Filtering method and equipment of short messages

Abstract

The invention discloses a filtering method and equipment of short messages and relates to the technical field of electronic information, which can flexibly deal constantly changed garbage messages, improves the classification precision of a classification model and filtering accuracy on the garbage messages and has no need of additionally establishing a training corpus for the classification model, thereby reducing cost. The filtering method of the short messages, provided in the embodiment of the invention, comprises the following steps of: judging the classes of the received short messages by utilizing the current classification model; when the short messages are garbage messages, filtering the short messages; when the short messages are normal messages, carrying out text finger extraction on the short messages to obtain finger information corresponding to the short messages; and when confirming that the short messages need to be audited according to the finger information and an auditing result is that the short messages are the garbage messages, updating the current classification model by utilizing the short messages, thereby executing next filtering process by utilizing the updated classification model.

Description

technical field [0001] The invention relates to the technical field of electronic information, in particular to a short message filtering method and equipment. Background technique [0002] Short message is one of the ways of information exchange that people often use. At the same time, spam text messages have also begun to flood gradually. Statistics show that among the huge number of short messages, about 30% belong to spam text messages. For ordinary users, spam text messages seriously interfere with daily life; for operators, spam text messages occupy a large amount of traffic space and reduce information transmission efficiency. [0003] Currently the most common spam filtering method is a method based on sensitive words and string matching, which filters spam text messages by arranging sensitive word lists on the server side and pattern matching. However, this method has a high "mistake rate" and cannot exhaust all sensitive word forms by enumerating sensitive words....

AI Technical Summary

Problems solved by technology

[0005] However, the existing spam filtering methods based on classification models also have many shortcomings. For example, mainstream classification models, such as support vector machine models, Bayesian models, and maximum entropy models, often need to build The training corpus is used to obtain the parameters for classification. When the training corpus is small in size, the accuracy of classification is low, and in order to build a training corpus of a certain scale, it needs to spend a lot of money, and the cost is too high; moreover, the sending of spam messages Operators will continue to deform spam messages or add new types of spam messages to avoid the message filtering mechanism. However, the existing spam message filtering method based on the classification model can only filter spam messages that have appeared, but cannot filter never-before-seen spam messages. Spam text messages that have appeared in the past cannot flexibly respond to changes in spam text messages, and the filtering accuracy is low

Images