KR20140127036A - Server and method for spam filtering - Google Patents

Abstract

Provides techniques to filter spam messages with precise probability. The spam filtering server according to an embodiment of the present invention stores predetermined words and determines whether a message transmitted from a message server providing a service for sending and receiving messages between terminals includes at least one of predetermined words A first filtering unit for classifying a transmitted message into a first message, which is a message having a possibility of spam, according to whether or not the message is spam; When the message classified as the first message is transmitted from the message server to the external terminal, the first request information is information for inquiring whether to classify the message classified as the first message into the second message as the spam message, A spam information transmission unit for controlling the transmission of the spam information; And receiving response information of an external terminal with respect to the first request information, and when the response information includes information that classifies a message classified as the first message into a second message, 2 message and storing the classified message in the spam message database.

Description

TECHNICAL FIELD [0001] The present invention relates to a spam filtering server and a spam filtering server,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for filtering a spam message, and more particularly, to a technique for improving the accuracy of spam message filtering by applying a multi-dimensional spam message filtering method.

Recently, as a technology for transmitting and receiving data has been developed, the utilization rate of services for transmitting and receiving messages among users is increasing by using a mobile terminal, a computer, and the like. The message transmission / reception service includes, for example, a real-time chat service, a note service, and a mail service.

Users can exchange information more quickly than sending and receiving messages off-line by exchanging private information or performing work using a message transmission / reception service.

However, in such a message transmission / reception service, users have suffered from a lot of inconvenience due to a spam message. Spam means primarily advertising messages that users do not want.

Research has continued to filter such spam messages to prevent them from being sent to users, or to save users' inconvenience by storing them in the spam message storage space included in the message reception module of users.

A spam message is filtered by determining whether a message transmitted using a certain spam filtering system is a spam message before transmitting a message transmitted from the message server. As a result, researches continue to develop a spam filtering system with high probability of spam filtering.

However, the existing spam filtering system classified spam messages as spam messages even though they are not spam messages, or classified spam messages as spam messages, and the accuracy of spam filtering is very low.

Accordingly, it is an object of the present invention to provide a more reliable spam filtering technique through feedback of a spam filtering system, having a very high spam filtering accuracy as compared with the existing spam filtering system.

In order to achieve the above object, a spam filtering server according to an embodiment of the present invention stores preset words, and a message transmitted from a message server providing a service for transmitting / A first filtering unit for classifying the transmitted message into a first message having a possibility of spam, depending on whether the message includes at least one of words; When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message and transmits the spam information; And when the response information includes information indicating that the message classified into the first message is classified into the second message, receiving the response information of the external terminal with respect to the first request information, And a second filtering unit for classifying the received message into the second message and storing the classified message in the spam message database.

The spam filtering method according to an embodiment of the present invention is a method in which a message sent from a message server providing a service for sending and receiving a message between terminals includes at least one of pre- Classifying the transmitted message into a first message having a possibility of spam, depending on whether the message is spam or not; When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message to be transmitted; And if the response information of the external terminal for the first request information includes information for classifying a message classified as the first message into the second message, And storing the classified message in a spam message database.

The spam filtering method according to another embodiment of the present invention determines whether or not the spam filtering server classifies a message transmitted from a message server providing a service for sending and receiving messages between terminals to a second message that is a spam message step; Classifying the transmitted message as a first message having a possibility of spam according to whether the transmitted message is not classified as the second message or not, if the transmitted message includes at least one of pre-stored words; When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message to be transmitted; And if the response information of the external terminal for the first request information includes information for classifying a message classified as the first message into the second message, And storing the classified message in a spam message database.

According to the present invention, a spam message is extracted by extracting a message that is primarily expected to be spam by the first filtering unit and finally determining the spam message as a second filtering unit spam message according to the response information of the external terminal, It is possible to provide a more accurate spam filtering system than the existing spam filtering system.

In addition, since a re-learning module for spam filtering is provided together with a diversified filtering system through reporting of spam from a user, monitoring from an administrator terminal, pattern analysis, etc., it is possible to correctly filter only spam messages, So that the reliability of the spam filtering can be improved.

1 is a configuration diagram of a spam filtering server according to an embodiment of the present invention.
2 to 8 are flowcharts of a spam filtering method according to each embodiment of the present invention.
FIG. 9 shows an example of a message reception screen displayed on a user terminal according to an implementation of each embodiment of the present invention.

Hereinafter, a spam filtering server and method according to embodiments of the present invention will be described with reference to the accompanying drawings.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. The following examples are intended to illustrate the present invention and should not be construed as limiting the scope of the present invention. Accordingly, equivalent inventions performing the same functions as the present invention are also within the scope of the present invention.

In the following description, the same reference numerals denote the same components, and unnecessary redundant explanations and descriptions of known technologies will be omitted.

In the embodiments of the present invention, " communication ", " communication network ", and " network " The three terms refer to wired and wireless local area and wide area data transmission and reception networks capable of transmitting and receiving a file between a user terminal, a terminal of another user, and a download server.

In the following description, " server " means a server computer that determines whether to classify a message as spam. In the case of a service having a small capacity or a small number of processing data, a plurality of services can be operated on one server. Also, in the case of a service having a very large capacity or a large amount of messages that need to be determined as spam in real time, there may be one or more servers for operating one service according to the function of the service.

In addition, middleware for the database and servers performing settlement processing may be connected to the server, but a description thereof will be omitted in the present invention.

In the present invention, the term 'spam filtering' refers to the entire process of analyzing each message transmitted from a message server providing a message transmission / reception service, which is a service capable of transmitting / receiving a message between users, and determining whether the corresponding message is a spam message.

Spam messages are e-mail messages sent in bulk to a large number of recipients on the Internet, or news articles that are simultaneously posted in a large number of newsgroups. It is the same as an advertisement or a junk mail sent to a lot of unspecified number of recipients by mail.

In most cases, spam is a message that the recipient does not want and does not care about, or that is not related to the discussion topic of each newsgroup. The act of sending such a message or porting an article is called spamming. Spamming is treated as an act of abusing the Internet for the purpose of advertising goods to a large number of people at a nominal cost, propagating a particular religion, or even denigrating a particular person, a particular product or a particular business.

Therefore, the spam filtering server and method according to each embodiment of the present invention may be comprised of a separate server connected to the message server or included in the message server. Hereinafter, the spam filtering server is referred to as a server, and the message server may be referred to as a message server.

Hereinafter, the description of FIG. 1 will be described with reference to the flowcharts of FIGS. Accordingly, it will be understood that the technical features of the spam filtering method according to each embodiment of FIGS. 2 to 8 can be performed by the respective components of the spam filtering server according to the embodiment of FIG.

1 is a configuration diagram of a spam filtering server according to an embodiment of the present invention.

Referring to FIG. 1, a spam filtering server 10 according to an embodiment of the present invention basically includes a first filtering unit 11, a spam information transmitting unit 12, and a second filtering unit 13, Learning module 14 may be additionally included.

Referring to FIG. 2 together with FIG. 1, the first filtering unit 11 determines whether a message transmitted from the message server 30 includes at least one of preset words stored in the first filtering unit 11 (S10) of classifying the transmitted message into the first message.

The first filtering unit 11 receives a message transmitted from the message server 30 before message transmission and determines whether the transmitted message corresponds to the first message. The first message means a message likely to be spam, and in the existing spam filtering system, it is classified as a spam message when it is classified as the first message.

The first filtering unit 11 specifically determines whether the message transmitted through the flow described in FIG. 3 is the first message.

Referring to FIG. 3, the first filtering unit 11 first extracts the content of the transmitted message, calculates a spam probability according to whether at least one of predetermined words is included in the content of the message (S11) , And classifying the transmitted message as the first message (S12) when the calculated spam probability exceeds a preset threshold probability (for example, 80%).

A representative example of the function performed by the embodiment of FIG. 3 and the first filtering unit 11 is a Bayesian filter. The Bayesian filter is a mathematical theory of the 1980s, and is a representative technique for filtering data based on probability.

Bayesian filter extracts and combines spam messages and non-spam samples, then learns the spam frequency (probability) of each word, compares the learned index with the message, extracts the word contained in the message, It is a filter that classifies spam messages by calculation.

Most of the existing spam filtering system is a filtering method using Bayesian filter, and the learning system has been applied to improve the functionality of Bayesian filter. The learning system updates the spam frequency of the word by reflecting the word contained in the message classified as spam through the Bayesian filter to the learned index.

However, Bayesian poisoning is a problem in spam filtering using existing Bayesian filter. Bayesian poison is to send a non-spam message to a large number of messages, so that the Bayesian filter will judge the normal message as spam.

In the present invention, to prevent the above-mentioned problems of the Bayesian filter and errors due to learning, it is only determined whether or not the message transmitted as the filtering result employing the Bayesian filter is the first message likely to be spam, Filtering accuracy is being improved.

In the first filtering unit 11, for example, a message transmitted using a Bayesian filter is classified as a first message, but it is natural that all other known automatic spam filtering methods can be used.

Referring back to FIGS. 1 and 2, when the message classified as the first message is transmitted to the external terminal 20 through the message server 30, the spam information transmitter 12 transmits the first message as a spam message (S20) controlling the message server 20 to include the first request information in the first message as information for inquiring whether to classify the message into two messages.

As described above, the first message means a message classified as a message likely to be spam. According to the present invention, when the message server 30 transmits a message to the external terminal 20 receiving the message, it is possible that the first message is not spam, and whether or not the message is determined as spam And transmits the first request information.

Referring to FIG. 9, an example of a reception screen of a message to which first request information is added is shown.

Referring to the message receiving screen 100, the contents information 101 of the message can be included and the first request information 102 can be included. The first request information 102 may include information indicating that the corresponding message is highly likely to be spam and information for inquiring whether to confirm the message as spam.

A user using the external terminal 20 can select the response information 103 and the response information is transmitted to the second filtering unit 13 through the message server 30. [

1 and 2, the second filtering unit 13 receives the above-described response information of the external terminal 20 with respect to the first request information, If the message is classified as the second message, that is, the spam message, the message classified as the first message is classified into the second message and stored in the spam message database (S30).

The spam message database is not shown in the drawing, but may be provided in the storage space of the second filtering unit 13 or the message server 30. [ In the spam message database, the contents information of the spam message and the transmission information of the spam message can be stored.

The transmission information of the spam message refers to all information capable of identifying the sender, such as, for example, transmission IP information and sender information.

On the other hand, the second filtering unit 13 can classify a certain message into the second message irrespective of the reception of the response information corresponding to the first request information for the first message. A description thereof is shown in Figs. 4 to 6. Fig.

For example, referring to FIG. 1 and FIG. 4, the second filtering unit 13 may receive second request information requesting classification of a message transmitted from the message server 30 as a second message from the external terminal 20, (S40), a step S50 of classifying a message corresponding to the second request information into a second message is performed.

For example, a user who uses the message transmission / reception service through the message server 30 may want to check a message not classified as spam and register the message as spam. In this case, the user can register the message as a spam message through the external terminal 20, and the second filtering unit 13 receives the information as the second request information and classifies the message as the second message .

Alternatively, the administrator terminal of the message server 30 may also be included in the external terminal 20. In this case, the external terminal 20 means a terminal performing a function of monitoring messages transmitted from the message server 30.

The administrator can manually check the message through the administrator terminal of the message server 30 to determine whether the corresponding message is spam or not. Alternatively, you can manually check the sending pattern of the message.

At this time, regardless of the use of the message transmission / reception service, the administrator can confirm some of the transmitted messages as spam. In this case, the second request information can be transmitted by the administrator terminal.

Meanwhile, as described above, when storing the information of the second message in the spam message database, all information that can identify the content of the message and the sender can be stored.

In this case, when the message transmitted from the message server 30 is a message corresponding to the sender identification information stored in the spam message database, the second filtering unit 13 classifies the message into the second message, Can be stored.

This technique can be exemplified as a spam filtering method of the DNS Query method which is recently used as a specific IP-based spam filtering method.

1, 5 and 6, the second filtering unit 13 may classify a specific message into a second message using information on the number of times the message is sent. Is also a function to perform using a message transmitted from the message server 30 regardless of the response information.

First, referring to FIG. 5, the second filtering unit 13 first receives information of a message classified as a first message from the first filtering unit 11, and stores the message classified as the first message in a predetermined first threshold A step S41 of detecting whether the data has been repeatedly transmitted exceeding the number of times is performed.

In the case of spam messages, it is common that they are repeatedly sent to random recipients, which means that the messages classified as the first message are very likely to be spam.

That is, when the message classified as the first message exceeds the first threshold number (for example, 1000 times) through the execution of step S41, the second filtering unit classifies the message into the second message (S51) Can be performed.

Referring to FIG. 6, the message server 30 detects whether there is a response message from the external terminal 20 in response to the previously transmitted message, and determines whether the number of origination The second filtering unit 13 performs a step S42 of detecting the number of times that the message having the same content as the message in which the response message does not exist from the message server 30 has a predetermined second threshold number (E.g., 2000 times), the step S52 of classifying the message into the second message is performed.

In the case of a normal message, the receiving user sends a response message to the user in many cases, and in the case of a message with no response, it is likely to be spam. Accordingly, the second filtering unit 13 detects the number of times the message is transmitted from the message server 30 counting the number of times the same message is transmitted in the case of a message that is only transmitted without a response, and when the number of times of transmission is a second threshold number , The message is determined as spam.

4 to 6, it is possible to effectively reduce the number of times that the external terminal 20 has to input the response information in the embodiment of classifying it into the second message through the response information of the external terminal 20, There is an effect that can be classified accurately. According to the embodiment of FIGS. 4 and 6, the number of messages that are simply classified into the first message can be reduced.

Meanwhile, for the same effect as the above effect, the first filtering unit 11 and the second filtering unit 13 can perform the line filtering function according to the embodiment of FIG.

1 and 7, before the second filtering unit 13 determines whether the message transmitted by the first filtering unit 11 is the first message, the transmitted message is stored in the spam message database (S1) of determining whether the message is the same as any one of the first and second messages, and if the transmitted message is the same message as the second message, classifying the message as a second message, (S53) is performed.

In the spam message database, information of a message classified into the second message is stored in a database as described above.

Therefore, if the transmitted message is the same as any one of the messages stored in the spam message database, the second filtering unit 13 performs the function by the first filtering unit 11 And classifies the transmitted message into the second message immediately.

If it is determined in step S1 that the transmitted message is a message different from all of the second messages stored in the spam message database, the message transmitted in the Locality-Sensitive Hashing (LSH) And compares each of the messages and determines whether the message transmitted according to the comparison result is a message that overlaps with the second message (S2).

The LSH method is a technique used in data mining techniques, which means a technique for calculating the probability of duplicate documents by comparing two documents.

LSH is a way to represent a document with several eigenvalues. For example, if a document consists of 100 words, it can be represented as a 100-dimensional vector. The LSH method is a technique of reducing this to n dimensions of limited size. If the size of n is made small, it is possible to judge duplication of documents or to enable clustering.

For example, when there are two documents, IDs are assigned to all the keywords included in the contents of the document (the IDs are assigned the same IDs for the same keywords among the two documents), and n-dimensional functions . For example, in each document, n term IDs are extracted at k intervals. The interval can be set arbitrarily. At this time, the smallest value is selected from the extracted n IDs. If the selected IDs are the same, the two documents are judged to be highly similar documents.

Specifically, there are the documents A and B, the keywords 3, 6, 9, and 12 in the document A are extracted and sorted, and the keyword K1 at the top is selected. Then, Extracts the keywords, sorts them, and selects the keyword K2 at the top. Finally, the 7th, 14th, and 21st keywords are extracted and sorted, and then the keyword K3 at the top is selected.

In document B, the same operation is performed on the same keyword. In this case, three keywords can be extracted as in document A.

At this time, when the keywords are the same, the document A and the document B are judged to be the documents having the same probability.

Through the LSH method, it is possible to judge whether two messages are identical or not at a high speed even if two messages are different. If it is determined that the message transmitted through step S2 is identical to any one of the second messages, the step S53 of classifying the transmitted message into the second message is performed.

On the other hand, the first filtering unit 11, which receives from the second filtering unit 13 information indicating that the message transmitted as a result of step S2 is not classified as the second message, transmits the transmitted message as a first message (S10) whether or not to sort.

Through the preprocessing process, the system load can be reduced by classifying a message, which is sure to be spam, into a second message without judging whether the message corresponds to the first message. This will be described in detail with reference to Fig. 9 below.

Meanwhile, in another embodiment of the present invention, the spam filtering server 10 may include a filter learning module 14.

As described above, the spam filtering system employs a re-learning system to more accurately filter spam.

The filter learning module 14 in the present invention means a re-learning system for the first filtering unit 11, and a description thereof is shown in FIG.

Referring to FIGS. 1 and 8, the filter learning module 14 extracts content information of messages classified as the second message from the spam message database stored in the message server 30 or the second filtering unit 13, 2 < / RTI > message (S60).

Thereafter, the predetermined word stored in the first filtering unit 11 is updated using the extracted word (S70).

That is, the filter learning module 14 performs a function of re-learning the first filtering unit 11 rather than the second filtering unit 13 by using a message accurately classified as spam.

This makes it possible to improve the spam filtering performance of the first filtering unit 11 and to re-learn the first filtering unit 11 instead of simply learning the spam filtering again. And the function of the automatic spam filtering function can be enhanced in providing the pluralized spam filtering system.

Meanwhile, according to each configuration of the spam filtering server according to each of the embodiments of FIGS. 1 to 8 and the spam filtering method according to each embodiment of the present invention performed by the configurations, the transmitted message is finally classified into the second message Is determined.

At this time, the message server 30 receives information on the message classified as the second message. In this case, the message server 30 blocks the message classified as the second message from being transmitted to the external terminal 20, or transmits a message classified as the second message, and when the message is transmitted to the external terminal 20 To the spam message inbox provided in the message inbox of the external terminal 20. [

The message receiving box is an interface displayed on the external terminal 20 and means an interface through which a message received by the user is displayed according to the use of the message transmission / reception service by the user.

As described above, the spam filtering method according to the embodiment of the present invention can be executed by an application installed in a terminal (which may include a program included in a platform or an operating system basically installed in the terminal) May be executed by an application (i.e., a program) installed directly on the terminal via an application providing server such as an application store server, an application, or a web server associated with the service. In this sense, the spam filtering method according to the above-described embodiment of the present invention is implemented as an application (i.e., a program) installed basically in a terminal or directly installed by a user and recorded in a computer-readable recording medium such as a terminal .

Such a program may be recorded on a recording medium that can be read by a computer and executed by a computer so that the above-described functions can be executed.

As described above, in order to execute the spam filtering method according to each embodiment of the present invention, the above-mentioned program may be stored in a computer-readable code such as C, C ++, JAVA, ).

The code may include a function code related to a function or the like that defines the functions described above and may include an execution procedure related control code necessary for the processor of the computer to execute the functions described above according to a predetermined procedure.

In addition, such code may further include memory reference related code as to what additional information or media needed to cause the processor of the computer to execute the aforementioned functions should be referenced at any location (address) of the internal or external memory of the computer .

In addition, when a processor of a computer needs to communicate with any other computer or server, etc., to perform the above-described functions, the code may be stored in a computer's communication module (e.g., a wired and / ) May be used to further include communication related codes such as how to communicate with any other computer or server in the remote, and what information or media should be transmitted or received during communication.

The functional program for implementing the present invention and the related code and code segment may be implemented by programmers in the technical field of the present invention in consideration of the system environment of the computer that reads the recording medium and executes the program, Or may be easily modified or modified by the user.

Examples of the computer-readable recording medium on which the above-described program is recorded include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical media storage, and the like.

Also, the computer-readable recording medium on which the above-described program is recorded may be distributed to a computer system connected via a network so that computer-readable codes can be stored and executed in a distributed manner. In this case, any of at least one of the plurality of distributed computers may execute some of the functions presented above and transmit the result of the execution to at least one of the other distributed computers, and transmit the result The receiving computer may also perform some of the functions described above and provide the results to other distributed computers as well.

In particular, a computer-readable recording medium storing an application that is a program for executing the spam filtering method according to each embodiment of the present invention includes an application store server, an application or a web server related to the service A storage medium (e.g., a hard disk, etc.) included in the application provider server, or an application providing server itself.

A computer capable of reading a recording medium on which an application, which is a program for executing a spam filtering method according to each embodiment of the present invention, can be read by a computer such as a smart phone, a tablet PC, a PDA Digital assistants, and mobile communication terminals. In addition, the present invention should be interpreted as all devices capable of computing.

Further, the computer readable recording medium on which the application, which is a program for executing the spam filtering method according to the embodiment of the present invention, can be read is a mobile terminal such as a smart phone, a tablet PC, a PDA (Personal Digital Assistants) In this case, the application may be downloaded to the general PC from the application providing server and installed in the mobile terminal through the synchronization program.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. That is, within the scope of the present invention, all of the components may be selectively coupled to at least one. In addition, although all of the components may be implemented as one independent hardware, some or all of the components may be selectively combined to perform a part or all of the functions in one or a plurality of hardware. As shown in FIG. The codes and code segments constituting the computer program may be easily deduced by those skilled in the art. Such a computer program can be stored in a computer-readable storage medium, readable and executed by a computer, thereby realizing an embodiment of the present invention. As a storage medium of the computer program, a magnetic recording medium, an optical recording medium, or the like can be included.

It is also to be understood that the terms such as " comprises, "" comprising," or "having ", as used herein, mean that a component can be implanted unless specifically stated to the contrary. But should be construed as including other elements. All terms, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted to be consistent with the contextual meanings of the related art, and are not to be construed as ideal or overly formal, unless expressly defined to the contrary.

The foregoing description is merely illustrative of the technical idea of the present invention and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas falling within the scope of the same shall be construed as falling within the scope of the present invention.

Claims (29)

And storing the preset words and determining whether the transmitted message is likely to be spam based on whether or not a message transmitted from a message server providing a service for transmitting and receiving messages between terminals includes at least one of the preset words A first filtering unit for classifying the first message into a first message;
When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message and transmits the spam information; And
Wherein when the response information includes information that classifies a message classified as the first message into the second message, the response information includes information on the first request information, And a second filtering unit for classifying the message into the second message and storing the classified message in the spam message database. The method according to claim 1,
Wherein the second filtering unit comprises:
Upon receipt of second request information requesting to classify a message not classified as the first message from the external terminal into a second message, a message corresponding to the second request information, And the second message is classified into the second message. The method according to claim 1,
Wherein the second filtering unit comprises:
And stores the sender identification information of the message classified as the second message in the spam message database. The method of claim 3,
Wherein the second filtering unit comprises:
And classifying the message corresponding to the sender identification information stored in the spam message database into the second message when the message transmitted from the message server is a message corresponding to the sender identification information stored in the spam message database. Spam filtering server. The method according to claim 1,
Wherein the second filtering unit comprises:
When receiving from the message server information indicating that a message classified as the first message is repeatedly transmitted beyond a predetermined first threshold number, transmitting a message repeatedly transmitted in excess of the predetermined first threshold number to the second message Wherein the spam filtering server classifies the spam filtering server into spam filtering servers. The method according to claim 1,
Wherein the second filtering unit comprises:
When receiving from the message server information indicating that the number of transmissions of the same message as the message in which the response message from the external terminal does not exist for the previously transmitted message exceeds a predetermined second threshold number, Wherein the second message classifies a message whose number of origination times of the same message as the message that does not exist in the response message from the external terminal exceeds a predetermined second threshold number. The method according to claim 1,
Wherein the second filtering unit comprises:
Determining whether the transmitted message is the same as any one of the second messages stored in the spam message database, and if the transmitted message is the same message as the second message, Is stored in the spam message database. 8. The method of claim 7,
Wherein the second filtering unit comprises:
If the transmitted message is a message different from the second messages stored in the spam message database, the transmitted message and the second message stored in the spam message database are compared using a Locality-Sensitive Hashing And determines whether to classify the transmitted message as the second message according to the comparison result. 9. The method of claim 8,
Wherein the first filtering unit comprises:
As a result of comparing the transmitted message from the second filtering unit with the second message stored in the spam message database using the LSH scheme, information indicating that the transmitted message is not classified as the second message is received And classifies the transmitted message as a first message according to whether the transmitted message includes at least one of the preset words. The method according to claim 1,
Wherein the first filtering unit comprises:
And calculating a spam probability according to whether or not at least one of the preset words is included in the transmitted message, and classifying the transmitted message as the first message when the calculated spam probability exceeds a preset threshold probability Wherein the spam filtering server is a spam filtering server. The method according to claim 1,
The message server comprising:
And blocks a message classified as the second message from being transmitted to the external terminal. The method according to claim 1,
The message server comprising:
Wherein the message is transmitted to a spam message inbox provided in a message inbox of the external terminal when the message classified as the second message is transmitted to the external terminal. The method according to claim 1,
The external terminal,
A user terminal using a message transmission / reception service provided by the message server, and an administrator terminal of the message server. The method according to claim 1,
Extracting content information of a second message stored in the spam message database, extracting at least one word included in the second message, and extracting at least one word from the at least one word stored in the first filtering unit Further comprising: a filter learning module for updating the words. The spam filtering server,
According to whether or not a message transmitted from a message server providing a service for sending and receiving a message between terminals includes at least one of pre-stored preset words, the transmitted message is divided into a first message ;
When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message to be transmitted; And
When the response information of the external terminal for the first request information includes information for classifying a message classified as the first message into the second message, And storing the classified spam message in a spam message database. 16. The method of claim 15,
Upon receipt of second request information requesting to classify a message not classified as the first message from the external terminal into a second message, a message corresponding to the second request information, Further comprising: classifying the first message into the second message. 16. The method of claim 15,
Wherein the step of storing in the spam message database comprises:
Wherein the sender identification information of the message classified as the second message is stored in the spam message database. 18. The method of claim 17,
Classifying a message corresponding to the sender identification information stored in the spam message database into the second message when the message transmitted from the message server is a message corresponding to the sender identification information stored in the spam message database; Wherein the spam filtering method further comprises: 16. The method of claim 15,
Wherein the step of storing in the spam message database comprises:
When receiving from the message server information indicating that the message classified as the first message is repeatedly transmitted beyond a predetermined first threshold number, transmitting a message repeatedly transmitted exceeding the predetermined first threshold number to the external terminal And the second message is classified into the second message and stored in the spam message database. 16. The method of claim 15,
When receiving from the message server information indicating that the number of transmissions of the same message as the message in which the response message from the external terminal does not exist for the previously transmitted message exceeds a predetermined second threshold number, Further comprising classifying a message whose number of times of transmission of the same message as that of the message having no response message from the external terminal exceeds a predetermined second threshold number as the second message Way. The spam filtering server,
Determining whether to classify a message transmitted from a message server providing a service for sending and receiving messages between terminals to a second message that is a spam message;
Classifying the transmitted message into a first message, which is a message with a possibility of spam, according to whether the transmitted message is not classified as the second message, whether or not the transmitted message includes at least one of pre-stored preset words;
When the message classified by the first message is transmitted from the message server to the external terminal, inquiring whether to classify the message classified as the first message into the second message as the spam message, 1 message to be transmitted; And
When the response information of the external terminal for the first request information includes information for classifying a message classified as the first message into the second message, And storing the classified spam message in a spam message database. 22. The method of claim 21,
The step of determining whether to classify into a second message comprises:
Upon receipt of second request information requesting to classify a message not classified as the first message from the external terminal into a second message, a message corresponding to the second request information, Is classified into the second message. 22. The method of claim 21,
Wherein the step of storing in the spam message database comprises:
Wherein the sender identification information of the message classified as the second message is stored in the spam message database. 24. The method of claim 23,
The step of determining whether to classify into a second message comprises:
And classifying the message corresponding to the sender identification information stored in the spam message database into the second message when the message transmitted from the message server is a message corresponding to the sender identification information stored in the spam message database. Spam filtering methods. 22. The method of claim 21,
The step of determining whether to classify into a second message comprises:
When receiving from the message server information indicating that the number of transmissions of the same message as the message in which the response message from the external terminal does not exist for the previously transmitted message exceeds a predetermined second threshold number, Wherein the second message classifies a message whose number of times of transmission of the same message as the message in which the response message from the external terminal does not exist exceeds a predetermined second threshold number. 22. The method of claim 21,
The step of determining whether to classify into a second message comprises:
And classifying the same message as the second message when the transmitted message is the same as any one of the second messages stored in the spam message database. 27. The method of claim 26,
The step of determining whether to classify into a second message comprises:
If the transmitted message is a message different from the second messages stored in the spam message database, the transmitted message and the second message stored in the spam message database are compared using a Locality-Sensitive Hashing And determines whether to classify the transmitted message into the second message according to the comparison result. 22. The method of claim 21,
Wherein the message server comprises:
And blocking a message classified as the second message from being transmitted to the external terminal. A computer-readable recording medium on which a program for implementing the method of claim 15 or 21 is recorded.

Images