Harassment Telephone Recognition System Based On User Behavior Analysis

As increasing people's income level and the continuous development of mobile communication service,price for mobile devices continues to fall.Until December 2018,the total number of mobile terminals of China mobile is 925 million,the total number of mobile phone users of China Telecom is 303 million,the total number of mobile phone users of China Unicom is 315 million,and the total number of mobile phone users in China has reached 1.543 billion,it can be seen that mobile phones have become a necessity of people's life.The rapid development of mobile communication brings convenience to people,at the same time,it also makes some people conduct telemarketing,harassment and fraud for commercial purposes.Even some groups,both domestic and foreign,use mobile communications to promote and disseminate illicit information for political purposes,leading to the proliferation of harassment calls.It has become one of the important reasons for users to complain about telecom operators,and also has a significant impact on the load of communication network equipment.In order to create a green network environment,the identification of harassing calls has become one of the key problems to be solved by the government,telecom operators and academia.This paper is to identify the characteristics of harassing telephone behavior through big data,data mining and other technologies.Firstly,China mobile no.7 signaling data is taken as the analysis object,all kinds of communication data related to users are collected from the business support network,network pipe network,signaling network,customer service system and other systems of a provincial mobile company,and they are stored in a database;then,using techniques such as classified data mining,an analysis model of suspected harassing calls is built to analyze users' behaviors of making calls and realize the identification of harassing calls;Finally,through automatic dial measurement and other technologies to carry out callback evidence,and they are provided to telecom operators for processing to achieve the identification and blocking of harassment calls.Technical key points and innovation points of this paper are,first,the Bayesian algorithm in classification mining technology is applied to screen the behavior of harassing calls,and clustering mining technology is applied to analyze the life cycle of harassing calls,so as to ensure the use of limited resources to implement efficient interception;second is automatic evidence collection of suspected harassing calls.The test run results of the system implemented in this paper in a provincial company of China mobile show that,the method of identifying the harassing behavior by dialing the phone through the analysis of user behavior has a significant effect,it can be used as a means of identifying harassing calls.