WO2016177148A1

WO2016177148A1 - Short message interception method and device

Info

Publication number: WO2016177148A1
Application number: PCT/CN2016/076791
Authority: WO
Inventors: 伏晓海
Original assignee: 中兴通讯股份有限公司
Priority date: 2015-08-18
Filing date: 2016-03-18
Publication date: 2016-11-10
Also published as: CN106470405A

Abstract

Disclosed is a short message interception method. Receiving a short message and acquiring the characteristic vectors of the short message; according to the characteristic vectors of the short message, searching for a short message spam template matching the short message; according to the short message spam template matching the short message and pre-set filtering conditions, using a corresponding interception frequency scheme to intercept the short message. Also disclosed is a short message interception device. The present invention performs interceptions using different schemes for different types of short messages, thereby implementing differentiated short message interception and satisfying the needs of service providers.

Description

SMS interception method and device

Technical field

The present invention relates to the field of communications, and in particular, to a short message intercepting method and apparatus.

Background technique

With the development of communication services, SMS services have won the favor of customers because of their low price, novel form, convenient and fast. However, with the widespread use of SMS services, the problem of spam messages has intensified. A large number of unscrupulous users or merchants use SMS platforms to send advertising messages, fraudulent messages and malicious harassment messages. The proliferation of spam messages not only disrupts the normal communication life of users, but also may cause financial losses to users.

In order to solve the problem that spam messages bring to users, operators generally use spam messages to intercept spam messages. The actual interception requirements of operators for different types of spam messages are different. For example, for badly-spoken text messages such as fraud and rumor communication, it needs to be strictly intercepted; for commercial advertisements or notifications, you can use a wider interception strategy. However, the current spam message interception system has a single configuration, and all the received spam messages are all adopted the same interception strategy, which cannot meet the needs of operators.

Summary of the invention

The main purpose of the embodiment of the present invention is to provide a short message intercepting method and device, which aims to solve the single technical problem of the short message intercepting strategy.

To achieve the above objective, an embodiment of the present invention provides a short message intercepting method, where the short message intercepting method includes the following steps:

Receiving a short message and obtaining a feature vector of the short message;

Searching for a spam template matching the short message according to the feature vector of the short message;

And according to the spam template matched with the short message, and the preset filtering condition, the short message is intercepted by using a corresponding intercept frequency policy.

In the embodiment of the present invention, the preset filtering condition includes: a calling number, a called number, and a sending time of the short message, and the step of receiving the short message and acquiring the feature vector of the short message includes:

Obtain the type of spam template and spam template input by the user;

Intercepting according to the type of the spam template input by the user, the calling number, the called number, and the sending time period Frequency strategy

Obtaining a feature vector of the spam template input by the user;

According to the feature vector of the spam message template, the spam message templates with the same feature vector are filtered out to form a list for searching.

In the embodiment of the present invention, the step of searching for a spam message template matching the short message according to the feature vector of the short message includes:

Finding a list of spam short message templates having the same feature vector according to the feature vector of the short message;

If a spam short message template list having the same feature vector is found, obtaining a spam short message template having the greatest similarity with the short message in the spam short message template list;

Determining whether the maximum similarity meets a threshold;

If the maximum similarity satisfies the threshold, the spam template with the greatest similarity with the short message is obtained as a spam template matching the short message.

In the embodiment of the present invention, after the step of searching for the spam message template matching the short message according to the feature vector of the short message, the method further includes:

If the spam template matching the short message is not found, the default spam template is used to intercept the short message.

In the embodiment of the present invention, the step of intercepting the short message by using a corresponding interception frequency policy according to the spam short message template matched with the short message and the preset filtering condition includes:

Obtaining a type of spam template that matches the short message;

The type of the spam template matched with the short message is used as the type of the short message;

Obtaining a corresponding interception frequency policy according to the type of the short message and the preset filtering condition;

Obtaining, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number;

Determining whether the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy;

If the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy, intercepting the short message;

If the sending frequency of the calling number sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy, Line the message.

In addition, in order to achieve the above object, the present invention further provides a short message intercepting device, where the short message intercepting device includes:

a receiving module, configured to receive a short message, and obtain a feature vector of the short message;

a matching module, configured to search for a spam template matching the short message according to the feature vector of the short message;

The intercepting module is configured to intercept the short message according to the spam template matched with the short message and the preset filtering condition by using a corresponding intercept frequency policy.

In the embodiment of the present invention, the preset filtering condition includes: a calling number of the short message, a called number, and a sending time, and the short message intercepting device further includes:

Obtaining a module, configured to obtain a type of the spam message template and the spam message template input by the user; and set an interception frequency policy according to the type of the spam message template input by the user, the calling number, the called number, and the sending time period;

a calculation module, configured to acquire a feature vector of the spam template input by the user;

The list module is configured to filter out spam templates having the same feature vector according to the feature vector of the spam template, and form a list for searching.

In the embodiment of the present invention, the matching module includes:

a searching unit, configured to search for a spam short message template list having the same feature vector according to the feature vector of the short message;

The similarity calculation unit is configured to obtain a spam short message template having the same similarity as the short message template in the spam short message template list if the spam short message template list having the same feature vector is found;

a determining unit, configured to determine whether the maximum similarity meets a threshold;

The matching unit is configured to acquire a spam template having the greatest similarity with the short message as the spam template matched with the short message if the maximum similarity satisfies the threshold.

In the embodiment of the present invention, the intercepting module is further configured to

In the embodiment of the present invention, the intercepting module includes:

a type unit, configured to obtain a type of a spam template that matches the short message; and use a type of the spam template that matches the short message as a type of the short message;

The policy unit is configured to obtain a corresponding interception frequency policy according to the type of the short message and a preset filtering condition;

The frequency unit is configured to obtain, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number;

a determining unit, configured to determine whether the sending frequency of the type of short message sent by the calling number exceeds an intercepting frequency of the intercepting frequency policy;

The intercepting unit is configured to intercept the short message if the sending frequency of the type of short message sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy;

The release unit is configured to release the short message if the sending frequency of the type of the short message sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy.

In the embodiment of the present invention, a computer storage medium is further provided, and the computer storage medium may store an execution instruction, where the execution instruction is used to execute the short message interception method in the foregoing embodiment.

The method and device for intercepting a short message according to an embodiment of the present invention obtains a short message by a carrier server, acquires a feature vector of the short message, searches for a spam message template matching the short message according to the feature vector of the short message, and then, according to the garbage matched with the short message SMS templates and filtering conditions, using the corresponding interception frequency strategy to intercept SMS messages. The embodiments of the present invention perform different types of short message interception processing on different types of short messages, and implement differentiated interception of short messages to meet the needs of operators.

DRAWINGS

1 is a schematic flowchart of a first embodiment of a short message interception method according to the present invention;

2 is a schematic flowchart of a second embodiment of a short message interception method according to the present invention;

3 is a schematic flowchart of a third embodiment of a short message interception method according to the present invention;

4 is a schematic flowchart of a fourth embodiment of a short message interception method according to the present invention;

FIG. 5 is a schematic flowchart diagram of a fifth embodiment of a short message interception method according to the present invention; FIG.

6 is a schematic diagram of functional modules of a first embodiment of a short message intercepting apparatus according to the present invention;

7 is a schematic diagram of functional modules of a second embodiment of a short message intercepting apparatus according to the present invention;

8 is a schematic diagram of functional modules of a third embodiment of a short message intercepting apparatus according to the present invention;

FIG. 9 is a schematic diagram of functional modules of a fifth embodiment of a short message intercepting apparatus according to the present invention.

The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.

detailed description

It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The main solution of the embodiment of the present invention is: receiving a short message, and acquiring a feature vector of the short message; searching for a spam message template matching the short message according to the feature vector of the short message; and spam message matching the short message The template, and the preset filtering conditions, use the corresponding intercept frequency policy to intercept the short message.

Because the prior art spam message interception system has a single configuration, the same interception strategy is adopted for all received spam messages, and therefore, the operator's different interception requirements for different types of spam messages cannot be met.

The present invention provides a solution, after receiving a short message, according to the type of the short message and the filtering condition, according to the corresponding interception strategy, the corresponding operation of the short message is intercepted, which satisfies the operator's interception requirement for the spam short message differentiation.

Referring to FIG. 1 , a first embodiment of the short message interception method of the present invention provides a short message intercepting method, where the short message intercepting method includes:

Step S10: Receive a short message, and acquire a feature vector of the short message.

The solution in this embodiment is mainly applied to an operator server, and can of course be applied to other servers having short message interception requirements. This embodiment is exemplified by an operator server.

Specifically, as an implementation manner, first, the operator server receives the short message, obtains the text content of the short message, and the related information of the short message, for example, the calling number, the called number, and the sending time of the short message.

Then, the feature vector of the short message is obtained, so as to obtain the spam template matching the short message.

Specifically, when acquiring the feature vector of the short message, the following scheme may be adopted:

First, the operator server removes the noise of the text message, that is, removes the interference words in the text of the message, such as punctuation.

Then, the obtained noise-removed short message text is subjected to word segmentation processing, the short message text is split into individual word segments, and the stop words in the obtained phrase are removed. The removed stop words include: adverbs, prepositions, modal particles, conjunctions, etc. These words usually have no clear meaning in themselves, and only have to be put into a complete sentence, such as common "of" and "at".

Then, the obtained individual word segments are sorted according to the encoding order of the characters to obtain the feature vector of the short message. In this embodiment, the UTF-8 encoding (8-bit Unicode Transformation Format, Unicode) is used to convert the obtained single word segment into UTF-8 encoding, and UTF-8 encoding for all word segments is obtained; according to UTF-8 The order of the encoding, sorting all the obtained UTF-8 codes to obtain the feature vector of the short message. Of course, other codes can also be used for sorting, which can be flexibly set according to actual needs.

Step S20: Search for a spam message template matching the short message according to the feature vector of the short message.

In this embodiment, the operator server is pre-configured with different types of spam templates, and the type of spam messages is determined according to the similarity with the spam templates. Then, according to different types of spam messages, combined with filtering conditions such as calling number and being Call the number and configure different interception strategies to meet the requirements of the operator's differentiated interception of SMS.

After obtaining the feature vector of the short message, the operator server searches for the spam short message template with the same feature vector in the memory by using the feature vector of the short message as a keyword, and calculates the similarity between the found spam short message template and the short message.

Then, it is determined whether the calculated similarity satisfies a preset threshold: if the obtained similarity satisfies a preset threshold, the found garbage short message template matches the short message, and the spam template is obtained.

If a plurality of spam templates having the same feature vector as the short message are found, the similarity between all the found spam templates and the short messages is calculated, and the spam template with the greatest similarity with the short message is obtained. Then, whether the maximum similarity is determined is The preset threshold is met: if the maximum similarity satisfies the preset threshold, it is determined that the spam template matches the short message.

Step S30: Perform interception processing on the short message according to the spam short message template matched with the short message and the preset filtering condition by using a corresponding intercept frequency policy.

After obtaining the spam template matching the short message, the type of the spam template is the type of the short message.

Then, according to the type of the short message and the preset filtering condition, find and obtain the corresponding interception frequency policy. The interception frequency policy in this embodiment refers to that the same calling user intercepts the short message after the frequency of sending the same type of short message reaches a set threshold within a certain period of time. Therefore, different interception frequencies can be set for different types of short messages. For example, fraudulent SMS can set the threshold of the interception frequency to a lower level, while the commercial advertisement and notification type SMS can set the threshold of the interception frequency to a higher level.

It should be noted that the preset filtering condition may be: the calling number of the short message, the called number, or the calling number, the called number, and the sending time of the short message, which can be flexibly set according to actual needs.

Then, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number is obtained.

Then, it is judged whether the frequency of sending the short message sent by the calling number exceeds the interception frequency threshold corresponding to the short message: if it exceeds, the short message is intercepted; if not, the short message is released.

In this embodiment, the operator server receives the short message, obtains the feature vector of the short message, searches for the spam short message template matching the short message according to the feature vector of the short message, and then, according to the spam message template matched with the short message and the preset filtering condition, The corresponding interception frequency strategy is used to intercept the short message, so that different types of short messages are intercepted by different strategies, and the differential interception of the short message is realized, which satisfies the needs of the operator users.

Further, referring to FIG. 2, the second embodiment of the short message interception method of the present invention provides a short message interception method. Based on the embodiment shown in FIG. 1, the preset filtering conditions include: a calling number of the short message, and a called party. The number and the sending time, before the step S10, further include:

Step S40: Obtain a type of the spam template and the spam template input by the user.

The operator user can input the spam template according to the short message that needs to be intercepted, and set the type of each spam template that is input. It should be noted that the type of the spam short message template can be set by the user, or the preset type can be selected by the user, and can be flexibly set according to actual needs.

The carrier server obtains the spam template input by the operator user and the type of spam template.

Step S50: Set an interception frequency policy according to the type of the spam message template input by the user, the calling number, the called number, and the sending time period.

After the operator server obtains the spam template input by the operator user, a different interception policy is set for different spam template types. The interception frequency policy for the spam template can be a preset interception frequency policy, or an interception frequency policy set by the operator according to actual needs, and can be flexibly set according to actual needs.

Specifically, as an implementation manner, an example of the interception frequency policy set by the operator user is obtained, according to the type of the spam short message template, different user groups to which the calling number belongs, different user groups to which the called number belongs, and different times of sending Segment, the operator user can set the corresponding interception frequency policy.

Thus, each interception frequency strategy obtained includes the following information:

Type of text message;

The user group to which the calling number belongs;

The user group to which the called number belongs;

The time period sent.

The obtained interception frequency policy may be a policy tree, that is, each spam message template, which is corresponding according to the type of the spam message template, the user group to which the calling number belongs, the user group to which the called number belongs, and the time period to be sent. Intercept frequency strategy.

The interception frequency policy is: the same calling user intercepts the short message after the frequency of sending the same type of short message reaches a set threshold within a certain period of time. For example, fraudulent SMS messages must be strictly intercepted, and the threshold for intercepting frequencies can be set lower; commercial advertising messages can relax the interception conditions, and the threshold for intercepting frequencies can be set higher. Of course, the interception frequency strategy can also be other strategies, which can be flexibly set according to actual needs.

The user group to which the calling number belongs includes the local user group, the external network user group, the local user group, the foreign user group, the different number segment group, and the SP (Service Provider) user group, or other user groups. , can be flexibly set according to actual needs.

The user group to which the called number belongs includes the local user group, the external network user group, the local user group, the foreign user group, the different number segment group, and the SP user group, or other user groups, which can be flexibly set according to actual needs.

The sending time period can be in hours or in days, or other time units can be used, which can be flexibly set according to actual needs.

Step S60: Acquire a feature vector of the spam template input by the user.

After obtaining the spam template input by the operator user, the carrier server obtains the feature vector of the spam template, so as to filter the spam template matching the short message.

Specifically, as an implementation manner, first, the noise of the spam short message template text is removed, that is, the interference words in the short message text, such as punctuation marks, are removed.

Then, the obtained noise-removed spam template text is subjected to word segmentation processing, the spam message template text is split into individual word segments, and the stop words in the obtained phrase are removed. The removed stop words include: adverbs, prepositions, modal particles, conjunctions, and so on.

Then, the obtained individual word segments are sorted according to the encoding order of the characters to obtain the feature vector of the spam short message template. In this embodiment, UTF-8 encoding is used, and the obtained single word segment is converted into UTF-8 encoding, and UTF-8 encoding for characterizing all word segments is obtained; all UTF-8s obtained in the order of UTF-8 encoding are obtained. The code is sorted to obtain the feature vector of the spam template.

Step S70: Filter out spam templates with the same feature vector according to the feature vector of the spam template, and form a list for searching.

After the operator server obtains the feature vectors of all the spam templates, all the spam templates are saved by using the feature vector of the spam template as a keyword.

When the spam template is saved, the spam templates with the same feature vector are saved in the same list, and a list of spam templates with the same feature vector is obtained, and the feature vector is saved as a keyword.

Thus, a list of spam templates is obtained, which can be used for subsequent searching for spam templates matching the received messages. use.

In this embodiment, the operator server first obtains the type of the spam template and the spam template input by the user, and then sets the corresponding interception frequency for the type of the spam template obtained, the calling number, the called number, and the sending time period. Strategy, get differentiated interception frequency strategy. The operator server obtains the feature vector of the spam template input by the user, and obtains a spam short message template list with the same feature vector for subsequent searching and matching the spam message template that matches the received short message. In this embodiment, the operator user can set different interception frequency policies for different spam short message templates according to actual needs, so that when the operator server receives the short message, the differentiated interception frequency policy can be used to perform the received short message. The interception process satisfies the interception requirements of operators for different SMS messages.

Further, referring to FIG. 3, the third embodiment of the short message interception method of the present invention provides a short message interception method. Based on the embodiment shown in FIG. 2, the step S20 includes:

Step S21: Search for a spam short message template list having the same feature vector according to the feature vector of the short message.

After obtaining the feature vector of the short message, the feature vector of the short message is used as a keyword to search for a list of spam templates having the same feature vector in the memory. The spam template list is saved by using the feature vector as a keyword.

Step S22: If a spam short message template list having the same feature vector is found, obtain a spam short message template with the greatest similarity with the short message in the spam short message template list.

After finding the list of spam templates with the same feature vector as the short message, the similarity between all spam templates and the short messages in the list is calculated. This example uses the minimum edit algorithm (Minimum edit distance) to calculate the similarity between the spam SMS template and the short message. Specifically, the short message is the string a, the spam template is the string b, and then:

Calculate the number of edits that need to pass from the string a to the string b. Each edit can only add, delete, or replace one character, and get the minimum number of edits c from the string a to the string b, that is, from the string a. The minimum edit distance to the string b: c.

Then, the length of the string a is L, and the similarity between the spam template and the short message is d, then:

Thereby, the similarity between the spam template and the short message is obtained.

Then, the spam template with the greatest similarity with the short message is obtained, and the maximum similarity is obtained.

Step S23: Determine whether the maximum similarity meets a threshold.

After obtaining the maximum similarity between the spam short message template and the short message, it is determined whether the obtained maximum similarity satisfies a preset threshold.

Specifically, if the maximum similarity is greater than or equal to the preset threshold, determining that the maximum similarity meets the preset threshold;

If the maximum similarity is less than the preset threshold, it is determined that the maximum similarity does not satisfy the preset threshold.

Step S24: If the maximum similarity satisfies the threshold, obtain a spam template with the greatest similarity with the short message as a spam template matching the short message.

If the maximum similarity between the spam short message template and the short message meets the preset threshold, the short message is determined to be a spam message.

Then, the spam template with the greatest similarity with the short message is obtained as a spam template matching the short message.

Thereby, a spam template matching the short message is obtained.

In this embodiment, the operator server first obtains the same spam template list as the received SMS feature vector, and then calculates the similarity between each spam template and the short message in the spam template list, and obtains the greatest similarity with the short message. The spam template, and the maximum similarity; determining whether the obtained maximum similarity satisfies the threshold; when the maximum similarity satisfies the threshold, obtaining the spam template with the greatest similarity with the short message as the spam template matching the short message. In this embodiment, the spam short message template that is most similar to the received short message template in the spam short message template list is obtained by determining the maximum similarity and the threshold value, which provides a basis for obtaining the most matching interception frequency strategy.

Further, referring to FIG. 4, a fourth embodiment of the short message interception method of the present invention provides a short message interception method, which is based on any of the embodiments shown in FIG. 1 to FIG. 3 (the present embodiment is illustrated by using FIG. 1). After step S20, the method further includes:

Step S80: If the spam template matching the short message is not found, the default spam template is used to intercept the short message.

When the operator server searches for the spam template with the same feature vector according to the feature vector of the short message, if it is not found, it is deemed that the spam template matching the short message is not found;

If a plurality of spam templates having the same feature vector as the short message are found, the similarity between all the found spam templates and the short messages is calculated, and the spam template with the greatest similarity with the short message is obtained. Then, it is determined whether the maximum similarity is satisfied. Pre-set threshold: If the maximum similarity does not meet the preset threshold, it is considered that the spam template matching the short message is not found.

When the operator server does not find the spam template matching the short message, the default spam template is used to intercept the short message.

Specifically, as an implementation manner, the business server may use a default spam template and a preset filter bar. , find and obtain the corresponding interception frequency strategy, intercept the SMS.

As another implementation manner, the operator server may adopt a default spam template to obtain a default interception frequency policy and intercept the short message.

In this embodiment, when the operator server does not find the spam template that matches the short message, the default spam template is used to intercept the short message, so that the operator server can intercept the received short message, and the pair is implemented. The differentiated interception processing of SMS meets the needs of operators.

Further, referring to FIG. 5, a fifth embodiment of the short message intercepting method of the present invention provides a short message intercepting method, which is based on any of the embodiments shown in FIG. 1 to FIG. 3 (the present embodiment is illustrated by using FIG. 1). Step S30 includes:

Step S31: Obtain a type of the spam short message template that matches the short message.

This embodiment can be implemented in combination with any of the embodiments shown in FIG. 1 to FIG. 3; the embodiment is implemented in combination with the embodiment shown in FIG.

In this embodiment, the operator server compares the similarity between the short message and the spam short message template to find a spam short message template that has the highest similarity with the short message and exceeds the set similarity threshold, and the spam template is used. Type as the type of the text message. Then, according to different types of short messages, combined with preset filtering conditions such as calling number and called number, the corresponding intercepting strategy is searched to realize differential interception of spam messages.

Specifically, first, after the operator server obtains the spam template matched with the short message, the type of the spam template is obtained. It should be noted that the type of the spam template is preset by the operator server.

Step S32: The type of the spam message template matching the short message is used as the type of the short message.

After obtaining the type of spam template matching the short message, the short message is regarded as the spam message of the spam template type.

Then, use the type of spam template as the type of SMS.

Step S33: Acquire a corresponding interception frequency policy according to the type of the short message and the preset filtering condition.

After the carrier server obtains the type of the short message, the corresponding interception frequency policy is obtained according to the type of the short message and the preset filtering condition.

Specifically, as an implementation manner, taking the preset filtering condition as the calling number, the called number, and the sending time as an example, the operator server sequentially according to the type of the short message, the calling number of the short message, the called number of the short message, and The sending time of the short message, find the interception frequency strategy. The detailed search process is as follows:

First, according to the type of the short message, all the interception frequency policies corresponding to the type are obtained;

Then, according to the calling number of the short message, all the interception frequency policies corresponding to the user group to which the calling number belongs are obtained in all the interception frequency policies corresponding to the type of the short message;

Then, according to the called number of the short message, the interception frequency policy corresponding to the user group to which the called number belongs is obtained in the total interception frequency policy corresponding to the type of the short message and the user group to which the calling number belongs;

Then, according to the sending time of the short message, the interception frequency policy corresponding to the time zone to which the sending time belongs is obtained in the total interception frequency policy of the type of the short message, the user group to which the calling number belongs, and the user group to which the called number belongs.

If the interception frequency policy of the short message is not found, the default intercept frequency policy is obtained.

Thus, the interception frequency strategy of the short message is obtained.

Step S34: Acquire, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number.

After the operator server obtains the interception frequency policy of the short message, the frequency of the short message sent by the calling number is obtained according to the type of the short message and the calling number.

Specifically, as an implementation manner, the sending frequency of the short message is saved by using a sliding window. In this embodiment, only the sending frequency of the short message is sent by the calling number within a certain period of time. The size of the sliding window can be flexibly set by the user according to actual needs.

Therefore, after the stored calling number is sent to send the frequency of sending the type of short message, the obtained frequency is incremented by one, and the sending frequency of the short message is sent by the calling number.

Step S35: Determine whether the sending frequency of the type of short message sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy.

The interception strategy of obtaining the short message at the operator server, according to the type of the short message and the calling number, the calling number is sent to the sending frequency of the short message, and it is determined whether the sending frequency of the calling number exceeds the intercepting frequency policy. The frequency threshold is intercepted, that is, whether the short message is reached or not.

Specifically, if the sending frequency of the calling number sent by the calling number is greater than or equal to the intercepting frequency threshold of the intercepting frequency policy, determining that the sending frequency of the calling number of the calling number exceeds the intercepting frequency threshold of the intercepting frequency policy, the short message Reach the conditions of being intercepted;

If the sending frequency of the calling number sent by the calling number is less than the intercepting frequency threshold of the intercepting frequency policy, determining that the sending frequency of the sending type of the calling number does not exceed the intercepting frequency threshold of the intercepting frequency policy, the short message does not reach the intercepted conditions of.

Step S36: If the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy, the short message is intercepted.

If the calling number of the calling number is more than the interception frequency threshold of the interception frequency policy, and the short message has reached the intercepted condition, the control intercepts the short message.

Step S37: If the sending frequency of the calling number sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy, the short message is released.

If the sending frequency of the calling number sent by the calling number does not exceed the intercept frequency threshold of the interception frequency policy, and the short message does not reach the intercepted condition, the message is controlled to be released.

In this embodiment, the operator server first uses the type of spam template that matches the short message as the type of the short message; according to the type of the short message and the preset filtering condition, the corresponding interception frequency policy is obtained; and then according to the type of the short message and the calling party The number obtains the sending frequency of the type of short message sent by the calling number, and determines whether the sending frequency of the sending type of the calling number exceeds the intercepting frequency of the intercepting frequency policy: if it exceeds, the short message is intercepted; if not, the short message is released. This embodiment implements different interception frequency policies for different types of short messages and filtering conditions, and performs differentiated interception to meet the needs of the operator users.

The present invention further provides a short message intercepting device. Referring to FIG. 6, the first embodiment of the short message intercepting device of the present invention provides a short message intercepting device, and the short message intercepting device includes:

The receiving module 100 is configured to receive a short message and obtain a feature vector of the short message.

The solution in this embodiment is mainly applied to an operator server, and can of course be applied to other servers having short message interception requirements.

Specifically, as an implementation manner, first, the receiving module 100 receives the short message, obtains the text content of the short message, and the related information of the short message, for example, the calling number, the called number, and the sending time of the short message.

First, the receiving module 100 removes the noise of the short text, that is, removes the interference words in the short text, such as punctuation.

Then, the receiving module 100 performs word segmentation on the obtained noise-removed short message text, splits the short message text into individual word segments, and removes the stop words in the obtained phrase. The stop words removed include: adverbs, prepositions, modal particles, Conjunctions, etc., these words usually have no clear meaning in themselves, and only have to be put into a complete sentence, such as the common "" and "at".

Then, the receiving module 100 sorts the obtained individual word segments according to the encoding order of the characters to obtain the feature vector of the short message. In this embodiment, UTF-8 encoding is used, and the obtained single word segment is converted into UTF-8 encoding, and UTF-8 encoding for characterizing all word segments is obtained; all UTF-8s obtained in the order of UTF-8 encoding are obtained. The code is sorted to obtain the feature vector of the short message. Of course, other codes can also be used for sorting, which can be flexibly set according to actual needs.

The matching module 200 is configured to search for a spam template matching the short message according to the feature vector of the short message.

After the receiving module 100 obtains the feature vector of the short message, the matching module 200 searches for the spam short message template with the same feature vector in the memory by using the feature vector of the short message as a keyword, and calculates the similarity between the found spam short message template and the short message.

Then, the matching module 200 determines whether the calculated similarity meets the preset threshold: if the obtained similarity meets the preset threshold, the found junk short message template matches the short message, and the spam template is obtained.

If the matching module 200 finds a plurality of spam templates having the same feature vector as the short message, respectively calculating the similarity between all the spam templates and the short messages, and obtaining the spam template with the greatest similarity with the short message; Whether the similarity satisfies the preset threshold: if the maximum similarity satisfies the preset threshold, it is determined that the spam template matches the short message.

The intercepting module 300 is configured to intercept the short message according to the spam template matched with the short message and the preset filtering condition by using a corresponding intercept frequency policy.

After the matching module 200 obtains the spam template that matches the short message, the intercepting module 300 uses the type of the spam template as the type of the short message.

Then, the intercepting module 300 searches for and acquires a corresponding interception frequency policy according to the type of the short message and the preset filtering condition. The interception frequency policy in this embodiment refers to that the same calling user intercepts the short message after the frequency of sending the same type of short message reaches a set threshold within a certain period of time. Therefore, different interception frequencies can be set for different types of short messages. For example, fraudulent SMS can set the threshold of the interception frequency to a lower level, while the commercial advertisement and notification type SMS can set the threshold of the interception frequency to a higher level.

Then, the intercepting module 300 obtains the calling number and sends the short message according to the type of the short message and the calling number. Send frequency.

Then, the intercepting module 300 determines whether the sending frequency of the type of the short message sent by the calling number exceeds the intercepting frequency threshold corresponding to the short message: if it exceeds, the short message is intercepted; if not, the short message is released.

In this embodiment, the receiving module 100 receives the short message and obtains the feature vector of the short message, and the matching module 200 searches for the spam short message template that matches the short message according to the feature vector of the short message, and then the intercepting module 300 according to the spam template matched with the short message. The preset filtering conditions use the corresponding interception frequency strategy to intercept the short message, so that different types of short messages are intercepted by different strategies, and the differential interception of the short message is realized, which satisfies the needs of the operator users.

Further, referring to FIG. 7, the second embodiment of the short message intercepting apparatus of the present invention provides a short message intercepting apparatus. Based on the embodiment shown in FIG. 6, the preset filtering condition includes: a calling number of the short message, and a called party. The number and the sending time, the short message intercepting device further includes:

The obtaining module 400 is configured to obtain a type of the spam message template and the spam message template input by the user; and set an interception frequency policy according to the type of the spam message template input by the user, the calling number, the called number, and the sending time period.

The obtaining module 400 first obtains the spam template input by the operator user, and the type of the spam template.

The acquisition module 400 then sets a differentiated interception policy for different spam template types. The interception frequency policy for the spam template can be a preset interception frequency policy, or an interception frequency policy set by the operator according to actual needs, and can be flexibly set according to actual needs.

Thus, each interception frequency policy obtained by the acquisition module 400 includes the following information:

Type of text message;

The user group to which the calling number belongs;

The user group to which the called number belongs;

The time period sent.

The interception frequency policy obtained by the obtaining module 400 may be a policy tree, that is, each spam short message template, according to the type of the spam short message template, the user group to which the calling number belongs, the user group to which the called number belongs, and the time period to be sent. Have Corresponding interception frequency strategy.

The calculating module 500 is configured to acquire a feature vector of the spam template input by the user.

After the obtaining module 400 obtains the spam short message template input by the operator user, the calculating module 500 obtains the feature vector of the spam short message template, so as to subsequently filter the spam short message template matched with the short message.

Specifically, as an implementation manner, first, the calculation module 500 removes the noise of the spam short message template text, that is, removes the interference words in the short message text, such as punctuation marks.

Then, the calculation module 500 performs word segmentation on the obtained noise-removed spam template text, splits the spam template template text into individual word segments, and removes the stop words in the obtained phrase. The removed stop words include: adverbs, prepositions, modal particles, conjunctions, and so on.

Then, the calculation module 500 sorts the obtained individual word segments according to the coding order of the characters, and obtains the feature vector of the spam short message template. In this embodiment, UTF-8 encoding is used, and the obtained single word segment is converted into UTF-8 encoding, and UTF-8 encoding for characterizing all word segments is obtained; all UTF-8s obtained in the order of UTF-8 encoding are obtained. The code is sorted to obtain the feature vector of the spam template.

The list module 600 is configured to filter out spam templates having the same feature vector according to the feature vector of the spam template, and form a list for searching.

After the calculation module 500 obtains the feature vectors of all the spam templates, the list module 600 saves all the spam templates by using the feature vectors of the spam templates as keywords.

When the spam template is saved, the list module 600 saves the spam templates with the same feature vector to the same list, and obtains a list of spam templates with the same feature vector, and saves the feature vector as a keyword.

Thus, the list module 600 obtains a list of spam short message templates, which can be used for subsequent searching for spam templates matching the received short messages.

In this embodiment, the obtaining module 400 first acquires the type of the spam message template and the spam message template input by the user, and then sets the corresponding interception frequency for the type of the spam message template, the calling number, the called number, and the sending time period. Strategy, get differentiated interception frequency strategy. The calculation module 500 obtains the feature vector of the spam template input by the user, and the list module 600 obtains the spam message template list with the same feature vector for subsequent use to search for the spam message template matched with the received message. In this embodiment, the operator user can set different interception frequency policies for different spam short message templates according to actual needs, so that when the operator server receives the short message, the differentiated interception frequency policy can be used to perform the received short message. The interception process satisfies the interception requirements of operators for different SMS messages.

Further, referring to FIG. 8, the third embodiment of the short message intercepting apparatus of the present invention provides a short message intercepting apparatus. Based on the embodiment shown in FIG. 7, the matching module 200 includes:

The searching unit 210 is configured to search for a spam short message template list having the same feature vector according to the feature vector of the short message.

After the receiving module 100 obtains the feature vector of the short message, the searching unit 210 searches for the spam short message template list having the same feature vector in the memory by using the feature vector of the short message as a key. The spam template list is saved by using the feature vector as a keyword.

The similarity calculation unit 220 is configured to obtain a spam message template having the same similarity as the short message template in the junk message template list if the junk message template list having the same feature vector is found.

After the searching unit 210 finds the spam short message template list having the same feature vector as the short message, the similarity calculating unit 220 calculates the similarity of all the spam short message templates and the short message in the list. This example uses the minimum edit algorithm (Minimum edit distance) to calculate the similarity between the spam SMS template and the short message. Specifically, the short message is the string a, the spam template is the string b, and then:

Then, the similarity calculation unit 220 acquires the spam template with the greatest similarity with the short message, and obtains the maximum similarity.

The determining unit 230 is configured to determine whether the maximum similarity satisfies a threshold.

After the similarity calculation unit 220 obtains the maximum similarity between the spam short message template and the short message, the determining unit 230 determines whether the obtained maximum similarity satisfies a preset threshold.

Specifically, if the maximum similarity is greater than or equal to the preset threshold, the determining unit 230 determines that the maximum similarity meets the preset threshold;

If the maximum similarity is less than the preset threshold, the determining unit 230 determines that the maximum similarity does not satisfy the preset threshold.

The matching unit 240 is configured to acquire a spam short message template having the greatest similarity with the short message as the spam short message template matching the short message if the maximum similarity satisfies the threshold.

Then, the matching unit 240 obtains the spam template with the greatest similarity with the short message as the spam template matching the short message.

Thereby, the matching unit 240 obtains a spam template that matches the short message.

In this embodiment, the search unit 210 first obtains the same spam message template list as the received short message feature vector, and then the similarity calculation unit 220 calculates the similarity between each spam short message template and the short message in the spam short message template list, and obtains the short message. The spam template with the greatest similarity, and the maximum similarity; the determining unit 230 determines whether the obtained maximum similarity satisfies the threshold; when the maximum similarity satisfies the threshold, the matching unit 240 obtains the spam template with the greatest similarity with the short message. As a spam template that matches the text message. In this embodiment, the spam short message template that is most similar to the received short message template in the spam short message template list is obtained by determining the maximum similarity and the threshold value, which provides a basis for obtaining the most matching interception frequency strategy.

Further, referring to FIG. 6, the fourth embodiment of the short message interception method of the present invention provides a short message interception method, which is based on any of the embodiments shown in FIG. 6 to FIG. 8 (the present embodiment is illustrated by using FIG. 6). The intercepting module 300 is also configured to

When the matching module 200 searches for the spam short message template having the same feature vector according to the feature vector of the short message, if not found, the garbage message template matching the short message is not found;

If the matching module 200 finds a plurality of spam templates having the same feature vector as the short message, respectively, all the calculations are calculated. The similarity between the spam template and the short message is obtained, and the spam template with the greatest similarity with the short message is obtained; then, it is determined whether the maximum similarity satisfies the preset threshold: if the maximum similarity does not satisfy the preset threshold, it is regarded as No spam template matching the text message was found.

When the matching module 200 does not find the spam template that matches the short message, the intercepting module 300 intercepts the short message by using the default spam template.

Specifically, as an implementation manner, the intercepting module 300 can find and obtain a corresponding interception frequency policy according to a default spam template and a preset filtering condition, and intercept the short message.

As another implementation manner, the intercepting module 300 may adopt a default spam template to obtain a default intercept frequency policy and intercept the short message.

In this embodiment, when the matching module 200 does not find the spam template that matches the short message, the intercepting module 300 uses the default spam template to intercept the short message, so that the operator server can intercept the received short message. The differentiated interception processing of the short message is realized, which satisfies the needs of the operator user.

Further, referring to FIG. 9, the fourth embodiment of the short message interception method of the present invention provides a short message interception method, which is based on any of the embodiments shown in FIG. 6 to FIG. 8 (the present embodiment is illustrated by using FIG. 6). The intercepting module 300 includes:

The type unit 310 is configured to obtain a type of the spam template that matches the short message, and use the type of the spam template that matches the short message as the type of the short message.

This embodiment can be implemented in combination with any of the above embodiments shown in FIG. 6 to FIG. 8; this embodiment is implemented in combination with the embodiment shown in FIG.

In this embodiment, by comparing the similarity between the short message and the spam short message template, a spam short message template with the greatest similarity to the short message and exceeding the set similarity threshold is found, and the type of the spam short message template is used as the The type of SMS. Then, according to different types of short messages, combined with preset filtering conditions such as calling number and called number, the corresponding intercepting strategy is searched to realize differential interception of spam messages.

Specifically, first, after the matching module 200 obtains the spam template that matches the short message, the type unit 310 obtains the type of the spam template, and regards the short message as the spam message of the spam template type. It should be noted that the type of spam template is preset.

Then, the type unit 310 uses the type of the spam template as the type of the short message.

The policy unit 320 is configured to obtain a corresponding interception frequency policy according to the type of the short message and the preset filtering condition.

After the type unit 310 obtains the type of the short message, the policy unit 320 acquires the corresponding intercept frequency policy according to the type of the short message and the preset filtering condition.

Specifically, as an implementation manner, taking the preset filtering condition as the calling number, the called number, and the sending time as an example, the policy unit 320 sequentially according to the type of the short message, the calling number of the short message, the called number of the short message, and The sending time of the short message, find the interception frequency strategy. The detailed search process is as follows:

Thus, the policy unit 320 obtains the interception frequency policy of the short message.

The frequency unit 330 is configured to acquire, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number.

After the policy unit 320 obtains the interception frequency policy of the short message, the frequency unit 330 obtains the transmission frequency of the short message sent by the calling number according to the type of the short message and the calling number.

Therefore, the frequency unit 330 obtains the stored frequency of the calling number and transmits the type of the short message, and adds 1 to the obtained frequency to obtain the sending frequency of the short message sent by the calling number.

The determining unit 340 is configured to determine whether the sending frequency of the calling number of the calling number exceeds the intercepting frequency of the intercepting frequency policy.

After the interception policy and the frequency unit 330 of the short message obtained by the policy unit 320 obtains the transmission frequency of the short message according to the type of the short message and the calling number, the determining unit 340 determines that the calling number sends the frequency of sending the short message of the type. Whether the interception frequency threshold of the interception frequency policy is exceeded, that is, whether the short message is reached or not is determined.

The intercepting unit 350 is configured to intercept the short message if the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy.

If the calling number of the calling number is more than the intercepting frequency threshold of the interception frequency policy, and the short message has reached the intercepted condition, the intercepting unit 350 controls to intercept the short message.

The releasing unit 360 is configured to release the short message if the sending frequency of the calling number sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy.

If the sending frequency of the calling number sent by the calling number does not exceed the intercept frequency threshold of the interception frequency policy, and the short message does not reach the intercepted condition, the release unit 360 controls the release of the short message.

In this embodiment, the type unit 310 first uses the spam template type matching the short message as the type of the short message; the policy unit 320 acquires the corresponding intercept frequency policy according to the type of the short message and the filtering condition; and then the frequency unit 330 according to the type of the short message. And the calling number obtains the sending frequency of the calling number to send the short message; the determining unit 340 determines whether the sending frequency of the calling number of the calling number exceeds the intercepting frequency of the intercepting frequency policy: if exceeded, the intercepting unit 350 intercepts the short message. If not exceeded, the release unit 360 releases the short message. This embodiment implements different interception frequency policies for different types of short messages and filtering conditions, and performs differentiated interception to meet the needs of the operator users.

The above are only the preferred embodiments of the present invention, and are not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the present invention and the drawings are directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of the present invention.

Industrial applicability

Claims

A short message intercepting method, the short message intercepting method includes the following steps:

Receiving a short message and obtaining a feature vector of the short message;

Searching for a spam template matching the short message according to the feature vector of the short message;

And according to the spam template matched with the short message, and the preset filtering condition, the short message is intercepted by using a corresponding intercept frequency policy.
The method for intercepting a short message according to claim 1, wherein the preset filtering condition comprises: a calling number of the short message, a called number, and a sending time, and the step of receiving the short message and acquiring the feature vector of the short message Previously, including:

Obtain the type of spam template and spam template input by the user;

Setting an interception frequency policy according to the type of the spam template input by the user, the calling number, the called number, and the sending time period;

Obtaining a feature vector of the spam template input by the user;

According to the feature vector of the spam message template, the spam message templates with the same feature vector are filtered out to form a list for searching.
The method for intercepting a short message according to claim 2, wherein the step of searching for a spam template matching the short message according to the feature vector of the short message comprises:

Finding a list of spam short message templates having the same feature vector according to the feature vector of the short message;

If a spam short message template list having the same feature vector is found, obtaining a spam short message template having the greatest similarity with the short message in the spam short message template list;

Determining whether the maximum similarity meets a threshold;

If the maximum similarity satisfies the threshold, the spam template with the greatest similarity with the short message is obtained as a spam template matching the short message.
The method for intercepting a short message according to claim 1, 2 or 3, wherein the step of searching for a spam template matching the short message according to the feature vector of the short message further comprises:

If the spam template matching the short message is not found, the default spam template is used to intercept the short message.
The short message interception method according to claim 1, 2 or 3, wherein the SMS is intercepted according to a spam template matched with the short message and a preset filtering condition by using a corresponding intercept frequency policy The steps include:

Obtaining a type of spam template that matches the short message;

The type of the spam template matched with the short message is used as the type of the short message;

Obtaining a corresponding interception frequency policy according to the type of the short message and the preset filtering condition;

Obtaining, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number;

Determining whether the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy;

If the sending frequency of the calling number sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy, intercepting the short message;

If the sending frequency of the calling number sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy, the short message is released.
A short message intercepting device, the short message intercepting device comprising:

a receiving module, configured to receive a short message, and obtain a feature vector of the short message;

a matching module, configured to search for a spam template matching the short message according to the feature vector of the short message;

The intercepting module is configured to intercept the short message according to the spam template matched with the short message and the preset filtering condition by using a corresponding intercept frequency policy.
The short message intercepting device of claim 6, wherein the preset filtering condition comprises: a calling number of the short message, a called number, and a sending time, and the short message intercepting device further comprises:

Obtaining a module, configured to obtain a type of the spam message template and the spam message template input by the user; and set an interception frequency policy according to the type of the spam message template input by the user, the calling number, the called number, and the sending time period;

a calculation module, configured to acquire a feature vector of the spam template input by the user;

The list module is configured to filter out spam templates having the same feature vector according to the feature vector of the spam template, and form a list for searching.
The short message intercepting device of claim 6, wherein the matching module comprises:

a searching unit, configured to search for a spam short message template list having the same feature vector according to the feature vector of the short message;

The similarity calculation unit is configured to obtain a spam short message template having the same similarity as the short message template in the spam short message template list if the spam short message template list having the same feature vector is found;

a determining unit, configured to determine whether the maximum similarity meets a threshold;

The matching unit is configured to acquire a spam template having the greatest similarity with the short message as the spam template matched with the short message if the maximum similarity satisfies the threshold.
The short message intercepting device according to claim 6, 7 or 8, wherein the intercepting module is further configured to

If the spam template matching the short message is not found, the default spam template is used to intercept the short message.
The short message intercepting device according to claim 6, 7 or 8, wherein the intercepting module comprises:

a type unit, configured to obtain a type of a spam template that matches the short message; and use a type of the spam template that matches the short message as a type of the short message;

The policy unit is configured to obtain a corresponding interception frequency policy according to the type of the short message and a preset filtering condition;

The frequency unit is configured to obtain, according to the type of the short message and the calling number, the sending frequency of the short message sent by the calling number;

a determining unit, configured to determine whether the sending frequency of the type of short message sent by the calling number exceeds an intercepting frequency of the intercepting frequency policy;

The intercepting unit is configured to intercept the short message if the sending frequency of the type of short message sent by the calling number exceeds the intercepting frequency of the intercepting frequency policy;

The release unit is configured to release the short message if the sending frequency of the type of the short message sent by the calling number does not exceed the intercepting frequency of the intercepting frequency policy.