Font Size: a A A

Analysis And Research On The Characteristics Of Telecom Harassment Fraud Information

Posted on:2020-05-14Degree:MasterType:Thesis
Country:ChinaCandidate:F YangFull Text:PDF
GTID:2506306104996859Subject:Master of Engineering Management
Abstract/Summary:PDF Full Text Request
Telecom fraud and harassment have occurred in many countries,and have also experienced development in China for more than ten years.Due to its large population and other reasons,China has not completely eliminated telecommunications fraud and harassment yet.The low-cost strategy of telecommunications fraud and harassment makes the call behavior of communication numbers appear significantly abnormal than the normal use of number behavior.Analysis of the communication behavior of normal and abnormal numbers can provide a basis for the screening of abnormal numbers.In this study,a large number of communication data of a user number group of an operator in city X were randomly extracted to form a normal call data matrix,and the communication data of suspected telecommunication fraud numbers identified by the public security department was used as the abnormal call data matrix.After analyzing the call characteristics of normal call data and abnormal call data,it is found that there are many differences between suspected fraud and normal calls.The comparison and screening of call characteristics makes the application of quantitative indicators more effective.The correlation analysis of characteristic parameters determines the number of callers.The six apparent characteristics of called number,roaming caller number,roaming caller number,calling number,and called number are highly correlated among normal callers and cannot be used for the selection of suspected fraud numbers at the same time.In the end,the combination of screening was used to determine the monthly caller as the first condition,and there were any two of the four indicators: online duration,number of times called,number of roaming times,and number of called numbers.Certain restrictions were set.A total of six combinations can meet the criteria for screening suspect numbers.This conclusion can provide accurate quantitative indicators for telecommunications operators to screen telecommunications fraud and harassment.At the same time,because the situation in different cities is different,the exploration of index screening in this study can also provide a reference for the screening of suspected numbers in other regions.As an attempt to distill the rules from the results,the screening model established by the results of this study cannot be applied to the changing environment overnight and needs to be continuously improved and supplemented in the application.
Keywords/Search Tags:Correlation analysis, Abnormal communication, Telecom fraud, Telecommunication harassment, Feature analysis, Screening model
PDF Full Text Request
Related items