Method and device for filtering short messages

Abstract

The invention discloses a method and a device for filtering short messages. The method comprises the following steps: calculating suspected coefficients with a predetermined number according to incoming and outgoing information for mobile phone numbers; removing ambiguous points of the suspected coefficients, obtained by calculating, converting the ambiguous points to suspected scores; determining that the mobile phone numbers with the suspected scores, greater than a score threshold, are suspected blacklist numbers; when short messages sent by the suspected blacklist numbers are waste short messages, determining the numbers as blacklist numbers and filtering sent short messages; otherwise, permitting sending of the short messages. According to the method and the device disclosed by the invention, by determining the suspected blacklist numbers through the incoming and outgoing information for the mobile phone number sending the short messages, judging whether the short messages sent by the suspected blacklist numbers are the waste short messages or not, and determining the blacklist number and filtering the short message when the waste short messages are determined, the processing efficiency of the waste short messages is improved, the pressure of judging whether all the short messages are the waste short messages or not is reduced, and the processing delay caused by sudden increase of the short messages is avoided.

Description

technical field [0001] The present application relates to the field of information analysis, and in particular, to a method and device for realizing short message filtering. Background technique [0002] Currently, there are two main methods for judging spam text messages. One is to judge by the frequency of sending text messages. That is, if the number of text messages sent by a number within a certain time interval is greater than the threshold set by technicians, the number will be If it is determined as a blacklisted number, the text messages sent by the blacklisted number will be confirmed as spam text messages. The blacklist determined in this way cannot achieve a unified standard for different regions, and can only be used for spam short messages with high frequency. Filtering, for spam messages that do not meet the set frequency, it is impossible to judge spam messages; another way is to judge whether spam messages are or not by keywords, for example, the key words s...

AI Technical Summary

Problems solved by technology

Using the first method to filter spam text messages, it is time-consuming to judge and there are problems of inconsistent standards, and the processing capacity is low; based on the spam text message processing method of the national short message processing center based on distributed stream computing, all text messages are sent to the centralized The processing center judges whether it is a spam text message, which puts a lot of pressure on the centralized processing center. During the busy period of business, the number of text messages surges, which may easily cause delays in the processing of text messages, and the actual filtering effect is not achieved.

Images