Bayes-based Intelligent Classification Spam SMS Interception Platform Build

As the rapid development of mobile communication network, mobile phonetext message is one of the important ways for peoples real-time exchange ofinformation. The number of spam attached to mobile phone text messages is a longtroubled for users. Although SMS spam blocking technology has early results, butonly focused on a single keyword to identify spam messages and filtering the callingnumber in according to the repeat and same SMS contents. Lawless elements,however, won enormous benefits by advertise, fraud or even illegal spam messagesand so on. After statistics found that the spam messages survival and spreadmorphology has undergone tremendous change in order to avoid this real-timestrategy. By the statistical analysis of the recent spam messages, we can updateappropriate keyword combinations and intercept strategy. But it is necessary to paythe enormous human cost, and statistical rules just solving a short time spammessages problem. That is the important reason why no best solution for the spammessages. Therefore, the combination of self-learning intelligent classification systemand real-time SMS spam blocking system has important implications for thelong-term control and identification for filtering spam messages.This article in full theoretical preparations premise, I made a brief exposition forChinese word, text classification and the key statistical learning method. Finally Idecided to use Bayes method as the main algorithm for intelligent classificationlearning. I made a detailed description for preliminary system requirements,improved junk SMS interceptor system analysis and system design. In particular,because of the system functions, system processes and module design, we find thecombination of the Bayes intelligent classification system and real-time application ofthe interception system has the possibility and perfection for realizing. And thesystem will eventually generate system reports for analyzing the follow-up spammessages.Test results show that the Bayes intelligent classification system with thepretreatment of SMS text content and the classification algorithm is very accurate for classification. Compared with the artificial classification test results, intelligentclassification system accuracy is higher than the artificial classification in processinglarge quantities of SMS text.In summary, SMS spam blocking system based on Bayes classification algorithm,improve the system self-learning ability and reduce great dependence on manualoperation. For the future long-term follow-up analysis changes in spam messages, aswell as means for real-time processing of spam messages, the system has a lot ofhelp. The improved junk SMS interceptor system with real-time interception andnon-real-time intelligent analysis interception has double protection. The effectiveintercept filtering will guarantee the SMS spam messages interceptor system on-linestably.