广东工业大学学报 ›› 2019, Vol. 36 ›› Issue (04): 24-30.doi: 10.12052/gdutxb.190052

• 综合研究 • 上一篇    下一篇

改进的聚类算法在恐怖袭击事件中的应用

何庆祥, 张巍   

  1. 广东工业大学 计算机学院, 广东 广州 510006
  • 收稿日期:2019-04-01 出版日期:2019-06-18 发布日期:2019-05-31
  • 作者简介:何庆祥(1994-),男,硕士研究生,主要研究方向为数据挖掘、大数据.
  • 基金资助:
    国家自然科学基金资助项目(61402118,61673123,61603100,61702110,61772141);教育部高教司项目(教高司函[2017]47号,教高司函[2018]4号);广东省科技计划项目(2015B090901016,2016B010108007);广东省教育厅项目(粤教高函[2018]179号,粤教高函[2018]1号,粤教高函[2015]133号,粤教高函[2014] 97号);广州市科技计划项目(201604020145,2016201604030034,201508010067,201604046017,201802030011,201802010042,201802010026)

Application of Improved Clustering Algorithm in Terrorist Attacks

He Qing-xiang, Zhang Wei   

  1. School of Computers, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2019-04-01 Online:2019-06-18 Published:2019-05-31

摘要: 恐怖袭击严重影响国际社会的稳定和人们生命财产安全,其形式、手段的多样化给反恐分析带来巨大挑战.为了把相似的恐怖袭击事件进行分组归类,并提高反恐分析员侦破案件的效率,本文基于全球恐怖主义数据库,提出了一种深度自编码表征(Deep Auto-Encoder Representation)的改进聚类算法,引入深度自编码器,将稀疏和嘈杂的原始数据映射为类内紧凑平滑的数据,提升聚类效果.实验结果表明,相比于传统的K-means聚类算法,改进后的算法可以提高聚类效果.本方法有利于反恐分析员将相似案件并案分析处理,找到案件的犯罪团伙.

关键词: 恐怖袭击, 深度自编码表征, 聚类

Abstract: The terrorist attacks have seriously affected the stability of the international community and the safety of people’s lives and property. The diversification of their forms and means has brought enormous challenges to the analysis of counter-terrorism. In order to classify similar terrorist attacks into groups, and to enhance the efficiency of counter-terrorism analysts in detecting cases, an improved clustering algorithm for Deep Auto-Encoder Representation is proposed based on the Global Terrorism Database. A deep self-encoder is introduced to map sparse and noisy raw data into compact and smooth data within the class, improving the clustering effect. The experimental results show that compared with the traditional K-means clustering algorithm, the improved algorithm can improve the clustering effect. This method is useful for counter-terrorism analysts to analyze similar cases and find criminal gangs in the case.

Key words: terrorist attack, deep auto-encoder representation, clustering

中图分类号: 

  • TP391
[1] 李国辉. 全球恐怖袭击时空演变及风险分析研究[D].合肥:中国科学技术大学,2014
[2] 滕少华, 吴昊, 李日贵, 等. 可调多趟聚类挖掘在电信数据分析中的应用[J]. 广东工业大学学报, 2014, 31(3):1-7 TENG S H, WU H, LI R G, et al. The application of the adjustable multi-times clustering algorithm in telecom data[J]. Journal of Guangdong University of technology, 2014, 31(3):1-7
[3] ALTMAN I, HUNTER M M. The employment and income effects of cleaner coal:the case of future gen and rural Illinois[J]. Clean Technologies and Environmental Policy, 2015, 17(6):1475-1485
[4] BARALIS E, CAGLIERO L, CERQUITELLI T, et al. Digging deep into weighted patient data through multiple-level patterns[J]. Information Sciences, 2015, 322:51-71
[5] 吴绍忠. 基于聚类分析的反恐情报中潜在恐怖团伙发现技术[J]. 警察技术, 2016(6):18-21 WU S Z. Potential terrorist gang discovery technology in counter-terrorism intelligence based on cluster analysis[J]. Police Technology, 2016(6):18-21
[6] 张巍, 麦志深. 核模糊谱聚类LOF降噪方法研究[J]. 广东工业大学学报, 2018, 35(6):77-82 ZHANG W, MAI Z S. Research on LOF noise reduction method based on kernel fuzzy spectral clustering[J]. Journal of Guangdong University of Technology, 2018, 35(6):77-82
[7] LIBRARY W E. National consortium for the study of terrorism and responses to terrorism[C]//International Symposium on High-power Laser Ablation. Santa Fe, NM:International Society for Optics and Photonics, 2010.
[8] PETER K, MICHAEL D I, JOHN P S. Countering terrorism and WMD:creating a global counter-terrorism network[M]. New York:Routledge, 2006.
[9] CHEN H, REID E, SINAI J, et al. Terrorism informatics:knowledge management and data mining for homeland security[M]. West Berlin:Springer Publishing Company, 2008.
[10] WANG X, MILLER E, SMARICK K, et al. Investigative visual analysis of global terrorism[J]. Computer Graphics Forum, 2008, 27(3):919-926
[11] JONES J, BUTKIEWICZ T, RIBARSKY W. Visualizing uncertainty for geographical information in the global terrorism database[J]. Proceedings of SPIE-The International Society for Optical Engineering, 2008, 15(9):1396
[12] 颜峻. 基于时空数据挖掘的社会安全(刑事)事件成因研究[D]. 北京清华大学, 2009.
[13] 彭如香, 张奥博, 杨涛, 等. 基于GTD的全球恐怖主义活动现状与发展趋势研究[J]. 计算机应用与软件, 2019, 36(1):1-5
[14] 李勇男, 梅建明. ROCK聚类在反恐情报分析中的应用研究[J]. 情报杂志, 2017(10):26-29 LI Y N, MEI J M. Application of ROCK clustering in anti-terrorism intelligence analysis[J]. Journal of Information, 2017(10):26-29
[15] AYTEKIN C, NI X, CRICRI F, et al. Clustering and unsupervised anomaly detection with L2 normalized deep auto-encoder representations[J]. arXiv preprint arXiv:1802.00187, 2018.
[16] CHANG C J, DAI W L. A grey silhouette coefficient for the small sample forecasting[C]//IEEE International Conference on Grey Systems & Intelligent Services. Macau:IEEE, 2014.
[17] LUKASIK S, KOWALSKI P A, CHARYTANOWICZ M, et al. Clustering using flower pollination algorithm and calinski-harabasz index[C]//2016 IEEE Congress on Evolutionary Computation. Vancouver:IEEE, 2016.
[1] 樊娟, 邓秀勤, 刘玉兰. 一种基于Fréchet距离的谱聚类算法[J]. 广东工业大学学报, 2023, 40(02): 39-44.
[2] 莫赞, 范梦婷, 刘洪伟, 严杨帆. 基于在线用户行为的产品非对称竞争市场结构研究[J]. 广东工业大学学报, 2023, 40(02): 111-119.
[3] 范梦婷, 刘洪伟, 高鸿铭, 何锐超. 电子商务平台下的竞争产品市场结构研究[J]. 广东工业大学学报, 2019, 36(06): 32-37.
[4] 谢振东, 冷梦甜, 吴金成. 基于一卡通数据的公交站点识别方法分析与研究[J]. 广东工业大学学报, 2019, 36(01): 23-28.
[5] 张巍, 麦志深. 核模糊谱聚类LOF降噪方法研究[J]. 广东工业大学学报, 2018, 35(06): 77-82.
[6] 马飞, 李娟. 基于聚类算法的MOOCs学习者分类及学习行为模式研究[J]. 广东工业大学学报, 2018, 35(03): 18-23.
[7] 王荣荣, 傅秀芬. 一种改进的mpts-HDBSCAN算法[J]. 广东工业大学学报, 2017, 34(03): 49-53.
[8] 陈丽, 曹熙, 林俊杰, 高鸿铭, 刘飞雅, 李艳艳. 基于数据挖掘的短期电力负荷风险预测分析[J]. 广东工业大学学报, 2017, 34(03): 105-109.
[9] 陈继峰, 刘广聪, 彭成平. 一种改进的无线传感器网络DV-Hop定位算法[J]. 广东工业大学学报, 2017, 34(02): 80-85.
[10] 申小敏, 李保俊, 孙旭, 徐维超. 基于卷积神经网络的大规模人脸聚类[J]. 广东工业大学学报, 2016, 33(06): 77-84.
[11] 王波, 钟映春, 陈俊彬. 融合AP和GMM的说话人识别方法研究[J]. 广东工业大学学报, 2015, 32(04): 145-149.
[12] 滕少华, 吴昊, 李日贵, 张巍, 刘冬宁, 梁路. 可调多趟聚类挖掘在电信数据分析中的应用[J]. 广东工业大学学报, 2014, 31(3): 1-7.
[13] 蒋盛益, 王连喜. 聚类分析研究的挑战性问题[J]. 广东工业大学学报, 2014, 31(3): 32-38.
[14] 李云, 鲍鸿. 语音分组识别技术的研究[J]. 广东工业大学学报, 2014, 31(2): 54-57.
[15] 姚蕾. 一种粒子群-Mamdani模糊神经网络的参数优化新算法[J]. 广东工业大学学报, 2014, 31(1): 36-39.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!