Studying of Classifying Junk Messages Based on The Data Mining

Traditional SMS filters are basically text-based filtering. However, the filtering mechanisms have limitations: first, the keyword can easily be replaced by other symbols, which increase the difficulty of filtering; second, is garbage in the ongoing renovation of keywords, filtering mechanism of the high demands of intelligent learning, which is difficult to achieve. Send acts based on the analysis of the filter will be able to solve the above problems. Filters in the design of messages added to the idea of data mining, through classification analysis of the sent messages, thus can be extracted to identify the rules of spam text messages and can send messages to the new classification behavior.