Design And Implementation Of Short Message Intelligent Classify Based On Contents

In the field of mobile communication, short message service even though the birth is not long, but with a simple, easy, cheap, open design concept deep into people’s work and life, become an important way of daily communication. According to statistics of the Ministry of Industry and Information, to the end of July2008, more than600million mobile phone users; while from January to April2008, short messages sent reached229billion, an increase of25.4%. However, the characteristics of SMS has also led to a large number of criminal elements and the business use of SMS channels broadcast transmission. Not only caused the majority of ordinary users of living problem, also caused a decline in the quality of service, the high volume of complaints, but some spam has also affected the social stability and national security. According to the China Internet Society survey released by China mobile, phone users received an average of8.29spam messages per week. Therefore, monitoring and blocking spam messages on the research into the current mobile operators a major topic. To solve this problem, this paper proposes the introduction of text classification technology to spam messages in the program monitoring system.This paper studies the issue of the status quo and development trend of mobile operators introduced monitoring and interception technology of common spam messages and project background. Then, the text classification sub-words, feature extraction, and the naive Bayesian classifier basic technical knowledge. Again, according to the actual needs of China Mobile proposed based on mobile phone SMS text classification technology solutions for intelligent classification system and the establishment of system model, and then integrated use of text classification technology, combined with traditional surveillance and interception of SMS technology to achieve efficient, accurate and on mobile phones SMS for intelligent classification. Finally, a comparison of the experimental system, and the system was evaluated.In this paper, based on the actual needs of the traditional SMS integrated monitoring in China Mobile Communications, interception technology advantages and text classification in the classification of technology to improve SMS in the spam messages recognition accuracy, reduce manual for the examination after the SMS monitoring. Unlike the traditional SMS monitoring system, only the spam messages and non-spam messages, according to the actual needs of this article SMS will be classified as category6, meet specific business needs of mobile companies. After testing, content-based SMS intelligent classification system, is able to meet the actual needs of mobile operators.