Harassing Call Identification Based On Data Mining Technology

Along with social progress and development of telecommunications industry, Communication by telephone has become an important part of everyday life. At the same time, In order to achieve their economic interests or political purposes, more and more companies, organizations and individuals harass their target populations by harassing calls. The main form of harassing calls is broadcasting advertising, forged or reactionary information to the specific populations. This behavior is not only disturbing normal life and work of people, taking up the limited network resources but also endangering national security and social stability.In the case of increasingly apparent harm of harassing, identification and monitoring of harassing calls has become one of the most problems which are needed to focus on by telecom operators. In the traditional case, telecom operators deal with harassing calls in passive mode. When the customer service staffs receive the complaints of customers about harassing calls, they will call the phone number which is complained to confirm whether the phone number is a real harass phone number. If it is confirmed, customer servicers will cancel call privileges of the harass phone number. Apparently it is less efficient because it will not find harassing calls forwardly and could not meet the demand of identification and monitoring of harassing calls.In this paper, the author studies the classification algorithm of data mining and applies related technologies to build a harassing calls identification system. This system analyses the call records and find suspected harassing calls forwardly for customer service. Through the effective use of historical data, this system can identify harassing calls forwardly and can reduce the customer service staffs' work effectively. It plays a significant role in harassing calls identification.The first chapter introduces the background, significance and the research status and development trend of domestic of this project.The second chapter introduces the data mining technology-related background knowledge, the concept of data mining, the relationship between data mining and knowledge discovery, the knowledge which can be found by data mining, the classification and the evaluation standards of data mining. This chapter also focuses on the classification algorithm in data mining.In the third chapter, the author introduces several major Bayesian classification algorithm were introduced firstly, secondly the author analyzes the characteristics of the harassing calls and combined with the practical demands sure to use minimum risk Bayesian classification to achieve the identification of the harassing calls, finally the author provided the implementation of the algorithm and the results of verification.Chapter IV introduces the design and implementation of the harassing calls recognition system which based on the improved classification algorithm.Chapter V describes the next step work and future prospects of the application of data mining technology to solve the problem of harassing calls identification.