Data Mining Based On Identification Of Spam Messages

SMS spam messages and malicious harassment as the development of the telecommunications industry has become intensified, result in serious harm to society. It is necessary to ensure that every normal user of the communication rights, but also accurate and timely for the malicious user restrictions and regulation, there is the biggest problem is "how the definition of spam messages accurately and efficiently." There is a large number of telecom operators by means of a malicious user restrictions. But the fundamental problem is, you need to distinguish this type of malicious clients that identify spam messages.Most of the existing means of identification with the spam messages are reviewed based on analysis of message content, such as the integration of reports from the public, by means of manual review. Because content-based audit to determine the maximum to ensure the accuracy of spam messages to avoid spam messages in the control of the process, making a lot of ordinary users of the normal demand for SMS communication is limited. But as long as the content is text-based means of identification, there is a poor response, timeliness is poor, high input costs, low efficiency. However, the incidence of spam messages in its behavior, has its many features. Is not limited to text content.Therefore, we hope that through the process of data mining to analyze the characteristics of many acts of spam messages. Summarized from a kind not based on message content, fast and efficient, the accuracy was high spam recognition method to replace or supplement the content of the current spam scanning means of identification. Based on this background and the objective needs, we have made through this research into customer behavior and data modeling and other means to establish a "spam messages recognition model " to improve the user to send spam messages to determine the accuracy and efficiency. Management component for the relevant telecommunications or telecommunications operators, the first time accurate to the spam messages to users to take the necessary restrictions provided by means of referential ideas and methods.