927\u591a\u6570BN\u7ed3\u6784\u5b66\u4e60\u7b97\u6cd5\u90fd\u662f\u9488\u5bf9\u96c6\u4e2d\u5f0f\u6570\u636e\u96c6\u5f00\u53d1\u7684\uff0c\u5176\u4e2d\u6240\u6709\u6570\u636e\u90fd\u88ab\u6536\u96c6\u5230\u4e00\u4e2a\u8ba1\u7b97\u673a\u8282\u70b9\u4e2d\u3002 \u4ece\u5927\u89c4\u6a21\u6570\u636e\u4e2d\u5b66\u4e60BN\u7ed3\u6784\u5f80\u5f80\u6210\u672c\u592a\u9ad8\u6216\u4e0d\u5207\u5b9e\u9645\u3002 \u901a\u8fc7\u5177\u6709map\u548creduce\u4e24\u4e2a\u529f\u80fd\u7684\u7b80\u5355\u754c\u9762\uff0cMapReduce\u6709\u52a9\u4e8e\u5e76\u884c\u5b9e\u73b0\u8bb8\u591a\u5b9e\u9645\u4efb\u52a1\uff0c\u4f8b\u5982\u641c\u7d22\u5f15\u64ce\u7684\u6570\u636e\u5904\u7406\u548c\u673a\u5668\u5b66\u4e60\u3002 \u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u901a\u8fc7\u4f7f\u7528MapReduce\u96c6\u7fa4\uff0c\u63d0\u51fa\u4e86\u4e00\u79cd\u57fa\u4e8e\u5927\u89c4\u6a21\u65e5\u671f\u96c6\u7684BN\u7ed3\u6784\u5e76\u884c\u7b97\u6cd5\u3002 \u6211\u4eec\u8ba8\u8bba\u4e86\u4f7f\u7528MapReduce\u8fdb\u884cBN\u7ed3\u6784\u5b66\u4e60\u7684\u597d\u5904\uff0c\u5e76\u901a\u8fc7\u5c06\u5176\u5e94\u7528\u4e8e\u8d22\u52a1\u5206\u6790\u9886\u57df\u7684\u73b0\u5b9e\u4e16\u754c\u8d22\u52a1\u56e0\u7d20\u5173\u7cfb\u5b66\u4e60\u4efb\u52a1\u6765\u6f14\u793a\u8be5\u65b9\u6cd5\u7684\u6027\u80fd\u3002<\/p>\n<p>ABSTRACT<\/p>\n<p>Bayesian Network (BN) is one of the most popular models in data mining technologies. Most of the algorithms of BN structure learning are developed for the centralized datasets, where all the data are gathered into a single computer node. They are often too costly or impractical for learning BN structures from large scale data. Through a simple interface with two functions, map and reduce, MapReduce facilitates parallel implementation of many real-world tasks such as data processing for search engines and machine learning. In this paper, we present a parallel algorithm for BN structure leaning from large-scale dateset by using a MapReduce cluster. We discuss the benefits of using MapReduce for BN structure learning, and demonstrate the performance of this approach by applying it to a real world financial factor relationships learning task from the domain of financial analysis.<\/p>\n<p>&nbsp;<\/p>\n<h2>2014 IEEE 33rd International Performance Computing and Communications Conference (IPCCC)<\/h2>\n<ul>\n<li><strong>Hot topic analysis and content mining in social media<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aYu Qian, Weng WeiTao, Zhang Kai, Lei Kai, Xu Kuai<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/ieeexplore.ieee.org\/document\/7017056?tp=&amp;arnumber=7017056\">https:\/\/ieeexplore.ieee.org\/document\/7017056?tp=&amp;arnumber=7017056<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u65b0\u6d6a\u5fae\u535a\u5df2\u6210\u4e3a\u4e2d\u56fd\u8d8a\u6765\u8d8a\u91cd\u8981\u7684\u793e\u4ea4\u5a92\u4f53\uff0c\u5206\u4eab\u6700\u65b0\u6d88\u606f\uff0c\u63a8\u5e7f\u65b0\u4ea7\u54c1\uff0c\u8ba8\u8bba\u6709\u4e89\u8bae\u7684\u95ee\u9898\u3002\u65b0\u6d6a\u5fae\u535a\u5bf9\u793e\u4f1a\u7684\u91cd\u8981\u6027\u65e5\u76ca\u63d0\u9ad8\uff0c\u56e0\u6b64\u4e86\u89e3\u6570\u767e\u4e07\u6d3b\u8dc3\u7528\u6237\u4e0d\u65ad\u53d1\u5e03\u548c\u641c\u7d22\u70ed\u95e8\u8bdd\u9898\u7684\u201c\u4ec0\u4e48\u201d\uff0c\u201c\u4f55\u65f6\u201d\uff0c\u201c\u8c01\u201d\u975e\u5e38\u91cd\u8981\u3002\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u5f00\u53d1\u4e86\u4e00\u79cd\u7cfb\u7edf\u7684\u65b9\u6cd5\u6765\u63cf\u8ff0\u65b0\u6d6a\u5fae\u535a\u7528\u6237\u5728\u56db\u4e2a\u6708\u7684\u65f6\u95f4\u8de8\u5ea6\u5185\u641c\u7d22\u7684\u70ed\u95e8\u8bdd\u9898\u7684\u65f6\u95f4\u5206\u5e03\uff0c\u5e76\u53d1\u73b0\u76f8\u5173\u7684\u70ed\u95e8\u8bdd\u9898\uff0c\u8fd9\u4e9b\u8bdd\u9898\u4e0d\u4ec5\u7531\u540c\u4e00\u7528\u6237\u53d1\u5e03\uff0c\u800c\u4e14\u8fd8\u51fa\u73b0\u5728\u7c7b\u4f3c\u7684\u4e00\u7ec4\u63a8\u6587\u6d88\u606f\u3002\u6211\u4eec\u5206\u6790\u5b9e\u65f6\u65b0\u6d6a\u5fae\u535a\u63a8\u6587\u6570\u636e\u6d41\uff0c\u7814\u7a76\u7528\u6237\u641c\u7d22\u548c\u70ed\u95e8\u8bdd\u9898\u63a8\u6587\u6d3b\u52a8\u4e4b\u95f4\u7684\u6570\u91cf\u76f8\u5173\u6027\u548c\u65f6\u95f4\u5dee\u8ddd\u3002\u6b64\u5916\uff0c\u6211\u4eec\u8fd8\u7814\u7a76\u4e86\u793e\u4ea4\u5a92\u4f53\u548c\u641c\u7d22\u5f15\u64ce\u4e0a\u70ed\u95e8\u8bdd\u9898\u641c\u7d22\u4e4b\u95f4\u7684\u76f8\u5173\u6027\uff0c\u4ee5\u4e86\u89e3\u4e0d\u540c\u5e73\u53f0\u4e0a\u7684\u70ed\u95e8\u8bdd\u9898\u548c\u7528\u6237\u884c\u4e3a\u3002\u9274\u4e8e\u5206\u6790\u5927\u91cf\u63a8\u6587\u6570\u636e\u7684\u6311\u6218\uff0c\u6211\u4eec\u63a2\u7d22\u4e86Hadoop MapReduce\u6846\u67b6\uff0c\u4ee5\u4fbf\u4ece\u6536\u96c6\u7684\u6570\u636e\u96c6\u4e2d\u6709\u6548\u5904\u7406\u6570\u767e\u4e07\u6761\u63a8\u6587\uff0c\u5e76\u91cf\u5316MapReduce\u5728\u5206\u6790\u63a8\u6587\u6d41\u65f6\u7684\u6027\u80fd\u4f18\u52bf\u3002\u636e\u6211\u4eec\u6240\u77e5\uff0c\u672c\u6587\u9996\u6b21\u5c1d\u8bd5\u5728\u65b0\u6d6a\u5fae\u535a\u4e0a\u63cf\u8ff0\u70ed\u95e8\u8bdd\u9898\u7684\u65f6\u95f4\u641c\u7d22\u6a21\u5f0f\uff0c\u5e76\u7814\u7a76\u5b83\u4eec\u4e0e\u63a8\u7279\u6570\u636e\u6d41\u4ee5\u53ca\u641c\u7d22\u5f15\u64ce\u7edf\u8ba1\u6570\u636e\u7684\u76f8\u5173\u6027\u3002<\/p>\n<p>Abstract<\/p>\n<p>Sina Weibo has become an increasingly critical social media in China for sharing latest news, marketing new products, and discussing controversial issues. The rising importance of Sina Weibo on the society makes it very important to understand \u201cwhat\u201d, \u201cwhen\u201d, \u201cwho\u201d on hot topics that are being continuously tweeted and searched by millions of active users. In this paper, we develop a systematic approach to characterize temporal distribution of hot topics searched by Sina Weibo users over a four-month time-span and to uncover correlated hot topics that are not only tweeted by the same users, but also appear in the similar set of tweet messages. We analyze real-time Sina Weibo tweet data streams and study volume correlations and temporal gaps between user searches and tweeting activities on hot topics. In addition, we examine the correlations between hot topic searches on social media and on search engines to understand hot topics and user behaviors across different platforms. Given the challenges of analyzing massive amount of tweet data, we explore Hadoop MapReduce framework to effectively process millions of tweets from the collected data-sets, and quantify the performance benefits of MapReduce on analyzing tweet streams. To the best of our knowledge, this paper is the first effort to characterize temporal search patterns of hot topics on Sina Weibo and to study their correlations with tweeting data streams as well as search engine statistics.<\/p>\n<p>&nbsp;<\/p>\n<h2>2015 IEEE International Conference on Communications (ICC)<\/h2>\n<ul>\n<li><strong>Profiling the followers of the most influential and verified users on Sina Weibo<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aWang Huiyu, Lei Kai, Xu Kuai<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/ieeexplore.ieee.org\/document\/7248479?tp=&amp;arnumber=7248479\">https:\/\/ieeexplore.ieee.org\/document\/7248479?tp=&amp;arnumber=7248479<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1aTwitter\u548c\u65b0\u6d6a\u5fae\u535a\u7b49\u65b0\u793e\u4ea4\u5a92\u4f53\u5df2\u6210\u4e3a\u4f20\u64ad\u5f71\u54cd\u529b\u7684\u65e5\u76ca\u53d7\u6b22\u8fce\u7684\u6e20\u9053\uff0c\u6311\u6218\u7535\u89c6\u548c\u62a5\u7eb8\u7b49\u4f20\u7edf\u5a92\u4f53\u3002\u65b0\u6d6a\u5fae\u535a\u4e0a\u6700\u5177\u5f71\u54cd\u529b\u548c\u7ecf\u8fc7\u9a8c\u8bc1\u7684\u7528\u6237\uff0c\u4e5f\u79f0\u4e3a\u5927V\u8d26\u6237\uff0c\u7ecf\u5e38\u5438\u5f15\u6570\u767e\u4e07\u7c89\u4e1d\u548c\u7c89\u4e1d\uff0c\u5728\u793e\u4ea4\u5a92\u4f53\u4e0a\u521b\u5efa\u5927\u91cf\u201c\u4ee5\u540d\u4eba\u4e3a\u4e2d\u5fc3\u201d\u7684\u793e\u4ea4\u7f51\u7edc\uff0c\u5728\u4f20\u64ad\u7a81\u53d1\u65b0\u95fb\uff0c\u6700\u65b0\u6d3b\u52a8\u65b9\u9762\u53d1\u6325\u5173\u952e\u4f5c\u7528\u5173\u4e8e\u793e\u4f1a\u95ee\u9898\u7684\u4e89\u8bae\u6027\u610f\u89c1\u3002\u9274\u4e8e\u8fd9\u4e9b\u5e10\u6237\u7684\u91cd\u8981\u6027\uff0c\u4e86\u89e3\u8fd9\u4e9b\u5e10\u6237\u7684\u793e\u4ea4\u7f51\u7edc\u548c\u7528\u6237\u5f71\u54cd\u4ee5\u53ca\u63cf\u8ff0\u5176\u5173\u6ce8\u8005\u7684\u884c\u4e3a\u975e\u5e38\u91cd\u8981\u3002\u4e3a\u6b64\uff0c\u672c\u6587\u76d1\u63a7\u65b0\u6d6a\u5fae\u535a\u4e0a\u4e00\u7ec4\u9009\u5b9a\u7684\u6709\u5f71\u54cd\u529b\u7684\u7528\u6237\uff0c\u5e76\u6536\u96c6\u4ed6\u4eec\u7684\u63a8\u6587\u6d41\u4ee5\u53ca\u6765\u81ea\u5176\u5173\u6ce8\u8005\u7684\u8fd9\u4e9b\u63a8\u6587\u4e0a\u7684\u8f6c\u53d1\u548c\u8bc4\u8bba\u6d3b\u52a8\u3002\u6211\u4eec\u5bf9\u6765\u81ea\u65b0\u6d6a\u5fae\u535a\u7684\u63a8\u6587\u6570\u636e\u6d41\u7684\u5206\u6790\u63ed\u793a\u4e86\u8ffd\u968f\u8005\u5728\u8fd9\u4e9b\u6709\u5f71\u54cd\u529b\u7684\u7528\u6237\u7684\u63a8\u6587\u4e0a\u53d1\u8868\u8bc4\u8bba\u7684\u65f6\u95f4\u548c\u5185\u5bb9\uff0c\u5e76\u5728\u8bc4\u8bba\u4e2d\u53d1\u73b0\u4e86\u4e0d\u540c\u7684\u65f6\u95f4\u6a21\u5f0f\u548c\u8bcd\u6c47\u591a\u6837\u6027\u3002\u57fa\u4e8e\u4ece\u8ffd\u968f\u8005\u7279\u5f81\u4e2d\u83b7\u5f97\u7684\u6d1e\u5bdf\u529b\uff0c\u6211\u4eec\u8fdb\u4e00\u6b65\u5f00\u53d1\u4e86\u7b80\u5355\u76f4\u89c2\u7684\u7b97\u6cd5\uff0c\u5c06\u8ffd\u968f\u8005\u5206\u7c7b\u4e3a\u5783\u573e\u90ae\u4ef6\u53d1\u9001\u8005\u548c\u666e\u901a\u7c89\u4e1d\u3002\u6211\u4eec\u7684\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u5728\u5bf9\u8fd9\u4e9b\u6709\u5f71\u54cd\u529b\u7684\u8d26\u6237\u7684\u63a8\u6587\u53d1\u8868\u8bc4\u8bba\u7684\u7c89\u4e1d\u4e2d\u68c0\u6d4b\u5783\u573e\u90ae\u4ef6\u53d1\u9001\u8005\u65f6\uff0c\u6240\u63d0\u51fa\u7684\u7b97\u6cd5\u80fd\u591f\u8fbe\u523095.20\uff05\u7684\u5e73\u5747\u51c6\u786e\u5ea6\u3002<\/p>\n<p>Abstract<\/p>\n<p>The new social media such as Twitter and Sina Weibo has become an increasingly popular channel for spreading influence, challenging traditional media such as TVs and newspapers. The most influential and verified users, also called big-V accounts on Sina Weibo often attract million of followers and fans, creating massive \u201ccelebrity-centric\u201d social networks on the social media, which play a key role in disseminating breaking news, latest events, and controversial opinions on social issues. Given the importance of these accounts, it is very crucial to understand social networks and user influence of these accounts and profile their followers&#8217; behaviors. Towards this end, this paper monitors a selected group of influential users on Sina Weibo and collects their tweet streams as well as retweeting and commenting activities on these tweets from their followers. Our analysis on tweet data streams from Sina Weibo reveals when and what the followers comment on the tweets of these influential users, and discovers different temporal patterns and word diversity in the comments. Based on the insight gained from follower characteristics, we further develop simple and intuitive algorithms for classifying the followers into spammers and normal fans. Our experimental results demonstrate that the proposed algorithms are able to achieve an average accuracy of 95.20% in detecting spammers from the followers who have commented on the tweets of these influential accounts.<\/p>\n<p>&nbsp;<\/p>\n<h2>2016 IEEE International Conference on Communications (ICC)<\/h2>\n<ul>\n<li><strong>Detecting spam comments posted in micro-blogs using the self-extensible spam dictionary<\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aLiu Chenwei, Wang Jiawei, Lei Kai<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<\/strong><a href=\"https:\/\/ieeexplore.ieee.org\/document\/7511605\">https:\/\/ieeexplore.ieee.org\/document\/7511605<\/a><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u5fae\u535a\u7684\u9ad8\u4eba\u6c14\u6781\u5927\u5730\u4e30\u5bcc\u4e86\u4eba\u4eec\u7684\u751f\u6d3b\uff0c\u5141\u8bb8\u5728\u7ebf\u7528\u6237\u901a\u8fc7\u53d1\u8868\u8bc4\u8bba\u6765\u5206\u4eab\u4ed6\u4eec\u7684\u611f\u53d7\u3002\u4f46\u662f\uff0c\u6b64\u793e\u4ea4\u5a92\u4f53\u4e0a\u7684\u7528\u6237\u535a\u5ba2\u4e2d\u4e5f\u4f1a\u53d1\u5e03\u8d8a\u6765\u8d8a\u591a\u7684\u5783\u573e\u8bc4\u8bba\u3002\u5728\u672c\u6587\u4e2d\uff0c\u4e3a\u4e86\u6709\u6548\u5730\u68c0\u6d4b\u4e2d\u6587\u5fae\u535a\u4e2d\u7684\u5783\u573e\u8bc4\u8bba\uff0c\u6211\u4eec\u5f15\u5165\u8bed\u4e49\u5206\u6790\u6765\u6784\u5efa\u81ea\u6269\u5c55\u5783\u573e\u90ae\u4ef6\u5b57\u5178\uff0c\u5f53\u9891\u7e41\u51fa\u73b0\u5728\u5fae\u535a\u4e0a\u65f6\uff0c\u81ea\u52a8\u6269\u5c55\u81ea\u8eab\u3002\u8bed\u4e49\u5206\u6790\u7684\u4f7f\u7528\u53ef\u4ee5\u4e3a\u6211\u4eec\u63d0\u4f9b\u6709\u52a9\u4e8e\u68c0\u6d4b\u5783\u573e\u8bc4\u8bba\u7684\u9644\u52a0\u529f\u80fd\u3002\u901a\u8fc7\u6839\u636e\u6211\u4eec\u7684\u81ea\u6269\u5c55\u5783\u573e\u90ae\u4ef6\u5b57\u5178\u6807\u51c6\u7b5b\u9009\u5fae\u535a\u8bc4\u8bba\u7684\u5783\u573e\u90ae\u4ef6\u6743\u91cd\u548c\u5783\u573e\u90ae\u4ef6\u6bd4\u4f8b\uff0c\u8fd8\u63d0\u51fa\u4e86\u6bd4\u4f8b\u6743\u91cd\u7b5b\u9009\u5668\uff08PWF\uff09\u6a21\u578b\u6765\u68c0\u6d4b\u4e24\u79cd\u5783\u573e\u8bc4\u8bba\uff08AD\u548c\u7c97\u4fd7\u8bc4\u8bba\uff09\u3002 \u3002\u6211\u4eec\u7684\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u5f53\u68c0\u6d4b\u5230AD\u548c\u7c97\u4fd7\u5783\u573e\u8bc4\u8bba\u7684\u7ec4\u5408\u65f6\uff0c\u6211\u4eec\u53ef\u4ee5\u5b9e\u73b087.9\uff05\u7684\u5e73\u5747\u68c0\u6d4b\u51c6\u786e\u5ea6\u3002\u7279\u522b\u662f\u5bf9\u4e8eAD\u5783\u573e\u8bc4\u8bba\u68c0\u6d4b\uff0c\u6211\u4eec\u53ef\u4ee5\u8fbe\u523096.2\uff05\u7684\u5e73\u5747\u51c6\u786e\u5ea6\uff0c\u8fd9\u6bd4\u4f7f\u7528\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u65f6\u66f4\u597d\u3002\u5bf9\u7ed3\u679c\u7684\u7edf\u8ba1\u5206\u6790\u9a8c\u8bc1\u4e86\u6211\u4eec\u63d0\u51fa\u7684\u65b9\u6cd5\u53ef\u4ee5\u6709\u6548\u5730\u8bc6\u522b\u5783\u573e\u8bc4\u8bba\u5e76\u5177\u6709\u76f8\u5bf9\u8f83\u9ad8\u7684\u51c6\u786e\u5ea6\u3002<\/p>\n<p>Abstract<\/p>\n<p>The high popularity of Weibo has greatly enriched people&#8217;s lives, allowing online users to share their feelings through posting comments. However, more and more spam comments are also being posted in users&#8217; blogs on this social media. In this paper, in order to effectively detect spam comments in Chinese micro-blogs, we introduce semantic analysis to construct a Self-Extensible Spam Dictionary which automatically expands itself when new words emerge on the micro-blogs frequently. The use of semantic analysis can provide us with additional features which are beneficial to detecting spam comments. A Proportion-Weight Filter (PWF) model is also proposed to detect two kinds of spam comments (AD and vulgar comments), by filtering the spam-weight and the spam-proportion of the Weibo comments based on our Self-Extensible Spam Dictionary criteria. Our experimental results demonstrate that when detecting a combination of both AD and vulgar spam comments, we can achieve an average detection accuracy of 87.9%. Particularly for AD spam comments detection, we can achieve an average accuracy of 96.2%, which is preferable compared to when using machine learning methods. The statistical analysis of the results verifies that our proposed methods can identify the spam comments effectively and to relatively high degrees of accuracy.<\/p>\n<p>&nbsp;<\/p>\n<h2>ACM-ICN &#8217;14 Proceedings of the 1st ACM Conference on Information-Centric Networking<\/h2>\n<ul>\n<li><strong>Scalable control panel for media streaming in NDN<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5\uff1a<a href=\"https:\/\/dl.acm.org\/citation.cfm?id=2660129.2660138&amp;coll=DL&amp;dl=GUIDE&amp;preflayout=flat\">https:\/\/dl.acm.org\/citation.cfm?id=2660129.2660138&amp;coll=DL&amp;dl=GUIDE&amp;preflayout=flat<\/a><\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aLei Kai, Yu Long Yu, Wei Jun<\/strong><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u672c\u6587\u8bbe\u8ba1\u5e76\u5b9e\u73b0\u4e86\u4e00\u4e2a\u57fa\u4e8eNDN\u7684\u5a92\u4f53\u6d41\u7cfb\u7edf\u53ef\u6269\u5c55\u63a7\u5236\u9762\u677f\u3002 \u8be5\u7cfb\u7edf\u662f\u57fa\u4e8e\u5148\u524d\u57fa\u4e8eIP\u7684P2P\u5a92\u4f53\u6d41\u7cfb\u7edfHippo [1]\u5f00\u53d1\u7684\uff0c\u8be5\u7cfb\u7edf\u5305\u542b\u4e00\u7ec4\u7528\u4e8e\u64cd\u7eb5P2P\u529f\u80fd\u7684\u63a7\u5236\u670d\u52a1\u5668\uff0c\u4f8b\u5982\u8ddf\u8e2a\u5668\u7b49\u3002\u7cfb\u7edf\u53ef\u6269\u5c55\u6027\u6210\u4e3a\u6700\u56f0\u96be\u7684\u95ee\u9898\u4e4b\u4e00\u3002 P2P\u7cfb\u7edf\u7684\u7528\u6237\u89c4\u6a21\u53d8\u5f97\u975e\u5e38\u5927\u3002 \u6211\u4eec\u5229\u7528SNC [2]\u7684\u76f8\u540c\u539f\u7406\u8bbe\u8ba1\u4e86NDN-Hippo\u7684\u63a7\u5236\u5c42\u3002 \u81f3\u4e8e\u5b9e\u73b0\uff0c\u6211\u4eec\u91c7\u53d6\u4e86\u4e24\u6b65\u7684\u65b9\u6cd5\uff1a\u9996\u5148\u5c06Hippo\u7684\u63a7\u5236\u5c42\u79fb\u690d\u5230\u57fa\u4e8eNDN\u7684\u7cfb\u7edf\uff0c\u7136\u540e\u518d\u79fb\u690d\u5a92\u4f53\u6d41\u91cf\u5c42\u3002 \u901a\u8fc7\u5206\u79bb\u63a7\u5236\u5c42\u548c\u5a92\u4f53\u5c42\uff0c\u6211\u4eec\u7684\u6f14\u793a\u8bc1\u660e\uff0c\u4e0d\u4ec5\u53ef\u4ee5\u5728NDN\u4e2d\u5de7\u5999\u5730\u548c\u672c\u80fd\u5730\u5b9e\u73b0\u8ddf\u8e2a\u5668\u7684\u67d0\u4e9b\u7ba1\u7406\u529f\u80fd\uff0c\u800c\u4e14\u8fd8\u5927\u5927\u63d0\u9ad8\u4e86NDN\u7248\u672c\u7684Hippo\u7684\u53ef\u6269\u5c55\u6027\u3002<\/p>\n<p>Abstract<\/p>\n<p>An NDN-based scalable control panel for a media streaming system was designed and implemented in this paper. The system is developed based on a previous IP-based P2P media streaming system named Hippo [1], which contains a group of control servers to manipulate P2P functionalities, such as the tracker, etc. System scalability becomes one of the most difficult problems when the user size of P2P system grows very large. We took the advantages from the same principle of SNC [2] to design the NDN-Hippo&#8217;s control layer. As for implementation, we took a two-step approach: First porting the control layer of Hippo to NDN-based system, then porting media traffic layer later. By separating control and media layers, our demo demonstrates that not only some management functions of tracker can be smartly and instinctively achieved in NDN, but also the scalability of NDN version of Hippo has been greatly improved.<\/p>\n<p>&nbsp;<\/p>\n<h2>2015 IEEE Global Communications Conference (GLOBECOM)<\/h2>\n<ul>\n<li><strong>MDPF: An NDN Probabilistic Forwarding Strategy Based on Maximizing Deviation Method<\/strong><\/li>\n<li><strong>\u6587\u7ae0\u94fe\u63a5:<a href=\"https:\/\/ieeexplore.ieee.org\/document\/7417024?arnumber=7417024\">https:\/\/ieeexplore.ieee.org\/document\/7417024?arnumber=7417024<\/a><\/strong><\/li>\n<li><strong>\u4f5c\u8005\uff1aLei Kai, Yuan Jie, Wang Jiawei<\/strong><\/li>\n<\/ul>\n<p>\u7b80\u4ecb\uff1a\u8f6c\u53d1\u7b56\u7565\u662f\u547d\u540d\u6570\u636e\u7f51\u7edc\uff08NDN\uff09\u5b9e\u73b0\u52a8\u6001\uff0c\u81ea\u9002\u5e94\u548c\u667a\u80fd\u8f6c\u53d1\u7684\u5173\u952e\u7279\u6027\uff0c\u4f46\u8be5\u9886\u57df\u7684\u5de5\u4f5c\u4ecd\u5904\u4e8e\u521d\u7ea7\u9636\u6bb5\u3002\u5728\u672c\u6587\u4e2d\uff0c\u9009\u62e9NDN\u4e2d\u591a\u4e2a\u5907\u9009\u65b9\u6848\u4e2d\u7684\u54ea\u4e2a\u8f6c\u53d1\u63a5\u53e3\u88ab\u5b9a\u4e49\u4e3a\u591a\u5c5e\u6027\u51b3\u7b56\uff08MADM\uff09\u95ee\u9898\u548c\u57fa\u4e8e\u6700\u5927\u5316\u504f\u5dee\u7684\u6982\u7387\u8f6c\u53d1\uff08MDPF\uff09\u7b56\u7565\uff0c\u4ee5\u9009\u62e9\u6982\u7387\u8f6c\u53d1\u63a5\u53e3\u3002\u7531\u4e8e\u591a\u4e2a\u7f51\u7edc\u6307\u6807\uff08\u5982\u63a5\u53e3\u72b6\u6001\uff0c\u5f85\u5904\u7406\u7684\u5174\u8da3\u7f16\u53f7\uff09\u88ab\u4e00\u8d77\u8003\u8651\uff0c\u56e0\u6b64\u53ef\u4ee5\u66f4\u51c6\u786e\u5730\u83b7\u5f97\u6bcf\u4e2a\u66ff\u4ee3\u63a5\u53e3\u7684\u53ef\u7528\u6027\u3002\u56e0\u6b64\uff0c\u53ef\u4ee5\u5b9e\u73b0\u66f4\u597d\u7684\u5185\u5bb9\u4f20\u9012\u6548\u7387\u3002\u6b64\u5916\uff0cMDPF\u63d0\u4f9b\u4e86\u826f\u597d\u7684\u53ef\u6269\u5c55\u6027\uff0c\u56e0\u4e3a\u53ef\u4ee5\u6dfb\u52a0\u4efb\u4f55\u9002\u5f53\u7684\u5ea6\u91cf\u6765\u589e\u5f3a\u6216\u81ea\u5b9a\u4e49\u5b83\u3002\u6211\u4eec\u5728ndnSIM\u4e2d\u5b9e\u65bd\u8be5\u63d0\u6848\uff0c\u5e76\u5c06\u5176\u4e0e\u5404\u79cd\u62d3\u6251\u548c\u65b9\u6848\u4e0b\u7684BestRoute\u548c\u57fa\u4e8ePI\u7684\u7b56\u7565\u8fdb\u884c\u6bd4\u8f83\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0cMDPF\u7b56\u7565\u5bf9\u7f51\u7edc\u53d8\u5316\u54cd\u5e94\u66f4\u5feb\uff0c\u66f4\u654f\u611f\uff0c\u53ef\u4ee5\u5b9e\u73b0\u66f4\u9ad8\u7684\u541e\u5410\u91cf\uff0c\u66f4\u4f4e\u7684\u4e22\u5f03\u7387\u4ee5\u53ca\u66f4\u597d\u7684\u8d1f\u8f7d\u5747\u8861\u3002<\/p>\n<p>Abstract<\/p>\n<p>Forwarding strategy is the key feature of Named Data Networking (NDN) to realize dynamic, adaptive and intelligent forwarding, but work in this area is still at a very preliminary stage. In this paper, selecting which forwarding interface among multiple alternatives in NDN is defined as a multiple attribute decision making (MADM) problem and a maximizing deviation based probabilistic forwarding (MDPF) strategy is proposed to select forwarding interface on probability. Since multiple network metrics such as interface status, pending Interest numbers are considered together, each alternative interface&#8217;s availability is obtained more accurately. Thus, better content delivery efficiency can be achieved. In addition, MDPF provides good extensibility, as any appropriate metric can be added to enhance or customize it. We implement the proposal in ndnSIM and compare it with BestRoute and PI-based strategies under various topologies and scenarios. Experimental results show that MDPF strategy is more responsive and sensitive to network changes, and can realize higher throughput, lower drop rate as well as better load balance.<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u5b9e\u9a8c\u5ba4\u7ecf\u8fc7\u8fd9\u51e0\u5e74\u7684\u79ef\u7d2f\uff0c\u53d1\u8868\u4e86\u4e0d\u5c11\u4f18\u79c0\u7684\u6587\u7ae0\uff0c\u73b0\u5728\u5728\u8fd9\u505a\u4e00\u4e2a\u5c55\u793a\u6982\u8981\uff0c\u5305\u62ec\u6587\u7ae0\u7684\u7b80\u4ecb\uff0cpdf\u94fe\u63a5\u4ee5\u53ca\u90e8\u5206\u6587\u7ae0\u7684 [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[16,31,4],"tags":[],"_links":{"self":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/6097"}],"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=6097"}],"version-history":[{"count":15,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/6097\/revisions"}],"predecessor-version":[{"id":6121,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=\/wp\/v2\/posts\/6097\/revisions\/6121"}],"wp:attachment":[{"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6097"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6097"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.icnlab.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6097"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}