Research And Realization Of Filtering Machine For Junk SMS Message Based On J2ME

With the swift development of mobile communication technology and rapid increase of mobile telephone popularization, short message has been more and more favored by many people, and it has become a kind of important method of communication and intercommunion in our daily life. However, it makes us in lots of unnecessary trouble because of junk short message. This phenomenon of junk short message not only exists strongly in our country, but also exists in some developed countries, such as America and Europe.In this paper, it researches the realization of filtering machine of junk short message in mobile telephone terminal based on content of short message, and we add Inbox and junk Box in the filtering machine like the form of e-mail. We put forward a new filtering method based on support vector machine, blacklist and white list.The method is as follows: At first, it intercepts the telephone number of short message sender. And then querying the white list, if the telephone number of short message sender is in white list, the short message is considered right message, and the mobile telephone will appear interface and short music to remind mobile telephone user. Otherwise, querying the blacklist, if the telephone number of short message sender is in blacklist, the short message is considered junk message. If the telephone number of short message sender is not in white list and blacklist, then filtrating machine will use support vector machine to categorize the new short message.There are kinds of mobile telephone OS, such as Symbian, Linux, Windows Mobile and Palm. We adopt J2ME technology to realize the filtering machine.Considering the upgrade and maintenance of filtering machine, we adopt MVC design pattern to design architecture and adopt the method of software product line to guide realization of filtering machine. In the filtering machine architecture, we show the commonality and variability particularly, and fix the section of commonality in a long interval. As to the section of variability, we save it with the form of user interface. Consequently, we can reuse the design and component of filtering machine in the future.