KR20110022245A - Spam filtering system using a front-end processing and spam filtering method thereof - Google Patents

Abstract

PURPOSE: A spam filter system including a preprocess function and a spam filtering method using the same are provided to be able to filter existing spam filtering rules by performing preprocessing process. CONSTITUTION: A text message transmitting service center(300) transmits a message to a spam filtering server. The spam filtering server includes a preprocessing unit(410), a spam determining unit(420), and a preprocessing DB(430). The spam filtering server received the message and transforms the received message into a preprocessed document. The spam filtering server determines whether it is a spam or not by using a filtering rule.

Description

Spam filtering system including pre-processing function and spam filtering method using same {SAM FILTERING SYSTEM USING A FRONT-END PROCESSING AND SPAM FILTERING METHOD THEREOF}

The present invention relates to a spam filtering system and a spam short message processing method for blocking a spam short message.

As the short message service (SMS) in mobile communication is activated, a mobile terminal user can transmit and receive information through text as well as conventional voice. However, complaints of mobile terminal users are increasing because spam short messages that are not intended to be received are often transmitted due to a negative phenomenon caused by short message services.

Accordingly, technologies for blocking spam short messages that the recipient does not want are being developed.

There are two main techniques for blocking such spam short messages.

In the early days, techniques for blocking spam short messages in mobile communication terminals were mainly devised. According to this technique, after the mobile terminal receives the short message, the mobile terminal determines whether the received short message includes a calling number corresponding to the received short message and includes the calling number corresponding to the received short message. The mobile terminal performs the short message rejection processing. Similarly, after receiving the short message, the mobile terminal performs the short message rejection processing when a string matching the string stored in the reject string database exists in the short message. However, according to the related art, the capacity of the mobile terminal is limited, so that the number or character string that can be stored is limited, and there is a problem in that the mobile terminal determines whether or not to reject the message.

Recently, in addition to determining whether a message is spam in a mobile terminal, a technique for determining whether a message is spam in a mobile communication network has been developed. The spam filtering system for a mobile communication network includes (1) receiving a short message, (2) filtering the short message based on at least one rule set, and (3) based on the filtering result. It is common to consist of three steps of the short message spam filtering method including the step of processing the short message. In the filtering step, the received short message is buffered to collect main parameters from the message, and a rule set is applied thereto to obtain a result of the filtering.

However, as spam short messages become more intelligent, spam filtering systems such as the above example cannot implement the level of spam short messages desired by the user. Especially in the recent mobile communication, spam short messages include meaningless special characters, periods and spaces (▶, ♣, ☜,., ^, __) among key words, or meaningful special characters (㉢, ㈀). , ㉲) is used instead of ordinary characters. Or, instead of "loan", use the case of composing the common letter by separating the morphemes of the letter like "ㄷ ㅐ ㅊ ㄹ ㄹ".

As spam short messages are transformed by special characters, punctuation marks, spaces, stemming, etc., instead of normal short messages, buffer the received short messages and collect main parameters from the messages. Therefore, it is impossible to obtain a normal result by applying a rule set.

The present invention was created in view of the above circumstances, and an object of the present invention is to use a modified short message rather than a normal short message as described above, and to filter a message that is difficult to apply to a conventional filtering rule. The present invention provides a method for effectively filtering spam by preprocessing the short message before applying the filtering rule.

A spam filtering system according to the present invention for achieving the above object comprises: a spam filtering system for determining whether a message is intended to be sent to a called terminal by a spam terminal, the SMS center transmitting the message to a spam filtering server; And a spam filtering server that receives the message, transforms the received short message into a preprocessed document that is in a form for applying to a filtering rule, and then determines whether to spam by using the filtering rule.

The SMS center receives the message from the mobile communication network, queries the home location register (HLR) whether the called party, the owner of the called terminal, has subscribed to the spam blocking service, and receives the caller from the home location register (HLR). When it is confirmed that the user has subscribed to the spam blocking service, it is preferable to deliver the short message to the spam filtering server.

On the other hand, a spam filtering server according to the present invention for achieving the above object is a spam filtering server for determining whether the originating terminal to send a message to be sent to the destination terminal, the form for applying the received short message to the filtering rule A preprocessing unit transforming the preprocessed document into a document; And a spam determination unit determining whether the preprocessed document is spam using a filtering rule.

And the short message preprocessing unit extracts a special character from the received short message; And a special character processing unit for converting the special character into a normal character when the extracted special character is meaningful, and deleting the special character when it is not meaningful.

The short message preprocessing unit may include a delimiter extracting unit for extracting a letter whose letter is separated from the received short message; And it is preferable to include a delimiter synthesizing unit to combine the extracted delimiter to transform into a normal character.

The short message preprocessor includes: a morpheme analysis unit configured to perform morpheme analysis to analyze the received short message in word units; And a keyword extraction unit for extracting a keyword from the morphologically analyzed message.

The preprocessing unit may further include a preprocessing DB in which data for transforming the short message into a preprocessed document is stored.

The preprocessing DB preferably includes at least one special character and a general character corresponding thereto.

In addition, the pretreatment DB preferably includes at least one of a morpheme dictionary and a term dictionary.

The apparatus may further include a preprocessing DB editing unit which additionally inputs prestored data in the preprocessing DB or edits the prestored data.

On the other hand, a portable terminal for achieving the above object according to an embodiment of the present invention, in the portable terminal capable of receiving a message, transforming the message received in the portable terminal into a pre-processed document which is a form for applying a filtering rule Pretreatment unit; And a spam determination unit determining whether the preprocessed document is spam using a filtering rule.

And the preprocessing unit extracts a special character from the received message; And a special character processing unit for converting the special character into a normal character when the extracted special character is meaningful, and deleting the special character when it is not meaningful.

The preprocessing unit may include a delimiter extracting unit configured to extract a letter having a separated letter from the received message; And it is preferable to include a delimiter synthesizing unit to combine the extracted delimiter to transform into a normal character.

The preprocessor includes: a morpheme analysis unit configured to perform morpheme analysis to analyze the received short message in word units; And a keyword extraction unit for extracting a keyword from the morphologically analyzed message.

On the other hand, a method for processing a short message spam to achieve the above object according to an embodiment of the present invention is a method for filtering a short message spam (short message spam), receiving a short message receiving a short message; A short message preprocessing step of transforming the received short message into a preprocessed document which is a form for applying to a filtering rule; And a spam determination step of determining whether the preprocessed document is spam using a filtering rule.

And the short message preprocessing step comprises: extracting a special character from the received short message; And a special character processing step of converting the special character into a normal character if the extracted special character is meaningful, and deleting the special character if it is not meaningful.

In addition, the short message preprocessing step may include a delimiter character extracting step of extracting a letter whose letter is separated from the received short message; And it is preferable to include a delimiter synthesis step of transforming the extracted delimiter to a general character.

The short message preprocessing step may include a morpheme analysis step of performing a morpheme analysis to analyze the received short message in word units; And a keyword extraction step of extracting a keyword from the morphologically analyzed message.

The special character processing step may be a preprocessed document which is a form for applying the received short message to a filtering rule by using data extracted from a preprocessing DB that stores data for transforming the short message into a preprocessed document. It is desirable to deform.

The preprocessing DB preferably includes at least one special character and a general character corresponding thereto.

In addition, the pretreatment DB preferably includes at least one of a morpheme dictionary and a term dictionary.

The method may further include a preprocessing DB editing step of additionally inputting prestored data in the preprocessing DB or editing the prestored data.

The method may further include a message processing step of processing the short message according to the determination result of the spam determination step.

As described in detail above, by using the spam filtering system according to the present invention, it is possible to efficiently block spam messages that the recipient does not want.

Particularly, in the present invention, before the spam message is applied to the conventional spam filtering rule, the pre-processing is performed through a pre-defined pre-processing rule and a pre-processing DB, so that the filter can be filtered even in the conventional spam filtering rule. Through this, it is possible to effectively block the modified spam message.

Hereinafter, a spam filtering system and a spam filtering method according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram illustrating a spam filtering system according to the present invention.

The originating terminal 100 is a terminal for transmitting a spam message, and after creating a spam short message, the terminal 100 transmits the message to the called terminal 110 using the mobile communication network 200.

That is, the short message prepared by the calling terminal 100 passes through the short message center (SMS center, 300) after passing through the base station (BTS), the control station (BSC) and the exchange (MSC), and then again exchange (MSC), control station ( The receiving terminal 110 receives the BSC and the base station BTS.

Meanwhile, after receiving the message from the mobile communication network 200, the short message center 300 according to an embodiment of the present invention blocks a spammer who is the owner of the called terminal with a home location register (HLR, 310). Inquire whether the user has subscribed to the service, and if it is confirmed from the home location register 310 that the called party subscribes to the spam blocking service, the message is transmitted to the spam filtering server 400.

The home location register 310 is a basic equipment of the mobile communication network that enables the user to check the location of the mobile communication subscriber and at the same time grasps various information about the subscriber and controls various additional services such as incoming and outgoing calls. It includes whether the subscriber has subscribed to the anti-spam service.

The spam filtering server 400 includes a preprocessor 410, a spam determination unit 420, and a preprocessing DB 430.

The preprocessor 410 transforms the received short message into a preprocessed document which is a form for applying a filtering rule.

The spam determination unit 420 determines whether or not the preprocessed document is spam using a filtering rule.

The preprocessing DB 430 stores data for transforming the short message into a preprocessed document in the preprocessor 410.

The spam filtering server 400 will be described in more detail below with reference to FIG. 2.

2 is a block diagram of a spam filtering server according to an embodiment of the present invention.

The spam filtering server 400 according to an embodiment of the present invention includes a preprocessor 410, a spam determination unit 420, and a preprocessing DB 430 as described above.

The preprocessor 410 again includes a special character processor 411, a delimiter processor 412, a morpheme processor 413, and a keyword extractor 414.

The special character processing unit 411 extracts a special character from the received short message, converts the special character into a normal character if the extracted special character is meaningful, and deletes the special character if it is not meaningful. do.

That is, when it is assumed that there is a content such as "loan ▶" in the short message, the special character processing unit 411 extracts the special character "▶" from the short message and performs a function of deleting it because it is meaningless. do. Accordingly, after passing through the special character processing unit 411, the short message having the contents of "loan ▶ loan" is preprocessed with the content of "loan", and the preprocessed short message is able to be filtered by the spam determination unit 420.

In addition, assuming that a short message has a meaningful special character, that is, for example, a special character such as "㉢, ㈀, ㉲", it is preprocessed as a general character such as "ㄷ, a, e". Accordingly, after passing through the special character processing unit 411, the short message having the content of "warning" is preprocessed with the content of "horse racing", and the preprocessed short message can be filtered by the spam determination unit 420.

The delimiter processing unit 412 extracts a letter whose letter is separated from the received short message, and combines the extracted delimiter and transforms it into a normal character.

That is, assuming that there is a content such as "ㄷ ㅐ ㅊ ㄹ ㄹ consultation" in the short message, the delimiter processing unit 412 extracts the letter "ㅐ ㄹ ㅅ ㄹ ㄹ." The extracted delimiter is combined and transformed into a general character of "loan". As a result, the short message in which the contents of the contents of the "c ㄷ ㅅ ㅜ ㄹ consultation" are described is preprocessed as "loan counseling", and the preprocessed short message may be filtered by the spam determination unit 420.

The morpheme processing unit 413 performs a morphological analysis to analyze the short message in word units. This removes the investigation from the short message it is possible to filter more efficiently than the spam determination unit 420.

The keyword extractor 414 extracts a keyword from a short message. By extracting only the keywords like this, more efficient filtering is possible in the spam determination unit 420.

 The preprocessing DB 430 stores data for transforming the short message into a preprocessed document in the preprocessor 410. The preprocessing DB 430 includes one or more special characters and general characters corresponding to the preprocessing DB, thereby providing data necessary for performing the preprocessing in the preprocessing unit 410.

The preprocessing DB 430 provides various data necessary for preprocessing, and in particular, includes a grammatical form dictionary, a lexical form dictionary, a field terminology dictionary, a user-defined dictionary, and a dictionary analysis dictionary.

In addition, by analyzing trends of short spam message types from the outside, it is sometimes found that a specific character pattern that has not been preprocessed is used in the spam message. In this case, it is also possible to register a new specific character pattern in the preprocessing DB, recognize the information in the preprocessing server, and replace the specific character pattern with a new string registered in the preprocessing DB.

To this end, the preprocessing DB includes a function of inputting a new specific string or editing an existing specific string through external spam collection, input by an operator, and transmission of information from other external DBs.

3 is a flowchart illustrating a spam short message processing method according to an embodiment of the present invention.

First, if the calling terminal 100 creates a short message, the short message is transmitted to the short message center 300 through the mobile communication network 200 (S510, S515).

Thereafter, the home location register (HLR) checks whether the recipient of the message, that is, the receiver of the called terminal, has subscribed to the spam blocking service (S520).

When it is confirmed that the called party subscribes to the spam blocking service, the short message is transmitted to the preprocessor 410 of the spam filtering server 400 (S530).

The preprocessing unit 410 imports data necessary for preprocessing from the preprocessing DB 430 and then preprocesses the short message using the data (S540).

Thereafter, the preprocessed document is transmitted to the spam determination unit 420 (S550).

The spam determination unit 420 determines whether or not the preprocessed document is spam (S560), and transmits the determined result to the short message center 300 (S570).

If the short message center 300 receives the result that the short message is not spam from the spam determination unit 420, and transmits the short message to the destination terminal 110 through the mobile communication network 200 (S580, S585). ).

4 is a flowchart illustrating a spam short message preprocessing method according to an embodiment of the present invention.

First, the preprocessor 410 receives a short message from the short message center 300 (S610).

After that, a special character is extracted from the received short message (S620).

If the extracted special character has no meaning (S630-N), the special character is deleted (S640). In other words, assuming that there is a content such as "loan" in a short message, the special character processing unit 411 extracts a special character "▶" from the short message (S620), which is meaningless (S630-). N) to perform the function of deleting (S640). Accordingly, the short message in which the content of "loan" is described is preprocessed with the content of "loan".

In addition, assuming that a short message has a meaningful special character (S630-Y), that is, for example, a special character such as "㉢, ㈀, ㉲" is converted to a general character such as "ㄷ, ,, 마". (S635). Through this pretreatment, a short message having a content of "warning" is preprocessed with "horse racing".

If there is a delimiter (S650-Y), the delimiter is combined and transformed into a normal character (S655). In other words, assuming that a short message has the same content as "ㄷ ㅐ ㅊ ㄹ ㄹ consultation", the short message is extracted with the letter "ㄷ ㅐ ㅊ ㄹ ㄹ." And the extracted delimiters are combined to " Loan ". As a result, the short message in which the contents of the "c ㅐ ㅅ ㅜ ㄹ consultation" are described is preprocessed with the contents of "loan counseling."

Subsequently, the morphological analysis is performed to analyze the short message in word units (S660). This removes the investigation from the short message it is possible to filter more efficiently than the spam determination unit 420.

The keyword is extracted from the short message (S665).

The preprocessed document is used to determine whether or not spam using filtering rules (S670).

If it is determined that the spam is spam (S670-Y), the short message is deleted (S685). On the contrary, if it is determined that the spam is not spam (S670-N), the short message is transmitted to the called terminal (S680).

So far, a spam filtering system including a preprocessing function for improving the accuracy of filtering a spam message and a spam filtering method using the same have been described.

Meanwhile, the present embodiment has been described with reference to a method applied to the spam filtering server for determining whether the message is spam in a mobile communication network, but is not limited thereto. Therefore, it can be applied to the case where the receiving terminal determines whether the received short message is spam. That is, since the mobile terminal has a built-in component for performing the preprocessing, the mobile terminal may be applied to a method for performing the preprocessing before spam filtering in the mobile terminal.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be construed as limiting the scope of the invention as defined by the appended claims. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

The spam filtering system including the preprocessing function and the spam filtering method using the same according to the present invention are methods for overcoming the conventional spam filtering rules, which are not filtered due to the spam messages used in recent years. As it goes beyond the limitations of the existing technology in that the pre-processing is performed as a general message and then the spam filtering rule is applied, the possibility of marketing or sales to the applied mobile communication system and terminal equipment as well as the use of the related technology. This is not only sufficient but also practically evident, and thus it is an invention with industrial applicability.

1 is a schematic block diagram showing a spam filtering system according to the present invention;

2 is a block diagram of a spam filtering server according to an embodiment of the present invention;

3 is a flowchart illustrating a spam message processing method according to an embodiment of the present invention;

4 is a flowchart illustrating a spam message preprocessing method according to an embodiment of the present invention.

Explanation of symbols on the main parts of the drawings

100: calling terminal 110: called terminal

200: mobile communication network 300: short message center

310: home location register 400: spam filtering server

410: preprocessing unit 420: spam determination unit

430: preprocessing DB

Claims (23)

In the spam filtering system for determining whether or not the originating terminal spams the message to be sent to the called terminal, An SMS center for transmitting the message to a spam filtering server; And And a spam filtering server for receiving the message, transforming the received message into a preprocessed document which is a form for applying a filtering rule, and then determining whether spam is detected using the filtering rule. . The method of claim 1, The SMS Center is Receive the message from the mobile communication network, and query the home location register (HLR) whether the called party, the owner of the called terminal has subscribed to the spam blocking service, and from the home location register (HLR) to the spam blocking service Spam filtering system, characterized in that for forwarding the message to the spam filtering server. In the spam filtering server that determines whether or not the originating terminal spams the message to be sent to the called terminal, A preprocessing unit transforming the received message into a preprocessed document which is a form for applying a filtering rule; And And a spam determination unit for determining whether the preprocessed document is spam using a filtering rule. The method of claim 3, The message preprocessing unit A special character extracting unit extracting a special character from the received message; And If the extracted special character is meaningful spam filtering server, characterized in that it comprises a special character processing unit for converting the special character into a normal character, and if not meaningful delete the special character. The method of claim 3, The message preprocessing unit A delimiter extracting unit extracting a letter having a separated letter from the received message; And Spam filtering server characterized in that it comprises a delimiter synthesis unit for combining the extracted delimiter to transform into a normal character. The method of claim 3, The message preprocessing unit A morpheme analysis unit for performing a morpheme analysis to analyze the received message in word units; And Spam filtering server comprising a keyword extraction unit for extracting a keyword from the stemmed message. The method of claim 3, And a pre-processing DB in which data for transforming the message into a pre-processed document is stored in the pre-processing unit. The method of claim 7, The preprocessing DB is spam filtering server, characterized in that it comprises one or more special characters and corresponding general characters. The method of claim 7, The preprocessing DB is spam filtering server, characterized in that it comprises any one or more of a morpheme dictionary and a term dictionary. The method of claim 7, Spam filtering server further comprises a pre-processing DB editing unit for additionally inputting the pre-stored data in the pre-processing DB or edit the pre-stored data. In a mobile terminal capable of receiving a message, A preprocessing unit transforming the message received by the mobile terminal into a preprocessed document which is a form for applying a filtering rule; And And a spam determination unit for determining whether the preprocessed document is spam using a filtering rule. The method of claim 11, The preprocessing unit A special character extracting unit extracting a special character from the received message; And And a special character processing unit for converting the special character into a normal character and deleting the special character if the extracted special character is meaningful. The method of claim 11, The preprocessing unit A delimiter extracting unit extracting a letter having a separated letter from the received message; And And a delimiter synthesizing unit for converting the extracted delimiter into a general character. The method of claim 11, The preprocessing unit A morpheme analysis unit for performing a morpheme analysis to analyze the received message in word units; And And a keyword extraction unit for extracting a keyword from the stemmed message. As a method of filtering spam messages, A message receiving step of receiving a message; A message preprocessing step of transforming the received message into a preprocessed document in a form for applying a filtering rule; And And a spam determination step of determining whether the preprocessed document is spam using a filtering rule. The method of claim 15 The message preprocessing step A special character extracting step of extracting a special character from the received message; And If the extracted special character is meaningful spam message processing method comprising the step of converting the special character to a normal character, if the meaning is not special character processing step of deleting the special character. The method of claim 15 The message preprocessing step A delimiter character extraction step of extracting a letter having a separated letter from the received message; And Spam message processing method characterized in that it comprises a delimiter synthesis step of transforming the extracted delimiter to a general character. The method of claim 15 The message preprocessing step A morpheme analysis step of performing morpheme analysis to analyze the received message in word units; And Spam message processing method characterized in that it comprises a keyword extraction step of extracting a keyword from the morphologically analyzed message. The method of claim 16, Special character processing step Spam, characterized in that for converting the received message into a preprocessed document which is a form for applying the filtering rule, using data extracted from the preprocessing DB that stores the data for transforming the message into a preprocessed document. How messages are handled. The method of claim 19 The preprocessing DB is spam message processing method characterized in that it comprises one or more special characters and corresponding general characters. The method of claim 19 The preprocessing DB is a spam message processing method comprising any one or more of a morpheme dictionary and a term dictionary. The method of claim 19 Spam message processing method characterized in that it further comprises a pre-processing DB editing step for additionally inputting the pre-stored data in the pre-processing DB or editing the pre-stored data. The method of claim 15 And a message processing step of processing the message according to the determination result of the spam determination step.

Images