|Table of Contents|

Network Hotspot Topic Detection Algorithm Based on Multi-center Model


Research Field:
Publishing date:


Network Hotspot Topic Detection Algorithm Based on Multi-center Model
WANG WeiYANG WuQI Hai-feng
Information Security Research Center,Harbin Engineering University,Harbin 150001,China
topic detection hotspot topics multi-center single-pass clustering
In order to effectively eliminate the effect that the diversification of network topic related focuses have on the accuracy of network hotspot topic detection,a network topic multiple-center model is proposed.The association between story contents is layered,and the description ability for network topic is improved.A network hotspot topic detection algorithm based on the multi-center model is proposed,the topic centers are used to distinguish whether a new story belongs to an existing topic.The proposed algorithms based on single-pass clustering,the computation overhead is optimized by introducing topic center.The experimental results show that network topics can be detected entirely and accurately,and at the same time the performance is comparatively perfect.The proposed algorithm can be applied in hotspot topic detection in large-scale networks with dynamic data stream environments.


[ 1] Allan J, Papka R. On-line new event detection and tracking[A]. Proceedings of the21stAnnual Interna-tionalACMSIGIRConference onResearchandDeve-l opment in Information Retrieval[ C]. Melbourne: ACMPress, 1998. 37- 45.
[ 2] SeoYW, SycaraK. Text clustering for topicdetection [ Z]. USA: CarnegieMellonUniversity, 2004.
[ 3] 雷震, 吴玲达,雷蕾, 等. 初始化类中心的增量K均值法及其在新闻事件探测中的应用[ J]. 情报学报, 2006, 25( 3): 289-295.
[ 4] 赵华,赵铁军, 张妹,等. 基于内容分析的话题检测研究[ J]. 哈尔滨工业大学学报, 2006, 38( 10): 1740- 1743.
[ 5] 曾依灵,许洪波. 网络热点信息发现研究[ J]. 通信学报, 2007, 28( 12): 141- 146.
[ 6] 周亚东,孙钦东, 管晓宏, 等. 流量内容词语相关度的网络热点话题提取[ J]. 西安交通大学学报, 2007, 41( 10): 1142-1145, 1150.
[ 7] YangC, ShiX, WeiC. Tracing the event evolutionof terror attacks fromon-line news[ A]. Proceedings of IEEEInternationalConference onIntelligence andSe-curity Informatics[ C]. SanDiego: Lecture Notes in Computer Science, 2006. 343- 354.
[ 8] 谢海光,陈中润. 互联网内容及舆情深度分析模式 [ J]. 中国青年政治学院学报, 2006( 3): 95-100.


Last Update: 2012-11-19