基于多维多粒度分析的电信网用户行为模式挖掘

doi:10.11959/j.issn.2096-109x.2018083

Abstract

Abstract:

In order to better understand the behavior of users in telecom networks,it takes CDR (call detail record) data of large-scale telecom network as the research object.By using the mixed probability model and feature engineering method,the multi-dimension characteristics of the call time,call frequency and connections are analyzed from the perspective of user groups and individuals.It is further refined from different time granularities such as hour,day,and week to realize effective discovery of call behavior patterns for different user groups.The distribution characteristics of user behavior are modeled by mixed probability model,which solves the problem of describing the distribution characteristics such as user's call time and frequency.Based on the dataset of a regional telecom network,the performance of decision tree,naive Bayes and SVM classification algorithm are compared.It proves the validity and computational feasibility of the proposed method.The differences in communication behavior patterns of different groups are also compared by taking the service numbers like express,flight and bank as examples.

Key words: telecom network, multi-dimension, multi-granularity,, mixture-of-Gaussian, behavior pattern mining

CLC Number:

TP393

Xiaotao CHENG,Lixin JI,Ruiyang HUANG,Hongtao YU,Yizhuo YANG. User behavior pattern mining method based on multi-dimension and multi-granularity analysis in telecom networks[J]. Chinese Journal of Network and Information Security, 2018, 4(10): 39-51.

Figures/Tables 17

References 22

[1]	刘衍珩, 李飞鹏, 孙鑫, 朱建启 . 基于信息传播的社交网络拓扑模型[J]. 通信学报, 2013,34(4): 1-9.
	LIU Y Y , LI F P , SUN X , ZHU J Q . Social network model based on the transmission of information[J]. Journal on Communications, 2013,34(4): 1-9.
[2]	曹玖新, 吴江林, 石伟 ,等. 新浪微博网信息传播分析与预测[J]. 计算机学报, 2014,37(4): 779-790.
	CAO J X , WU J L , SHI W ,et al. Sina microblog information diffusion analysis and prediction[J]. Journal of Computers, 2014,37(4): 779-790.
[3]	杨杰, 李小平, 陈湉 . 基于增量时空轨迹大数据的群体挖掘方法[J]. 计算机研究与发展, 2014(s2): 76-85.
	YANG J , LI X P , CHEN T . A group mining method for incremental spatio-temporal trajectory bigdata[J]. Journal of Computer Research and Development, 2014(s2): 76-85.
[4]	李艳丽, 刘阳, 谢文波 ,等. 大数据发现非法传销网络[J]. 大数据, 2017,3(5): 106-112.
	LI Y L , LIU Y , XIE W P ,et al. Uncovering the illegal pyramid networks by big data[J]. Big Data Research, 2017,3(5): 106-112.
[5]	史殿习, 李寒, 杨若松 ,等. 用户日常频繁行为模式挖掘[J]. 国防科技大学学报, 2017,39(1): 74-80.
	SHI D X , LI H , YANG R S ,et al. Mining user frequent behavior patterns in daily life[J]. Journal of National University of Defense Technology, 2017,39(1): 74-80.
[6]	文雨, 王伟平, 孟丹 . 面向内部威胁检测的用户跨域行为模式挖掘[J]. 计算机学报, 2016,39(8): 1555-1569.
	WEN Y , WANG W P , MENG D . Mining user cross-domain behavior patterns for insider threat detection[J]. Journal of Computer, 2016,39(8): 1555-1569.
[7]	李全刚, 时金桥, 秦志光 ,等. 面向邮件网络事件检测的用户行为模式挖掘[J]. 计算机学报, 2014,37(5): 1135-1146.
	LI Q G , SHI J Q , QIN Z G ,et al. Mining user behavior patterns for event detection in email networks[J]. Journal of Computer, 2014,37(5): 1135-1146.
[8]	郝秀兰, 胡运发, 申倩 . 中文论坛内容监测的方法研究[J]. 中文信息学报, 2012,26(3): 129-137.
	HAO X L , HU Y F , SHEN Q . Research on content monitoring on Chinese web forums[J]. Journal of Chinese Information Processing, 2012,26(3): 129-137.
[9]	AIELLO W , CHUNG F , LU L . A random graph model for power law graphs[J]. Experimental Mathematics, 2001,10(1): 53-66.
[10]	ONNELA J P , SARAM?KI J , HYV?NEN J ,et al. Structure and tie strengths in mobile communication networks[J]. Proceedings of the national academy of sciences, 2007,104(18): 7332-7336.
[11]	JIANG Z Q , XIE W J , LI M X ,et al. Calling patterns in human communication dynamics[J]. Proceedings of the National Academy of Sciences, 2013,110(5): 1600-1605.
[12]	余晓平, 裴韬 . 手机通话网络度特征分析[J]. 物理学报, 2013,62(2): 1-9.
	YU X P , PEI T . Analysis on degree characteristics of mobile call network[J]. Acta Phys Sin, 2013,62(2): 1-9.
[13]	CAPDEVILA J , CERQUIDES J , TORRES J . Recognizing warblers:a probabilistic model for event detection in Twitter[C]. ICML2016 Anomaly Detection Workshop,New York,USA, 2016.
[14]	MANAVOGLU E , PAVLOV D , GILES C L . Probabilistic user behavior models[C]// Proceedings of the Third IEEE International Conference on Data Mining,Melbourne,Florida. 2003: 203-210.
[15]	BEUTEL A . User behavior modeling with large-scale graph analysis[D]. Computer Science Department School of Computer Science Carnegie Mellon University Pittsburgh,PA, 2016.
[16]	XU Z H , ZHANG Y,WU ,et al. Modeling user posting behavior on social media[C]. SIGIR’12,Portland,Oregon,USA. 2012.
[17]	张啸剑, 王淼, 孟小峰 . 差分隐私保护下一种精确挖掘 top-k 频繁模式方法[J]. 计算机研究与发展, 2014,51(1): 104-114.
	ZHANG X J , WANG M , MENG X F . An accurate method for mining top-k frequent pattern under differential privacy[J]. Journal of Computer Research and Development, 2014,51(1): 104-114.
[18]	CUI P , LIU H , AGGARWAL C ,et al. Uncovering and predicting human behaviors[J]. IEEE Intelligent Systems, 2016,03.
[19]	张宇翔, 孙菀, 杨家海 ,等. 新浪微博反垃圾中特征选择的重要性分析[J]. 通信学报, 2016,37(8):2016152-10.
	ZHANG Y X , SUN Y , YANG J H ,et al. Feature importance analysis for spammer detection in Sina Weibo[J]. Journal of Communications, 2016,37(8):2016152-10.
[20]	曹菁华, 胡访宇, 华烨 . 基于电信数据的通话行为模式发现[J]. 计算机仿真, 2014,11(31): 163-167.
	CAO J H , HU F Y , HUA Y . Call behavior pattern discovery based on telecom data[J]. Computer Simulation, 2014,11(31): 163-167.
[21]	OLSZEWSKI D . A probabilistic approach to fraud detection in telecommunications[J]. Knowledge-Based Systems, 2012,26(3): 246-258.
[22]	STEINLEY D . Properties of the hubert-arable adjusted rand index[J]. Psychological methods, 2004,9(3): 386.

Metrics

Recommended 0

No Suggested Reading articles found!

类别	号码	Top-1	Top-2	Top-3	Top-4	Top-5	累加和
私人号码	185****6507	0.161 62	0.141 41	0.101 01	0.101 01	0.060 61	0.565 66
	138****0696	0.283 87	0.103 23	0.070 97	0.070 97	0.064 52	0.593 55
	95558	0.021 53	0.016 28	0.009 75	0.009 72	0.009 59	0.066 87
服务类号码	95566	0.007 06	0.007 06	0.005 29	0.005 29	0.005 29	0.029 98
	95339	0.025 86	0.017 24	0.008 62	0.008 62	0.008 62	0.068 97
	95530	0.008 25	0.005 50	0.004 58	0.003 67	0.003 67	0.025 66

呼损码	失败原因
0	呼叫成功应答(主叫先挂机)
1	拨号中途挂机
2	振铃中途挂机
3	空号
4	用户市话忙
5	用户长话忙
6	呼叫异常释放
7~14	其他
15	呼叫故障
16	呼叫成功应答(被叫先挂机)

特征类别	呼叫号码特征	呼叫时长特征	呼叫频次特征	通联关系特征
	呼叫离散度	呼叫时长(天)	呼叫频次	拨打空号次数
	呼叫区域分布	最大呼叫时长(天)	呼叫号码数	出度
	主被叫占比	最小呼叫时长(天)	通话频次分布(天)	入度
指标	Top-k呼叫占比	通话时长分布(天)	通话频次分布(时)	呼叫离散度
	—	通话时长分布(时)	主叫挂机次数	频繁主叫
	—	—	被叫挂机次数	频繁被叫

用户数	记录数	时间	数据大小/GB
14 560 000	20亿	2017.4.1～2017.4.30	200

算法	准确率	召回率	AUC
决策树	0.82	0.85	0.862
朴素贝叶斯	0.84	0.76	0.848
SVM	0.85	0.79	0.853

User behavior pattern mining method based on multi-dimension and multi-granularity analysis in telecom networks

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 17

References 22

Related Articles 2

Metrics

Recommended 0

[1]	Baolin QIU, Ping YI. Adversarial examples defense method based on multi-dimensional feature maps knowledge distillation [J]. Chinese Journal of Network and Information Security, 2022, 8(2): 88-99.
[2]	Xuetao ZHANG,Meng SUN,Jinshuang WANG. Multi-granularity Android malware fast detection based on opcode [J]. Chinese Journal of Network and Information Security, 2019, 5(6): 85-94.