Garbage short message interception method based on characteristic similarity

Abstract

The present invention provides a garbage short message interception method based on the characteristic similarity, which realizes a garbage short message interception method capable of only determining whether a short message is group sending garbage information by tracking the appearance probability of specific similar characteristic information in short message contents at a period of time, includes two filtering algorithms respectively for two modes of appearing contact information (like a telephone number, a web address and a bank card number) of a communication person in the short message content and not appearing a contact information, intercepts communication contact person content and dynamic random short message content after the pretreatment for converting special characters in short message text, then performs similarity comparison with the contents in database and counting, and regards the short message with an exceeding number of corresponding content as garbage short message, and adds the short message content into garbage short message characteristic information content data, so that the method not only realizes the fuzzy matching text identification, but also possesses the intelligent learning identification garbage information function.

Description

【Technical field】 [0001] The invention relates to a spam short message filtering method, in particular to a spam short message interception method based on feature similarity. 【Background technique】 [0002] With the development of mobile communication technology, the coverage of mobile phone applications is very extensive. Wherein, the use of mobile phones to send short messages has become an important way for people to communicate in daily life. For this reason, some unscrupulous merchants have also begun to use text messages to wantonly spread advertising text messages and reactionary text messages. Diverse forms such as multimedia messages and mobile emails have appeared, and the problem of spam messages has become increasingly serious. In order to avoid the flooding of a large number of spam messages in the system, reduce the interference to the public and reduce the impact on the normal business development of mobile operators, it is urgent to develop a set of relati...

AI Technical Summary

Problems solved by technology

However, when the existing spam filter is designed, the content of the text message is screened and filtered after analyzing the popular spam samples in advance to form spam keywords. Since the keywords are all preset and the system lacks self-learning ability, Therefore, this type of spam filter is less intelligent and cannot respond to the changing spam messages in a timely manner. Spam messages can easily bypass the filter by simply changing the content.

Images