A General Automatic Method For Detecting And Analysing Of Junk Messages And The Preventable Strategy For Telecom Operators

How to find out an effective method to detect and control the junk messages has bothered the telecommunication operators for a long time. On one hand, the junk messages may annoy the freedom of communication for people, especially a part of illegal and baleful messages hugely damaging the spirit and property of people. It leads to a increasingly demand to detect and control junk messages from the perspectives of nation and customers. On the other hand, economically, it is unprofitable for telecommunication operators to conduct this detection and control, and the accuracy of detection and control may negatively affect the satisfaction degrees of customers. Basically, junk messages are difficult to be detected because they possess a great variety of characteristics both in form and feature, and manual or lower automatic methods could cause a risk on the user’s privacy and freedom of communication.Firstly, this thesis introduces serveral popular personal communiations systems, analyzes the weakpoints of the existed mechanisms of processing and intercepting junk messages, and extracts the key characteristics of message dissemenation to identify the disadvantages of current control methods. then, several typical methodologies of deteting junk messages are deeply investigated, including intelligent semantic recognizing(a method used to recognize the semanteme in texts), and a methed used to identify the properties which is base on statistics and pattern recognition toward frequency, breadth and similarity for junk messages. Then this thesis provides a model which take completely into account all above-mentioned factors to analyze and detect junk messages accurately. After that, a scheme integrating the inference of machine intelligence and manual decision is developed.Finally, we discuss the significance of building a general automatic detection system by operator, illustrate the advantage of fulfilling the joint control among various operators, through the automatic detection system, and and summarize how to construct the cloud-based and automic management and control platform to provide full-fledged and multi-function junk message detection and control system.