CAJ | 학술논문

随着 Web2．0技术的发展，微博网络已成为当前互联网中非常流行的网络应用，以即时性、碎片化、聚合性为特点的信息传播特性受到广大用户的欢迎。意见领袖在信息传播过程中发挥了重要的推动作用，有效地识别网络意见领袖，对于提高信息传播效果和可控性具有重要的现实意义。根据微博网络特点，提出一种基于多重链接的微博网络节点权重计算方法，根据节点权重计算结果并排序来识别意见领袖。采用 T -Test、Kendall -tau 和 Spareman Rank 三种统计学检验标准对不同的意见领袖识别算法进行对比实验，同时还对算法的准确率和召回率以及计算时间进行了实验。实验结果表明，从意见领袖识别能力、准确率和召回率以及计算时间等综合指标来看，该方法更具优势。
수착 Web2．0기술적발전，미박망락이성위당전호련망중비상류행적망락응용，이즉시성、쇄편화、취합성위특점적신식전파특성수도엄대용호적환영。의견령수재신식전파과정중발휘료중요적추동작용，유효지식별망락의견령수，대우제고신식전파효과화가공성구유중요적현실의의。근거미박망락특점，제출일충기우다중련접적미박망락절점권중계산방법，근거절점권중계산결과병배서래식별의견령수。채용 T -Test、Kendall -tau 화 Spareman Rank 삼충통계학검험표준대불동적의견령수식별산법진행대비실험，동시환대산법적준학솔화소회솔이급계산시간진행료실험。실험결과표명，종의견령수식별능력、준학솔화소회솔이급계산시간등종합지표래간，해방법경구우세。
With the development of Web2.0 technology,microblogging networks has become the extremely prevailing networks application in current Internet,and is welcomed by a vast numbers of users for its information dissemination characteristics of instantaneity,fragmentation and polymerisation.Opinion leaders play an important role in promoting the information dissemination process,and to effectively recognise the network opinion leaders has the important actual significance for improving the effect of information communication and controllability. According to the characteristics of microblogging networks,we propose a multilink-based weight computing method for microblogging networks node,and the opinion leaders are to be identified in light of the nodes weight calculation result as well as sorting.Three statistic testing criteria of T-Test,Kendall-tau and Spareman Rank are adopted in contrastive experiments of different opinion leaders’recognition algorithms, meanwhile,the experiments have also been performed on accurate rate and recall rate of the algorithm and on the computing time. Experimental results show that this method is more competitive from the comprehensive measures of recognition capability on opinion leaders, accurate rate,recall rate and computing time.