Junk short message detecting method and device

Abstract

The invention provides a junk short message detecting method and device, relates to the field of information processing, and aims at improving the accuracy and efficiency of junk short message detection. The method comprises that a training sample flow of a present training time window is obtained according to a last short message time window; feature information is extracted from the training sample flow of the present training time window, comprises a short message sending feature and a short message content feature, and includes at least one feature parameter; the training sample flow in the present training time window is trained and classified on the basis of the feature information, and values of the at least one feature parameter corresponding to a normal short message and a junk short message are obtained; and on the basis of the values of the at least one feature parameter, a short message flow in a present short message time window is classified and determined, and the junk short message of the short message flow in the present short message time window is determined.

Description

technical field [0001] The invention relates to the field of information processing, in particular to a spam short message detection method and device. Background technique [0002] With the rapid development of communication technology, the use of mobile phones is becoming more and more extensive. Among them, using mobile phones to send short messages has become an important way for users to communicate. However, the emergence of spam text messages has also caused great troubles to users. Some criminals use text messages to spread advertisements, rumors, or conduct fraud, which seriously affects the lives of users. [0003] At present, there are many detection methods for spam text messages, such as content feature detection based on text messages, such as keyword matching and text classification filtering, etc., and sender detection based on text messages, such as black and white lists and sending frequency filtering, etc. . However, these detection methods have more or ...

AI Technical Summary

Problems solved by technology

However, these detection methods have more or less certain defects, such as the method based on keyword matching, which is related to the preset thesaurus capacity. If the preset thesaurus capacity is large, it will affect the detection performance. Speed, if the preset thesaurus capacity is small, there will be missed detection. Among them, it will take a certain amount of time to preprocess and segment the text message content, and the detection results will also have errors, so that the detection is accurate limited accuracy and detection efficiency

Images