KR20170006288A - Apparatus and method for analyzing voice phishing pattern based on probability - Google Patents

Abstract

Provided are an apparatus and a method for analyzing a voice phishing pattern based on probability. The apparatus for analyzing a voice fishing pattern based on probability includes: a communication unit for receiving text data from an external request apparatus and transmitting the text data to a keyword analyzing unit, receiving voice phishing pattern information and a voice phishing probability value transmitted from a voice phishing pattern detecting unit, and transmitting the voice phishing pattern information and the voice phishing probability value to the request apparatus; a keyword analysis unit for recognizing an identifier of a voice phishing pattern corresponding to the text data received from the communication unit with reference to a keyword set list stored in a voice phishing pattern database and transmitting a keyword corresponding to the text data to the voice phishing probability analyzing unit; the voice phishing probability analyzing unit for calculating a voice phishing probability value based on the identifier of the voice phishing pattern received from the keyword analyzing unit and the keyword; a voice phishing pattern detecting unit for receiving the voice phishing pattern identifier and the voice phishing probability value from the voice phishing probability analyzing unit to determine a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing probability value, and transmitting voice phishing pattern information about the determined voice phishing pattern and the voice phishing probability value to the communication unit; and a voice phishing pattern database for storing the keyword set list about the voice phishing pattern.

Description

[0001] APPARATUS AND METHOD FOR ANALYZING VOICE PHISHING PATTERN BASED ON PROBABILITY [0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and method for analyzing a voice phishing pattern, and more particularly, to an apparatus and method for analyzing a voice phishing pattern capable of analyzing a voice phishing pattern more accurately and efficiently based on probability.

"Voice Phishing" refers to a fraudulent method of impersonating a relevant organization via telephone and extracting personal information or inducing remittance. As voice phishing becomes a social issue and people are aware of the danger of voice phishing to some degree, voice phishing techniques are also evolving, and recent cases have been reported.

As a conventional technique for preventing voice phishing, there is known a service in which a telephone number used for voice phishing is stored in a communication service provider network and the call is originally blocked or informed to a called terminal. However, In this case, it takes much time and effort to acquire initial data. In addition, such information can only be confirmed on the receiving screen before receiving the call, and there is a limit to that the user can not be provided with the risk of voice phishing after receiving the call, that is, during communication.

In addition, as the voice phishing method evolves, a new voice phishing method appears. It is difficult for the user to know how to cope with all the voice phishing cases beforehand, and the damage occurs due to the momentary vigilance. It is not possible to provide such a limit.

Also, a method of storing a voice phishing suspicious phone number in a user terminal has been proposed, but the same problem also exists. Furthermore, there is also a problem that a telephone necessary for an actual called party is marked as spam, so that the reception of the telephone is cut off, or accurate information can not be provided.

In addition, the conventional art does not provide an algorithm that can specifically determine whether or not voice phishing is performed. Therefore, there is a limitation in that it is not possible to implement a device capable of determining whether or not voice phishing is performed in real time based on the contents of a call.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems and it is an object of the present invention to provide an apparatus and method for preventing voice phishing from occurring in real time by verifying whether voice phishing is occurring during a call.

It is another object of the present invention to provide a more accurate and efficient voice phishing analysis apparatus and method by determining the presence or absence of voice phishing based on the mathematical probability theory and increasing the probability of detecting voice phishing, do.

It is another object of the present invention to provide an apparatus and method for distinguishing voice phishing by taking into consideration the relationship between a keyword of a call content and voice phishing.

It is another object of the present invention to provide an apparatus and method for increasing the accuracy of voice phishing diagnosis by discriminating voice phishing using the concept of conditional probability together with the order and continuity of keywords appearing in voice phishing .

According to another aspect of the present invention, there is provided an apparatus for analyzing a voice phishing pattern based on probability, comprising: a text analyzing unit that receives text data from an external requesting apparatus and transmits the text data to a keyword analyzing unit, A communication unit for receiving the information and the voice phishing probability value and transmitting the received information to the requesting apparatus; A keyword analyzer for identifying an identifier of a voice phishing pattern corresponding to the text data received from the communication unit by referring to a keyword set list stored in the voice phishing pattern database and transmitting the keyword corresponding to the text data to the voice phishing probability analyzer; A voice phishing probability analyzing unit for calculating a voice phishing probability value based on the identifier of the voice phishing pattern received from the keyword analyzing unit and the keyword; A voice phishing pattern identifier and a voice phishing probability value from the voice phishing probability analyzer to determine a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing probability value, and voice phishing pattern information and voice phishing probability value for the determined voice phishing pattern, A pattern detection unit; And a voice phishing pattern database storing a keyword set list for the voice phishing pattern.

Here, the keyword set list may include at least one or more keyword sets for each voice phishing pattern, an identifier for the voice phishing pattern, and the total number of keyword sets for each voice phishing pattern.

The keyword analyzing unit may refer to a keyword set list of a voice phishing pattern database to identify a voice phishing pattern identifier corresponding to a set of keywords including received text data and a keyword corresponding to the received text data, To-be-transmitted.

Also, the voice phishing probability analyzing unit may include:

(여기서, n은 보이스 피싱 패턴의 총 키워드 집합의 갯수, r은 현재까지 키워드 분석부(20)로부터 키워드가 전송된 횟수)의 수식에 의하여 보이스 피싱 확률값(P)을 계산할 수 있다.

(Where n is the number of total keyword sets of the voice phishing pattern, r is the number of times the keyword has been transmitted from the keyword analyzing unit 20).

Also, the voice phishing probability analyzer may increase r by 1 every time a keyword is delivered until the analysis of the text data is completed, with r = 1 when the keyword is transmitted from the keyword analyzing unit.

In addition, the voice phishing probability analyzing unit may calculate each voice phishing pattern probability value P for each voice phishing pattern using the above formula when there are a plurality of identifiers of the voice phishing patterns.

The keyword set list may further include a predetermined weight for each keyword set.

Also, the voice phishing probability analyzing unit may include:

(Where n is the number of total keyword sets of the voice phishing pattern, r is the number of times the keyword has been transmitted from the keyword analyzing unit 20, and w is a cumulative weight which is the sum of the weights of the keywords detected so far) The voice phishing probability value P may be calculated.

According to another aspect of the present invention, there is provided a method of analyzing a voice phishing pattern performed in a voice phishing pattern apparatus as described above, the method comprising: a first step of receiving a text data from an external request apparatus and transmitting the received text data to a keyword analyzing unit; A second step of discriminating an identifier of a voice phishing pattern corresponding to the received text data by referring to a keyword set list stored in a voice phishing pattern database and delivering a keyword corresponding to the text data to a voice phishing probability analyzing unit; A third step of calculating a probability value for voice phishing based on the identifier and the keyword of the received voice phishing pattern; The voice phishing pattern detection unit receives the voice phishing pattern identifier and the voice phishing pattern probability value from the voice phishing probability analyzing unit, determines a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing pattern probability value, A fourth step of transmitting a probability value to a communication unit; And a fifth step of transmitting the voice phishing pattern information and the voice phishing probability value received from the voice phishing pattern sensing unit to the requesting unit.

According to another aspect of the present invention, there is provided an apparatus for analyzing a voice phishing pattern based on probability, comprising: a text analyzing unit for receiving text data from an external requesting apparatus and transmitting the text data to a keyword analyzing unit; A communication unit for receiving the probability value and delivering it to the requesting device; A keyword analyzer for referring to a keyword set list stored in a voice phishing pattern database and transmitting a keyword corresponding to text data received from the communication unit and a phishing index for the keyword to a voice phishing probability analyzer; A voice phishing probability analyzing unit for calculating a voice phishing probability value based on the keyword and the phishing index received from the keyword analyzing unit; A voice phishing pattern detection unit for receiving a keyword and a voice phishing probability value from the voice phishing probability analysis unit, determining a voice phishing pattern based thereon, and transmitting voice phishing pattern information and a voice phishing probability value for the determined voice phishing pattern to a communication unit; And a voice phishing pattern database storing a keyword set list for the voice phishing pattern.

Here, the phishing index is a probability value indicating the association of voice phishing for each keyword.

Also, the voice phishing probability analyzing unit may include:

(Where A to N are the phishing indices of the keywords), the voice phishing probability value P can be calculated.

In addition, the voice phishing pattern detecting unit may determine whether a voice phishing pattern or voice phishing is based on at least one of a sequence of occurrence, continuity, a number of times, and a probability value of a keyword.

According to another aspect of the present invention, there is provided a method for analyzing a voice phishing pattern performed in an apparatus for analyzing a voice phishing pattern, comprising: a first step of receiving a text data from an external request apparatus and transmitting the received text data to a keyword analyzing unit; A second step of referring to the keyword aggregation list stored in the voice phishing pattern database and transmitting the keyword corresponding to the text data received from the communication unit and the phishing index for the keyword to the voice phishing probability analyzer; A third step of calculating a voice phishing probability value based on the keyword and the phishing index received from the keyword analyzing unit; A voice phishing pattern detection unit receives a keyword and a voice phishing probability value from the voice phishing probability analysis unit, determines a voice phishing pattern based on the keyword, and transmits voice phishing pattern information and a voice phishing probability value for the determined voice phishing pattern to a communication unit Step 4; And a fifth step of receiving the voice phishing pattern information and the voice phishing probability value transmitted from the voice phishing pattern sensing unit and transmitting the received voice phishing pattern information and the voice phishing probability value to the requesting device.

According to the present invention, it is possible to provide an apparatus and method for preventing voice phishing damage in advance by confirming whether voice phishing is performed in real time during a call.

In addition, the present invention can increase the probability of voice phishing detection by discriminating whether or not voice phishing is performed based on mathematical probability theory, and can provide a more accurate and efficient voice phishing analysis apparatus and method by converting a voice phishing pattern into a database.

In addition, the present invention can provide an apparatus and method for discriminating voice phishing by taking into account the relationship between keywords of a call content and voice phishing.

In addition, the present invention can provide an apparatus and method for increasing the accuracy of voice phishing diagnosis by discriminating voice phishing using the concept of conditional probability together with the order and continuity of keywords appearing in voice phishing.

In addition, the present invention is applicable not only to the field of voice phishing but also to the field of spam phone, advertisement, and the like, and has an effect of analyzing patterns of conversation contents.

1 is a diagram illustrating an embodiment of an apparatus 100 for analyzing a voice phishing pattern based on probability according to the present invention.
FIG. 2 is a diagram for explaining a keyword set list stored in the voice phishing pattern database 50. FIG.
FIG. 3 is a flowchart illustrating an embodiment of a method for analyzing a voice phishing pattern based on a probability performed by the analyzing apparatus 100 described with reference to FIGS. 1 and 2. FIG.
FIG. 4 shows another example of the keyword set list, and is a diagram for explaining another embodiment of the present invention.
5 is a view for explaining another embodiment of the voice phishing pattern database 50 according to the present invention.
6 is a flowchart illustrating a method of analyzing a voice phishing pattern according to another embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating an embodiment of an apparatus 100 for analyzing a voice phishing pattern based on probability according to the present invention.

1, a probability phonetic phonetic pattern analyzer 100 (hereinafter simply referred to as "analyzer 100"), a communication unit 10, a keyword analyzer 20, a voice phishing probability analyzer 30, A voice phishing pattern detecting unit 40, and a voice phishing pattern database 50. [

The communication unit 10 receives the text data from the external request device 200 and transmits the text data to the keyword analysis unit 20 and receives the voice phishing pattern information and the voice phishing probability value transmitted from the voice phishing pattern sensing unit 50 And transmits it to the requesting device 200.

Here, the request apparatus 200 may be a mobile communication terminal having a function of transmitting text data to the analysis apparatus 100 and generating text data from the conversation contents by a speech recognition function. In this case, the mobile communication terminal is automatically activated when a call is performed in cooperation with the voice phishing pattern analyzing apparatus 100 according to the present invention to perform a voice recognition from a voice voice signal, a voice phishing pattern analyzing apparatus 100, It is preferable that the smartphone is a smart phone having a function of transmitting and receiving in the form of an application.

The keyword analyzing unit 20 refers to the keyword set list stored in the voice phishing pattern database 50 to determine the identifier of the voice phishing pattern corresponding to the text data received from the communication unit 10, And transmits the keyword to the voice phishing probability analyzer 20.

The voice phishing pattern database 50 is a means for storing a keyword set list for a voice phishing pattern.

FIG. 2 is a diagram for explaining a keyword set list stored in the voice phishing pattern database 50. FIG.

Referring to FIG. 2, the keyword set list includes at least one set of keywords for each voice phishing pattern, and an identifier (ID) is assigned to each voice phishing pattern. In addition, the total number of keyword sets for each voice phishing pattern is calculated and stored in advance.

Each keyword set for the voice phishing pattern is made up of keywords that are correlated with each other, and each keyword is a pre-classified keyword that typically appears in the corresponding voice phishing pattern.

On the other hand, each keyword set includes similar keywords, and the similar keywords indicate keywords that are different in spelling but can be judged to be similar due to difference in pronunciation.

For example, the "Prosecutors' Office, Seoul Prosecutors' Office, Prosecutors" constitute a keyword set 1 of the prosecution office's false-voice phishing patterns as typical keywords in the prosecution offices' Also, a keyword such as "impersonation, account" constitutes a keyword set 4 of the false-voice phishing pattern of the prosecution office. In Fig. 2, the total number of keyword sets of the "prosecuting attorney-type" voice phishing pattern is four.

The keyword analyzing unit 20 refers to the keyword set list of the voice phishing pattern database 50 as shown in FIG. 2 to determine which keyword set of the voice phishing pattern the received text data is included in. When the received text data is included in a specific keyword set of a specific voice phishing pattern, the keyword is discriminated as an identifier of the corresponding voice phishing pattern and transmitted to the voice phishing probability analyzer 30.

Here, the received text data may be included in a set of keywords of a plurality of voice phishing patterns, because a specific keyword may appear commonly in a plurality of sets of keyword phishing patterns. In this case, the voice phishing probability analyzer 30 transmits all the identifiers of the plurality of voice phishing patterns to the voice phishing probability analyzer 30.

For example, when the received text data is "impersonation ", it is included in the keyword set 4 of the false office of the Attorney's Office in FIG. 2 and is also included in the post-office false keyword set 3, And transmits the identifier AA of the prosecution office to the voice phishing probability analyzer 30 as the identifier of the post office false name type.

The voice phishing probability analyzing unit 30 is responsible for calculating a voice phishing probability value based on the identifier of the voice phishing pattern received from the keyword analyzing unit 20 and the keyword.

The voice phishing probability analyzing unit 30 first stores the voice phishing pattern identifier and the keyword transmitted from the keyword analyzing unit 20 and confirms the number of times the keyword is transmitted from the keyword analyzing unit 20 until now. Then, referring to the voice phishing pattern database 50, the number (n) of keyword sets of the voice phishing pattern is confirmed. Here, r is the same as the number of sets of keywords detected so far, and the voice phishing probability analyzer 30 sets r = 1 when the keyword is delivered from the keyword analyzer 20, Increase the r value by 1 each time you receive a keyword until it is complete.

Then, using this, a voice phishing probability value P is calculated using the following equation.

Equation (1) means that probability values are calculated using redundant permutation.

For example, when the number of the total keyword sets of the prosecution offices in FIG. 2 is 4 (= n), the keyword "Prosecutor's Office" is transmitted from the keyword analyzing unit 20 for the first time, , And the voice phishing probability value (P) is 18.75%.

Thereafter, if the second keyword is discriminated from the keyword analyzing unit 20 (r = 2), the voice phishing probability value P is calculated according to the above formula 43.75%, and the further third keyword is determined (r = 3), the probability of voice phishing (P) becomes 90.63%.

If there are a plurality of identifiers of the voice phishing patterns transmitted from the keyword analyzing unit 20, the voice phishing probability analyzing unit 30 calculates the respective voice phishing probability values P using the above formula for each voice phishing pattern .

The voice phishing probability analyzer 30 transmits the calculated probability value and the voice phishing pattern identifier to the voice phishing pattern detector 40 when the voice phishing probability P is calculated.

The voice phishing pattern detecting unit 40 determines a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing probability value received from the voice phishing probability analyzing unit 30 and outputs voice phishing pattern information for the determined voice phishing pattern, And transmits the phishing probability value to the communication unit 10.

When a voice phishing pattern identifier and a probability value are received, the voice phishing pattern detector 40 transmits the voice phishing pattern identifier and the probability value to the communication unit 10 as it is. When a plurality of voice phishing pattern identifiers and a probability value are received, And transmits the voice phishing pattern identifier and the probability value to the communication unit 10.

FIG. 3 is a flowchart illustrating an embodiment of a method of analyzing a voice phishing pattern based on the probability performed by the analyzing apparatus 100 described with reference to FIGS. 1 and 2. FIG.

Referring to FIG. 3, the requesting apparatus 200 transmits text data to the communication unit 10 of the analyzing apparatus 100 (S100). The text data may be transmitted one by one or a plurality of sets of text data may be transmitted at a time.

When the communication unit 10 receives the text data from the external request apparatus 200, the communication unit 10 transmits the text data to the keyword analysis unit 20 (S110).

When receiving the text data, the keyword analyzing unit 20 refers to the keyword set list stored in the voice phishing pattern database 50, determines the identifier of the voice phishing pattern corresponding to the text data received from the communication unit 20, And transmits the keyword corresponding to the text data to the voice phishing probability analyzing unit 30 (S120, S130).

When the voice phishing probability analyzing unit 30 receives the identifier and the keyword, the voice phishing probability analyzing unit 30 calculates the voice phishing probability value P by the above-described method (S140).

When the probability value P is calculated, the voice phishing probability analyzing unit 30 transmits the voice phishing pattern identifier and the probability value to the voice phishing pattern detecting unit 40 (S150).

Upon receiving the voice phishing pattern identifier and the voice phishing probability value from the voice phishing probability analyzing unit 30, the voice phishing pattern detecting unit 40 detects the voice phishing pattern identifier and the voice phishing probability value in the manner described above, (S160), and transmits the voice phishing pattern information and the voice phishing probability value for the determined voice phishing pattern to the communication unit 10 (S170).

The communication unit 10 transmits the received voice phishing pattern information and the probability value to the requesting device 200 (S180).

FIG. 4 shows another example of the keyword set list, and is a diagram for explaining another embodiment of the present invention.

Fig. 4 is basically the same as the keyword set list of Fig. 2, but differs in that a weight value is preset in each keyword set.

This weight is set in consideration of importance for each keyword set. This is for calculating a probability value of a voice phishing pattern to reflect the importance of the keyword sets.

In consideration of this weight, the voice phishing probability analyzer 30 may calculate the voice phishing probability value using the following equation.

Here, r and n are the same as in the above-described embodiment. That is, r is the number (r) of times the keyword has been transmitted from the keyword analyzing unit 20 until now, and n is the number (n) of sets of keywords of the voice phishing pattern.

Meanwhile, w denotes a cumulative weight. The cumulative weight means a value obtained by accumulating each weight added to a set of keywords to which the keyword belongs when the keyword is transmitted.

For example, if the total number of keyword aggregations of the prosecutor's office impersonation pattern is 4 and one keyword is detected for the first time and the weight at this time is 10%, the probability value is 7.5% according to Equation 2, Is 20%, the cumulative weight is 30%. And the probability value at this time is 26.25%. Next, if the third keyword is detected and the weight is 30%, the cumulative weight is 60% and the probability value is 54.37%. If the fourth keyword is detected and the weight is 40%, the cumulative weight becomes 100%, and the probability value at this time becomes 90.63%.

5 is a view for explaining another embodiment of the voice phishing pattern database 50 according to the present invention.

FIG. 5 shows another example of the keyword set list stored in the voice phishing pattern database 50, which is provided in the form of phishing indexes for keywords used for voice phishing and keywords.

Here, the phrase " phishing index " means a probability value corresponding to voice phishing when the keyword appears. That is, the correlation between the keyword and voice phishing is represented by a probability value. The phishing index can analyze a large number of voice phishing cases and statistically calculate the probability that the keyword exists in a voice phishing case.

The operation of the analysis apparatus 100 when such a keyword set list is used will be described with reference to FIG.

6 is a flowchart illustrating a method of analyzing a voice phishing pattern according to another embodiment of the present invention.

Referring to FIG. 6, the keyword analyzing unit 20 receives text data in the same manner as described with reference to FIG. 3 (S100, S110).

The keyword analyzing unit 20 refers to the voice phishing pattern database 50 to check whether the text data received from the communication unit 10 exists in the keyword set list (S220).

If there is text data in the keyword set list, the keyword analyzing unit 20 transmits the keyword and the phishing index to the voice phishing probability analyzing unit 30 (S230)

The voice phishing probability analyzing unit 30 calculates a voice phishing probability value using the received keyword and the phishing index using the following equation (S240).

Here, A to N denote the phishing indexes of the keywords displayed so far.

The above expression means a conditional probability that a condition is satisfied when a plurality of keywords are present and means that a voice phishing probability value is calculated by a conditional probability that the condition exists when the corresponding keywords exist.

As in the above-described embodiment, the voice phishing probability analyzing unit 30 calculates the continuity probability value by the above-mentioned Equation 3 every time the text data is received until the text data is no longer received from the keyword analyzing unit 20 do. For this, the voice phishing probability analyzer 30 stores the previously received keywords and calculates the cumulative probability value according to Equation (3) every time a new keyword is received.

When the voice phishing probability value is finally calculated, the voice phishing probability analyzing unit 30 transmits the keyword and the calculated probability value to the voice phishing pattern detecting unit 40 (S250).

The voice phishing pattern detecting unit 40 determines a voice phishing pattern based on the received keyword and the probability value (S260).

Here, the voice phishing pattern can be determined by comprehensively considering the appearance order, continuity, number of times, and probability value of the keyword.

For example, if certain keywords appear in the same order, if the probability of voice phishing is known to be high, the weighting can be given when certain keywords appear in the same order.

In addition, a weight can be given to the total number of times the keyword is displayed.

In addition, weights can also be given to the probability values calculated by Equation (3).

In this way, the voice phishing pattern detecting unit 40 can finally determine a voice phishing pattern (or voice phishing) considering a plurality of conditions comprehensively.

When the voice phishing pattern (or voice phishing) is determined, the voice phishing pattern detecting unit 40 transmits the voice phishing pattern information (or voice phishing information) and the voice phishing probability value for the determined pattern to the communication unit 10 S170), and the communication unit 10 transfers the received information to the request apparatus 200 (S180).

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments.

For example, the methods of Equations (1) to (3) described as methods for calculating the probability of voice phishing in the voice phishing probability analyzer 30 can be used in combination with each other.

Also, in the above embodiment, the request device 200 is a mobile communication terminal having a voice recognition function, but may be a separate voice recognition server for performing voice recognition. In this case, the mobile communication terminal transmits a voice signal to the voice recognition server, and the voice recognition server performs voice recognition to transmit the text data to the voice phishing pattern analyzer 100 according to the present invention.

In addition, the present invention can be applied to various advertisements, spam phones, etc., rather than voice phishing. For example, keywords used in advertisements or spam telephones are databaseized in advance, and keywords are analyzed using the databases to provide information about advertisements or spam phones to users can do.

100 ... Voice phishing pattern analyzer
10 ... communication unit
20 ... keyword analysis unit
30 ... Voice Phishing Probability Analysis Unit
40 ... voice phishing pattern detection unit
50 ... Voice phishing pattern database
200 ... request device

Claims (14)

As a probability phonetic pattern analyzing apparatus,
A communication unit for receiving text data from an external requesting device and delivering the text data to the keyword analyzing unit, receiving the voice phishing pattern information and the voice phishing probability value transmitted from the voice phishing pattern detecting unit, and transmitting the received voice phishing pattern information and voice phishing probability value to the requesting device;
A keyword analyzer for identifying an identifier of a voice phishing pattern corresponding to the text data received from the communication unit by referring to a keyword set list stored in the voice phishing pattern database and transmitting the keyword corresponding to the text data to the voice phishing probability analyzer;
A voice phishing probability analyzing unit for calculating a voice phishing probability value based on the identifier of the voice phishing pattern received from the keyword analyzing unit and the keyword;
A voice phishing pattern identifier and a voice phishing probability value from the voice phishing probability analyzer to determine a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing probability value, and voice phishing pattern information and voice phishing probability value for the determined voice phishing pattern, A pattern detection unit; And
A voice phishing pattern database storing a keyword set list of voice phishing patterns
Wherein the voice phishing pattern analyzer comprises:
The method according to claim 1,
Wherein the keyword set list includes at least one or more keyword sets for each voice phishing pattern, an identifier for the voice phishing pattern, and a total number of keywords for each voice phishing pattern.
3. The method of claim 2,
The keyword analyzing unit refers to the keyword set list of the voice phishing pattern database and transmits an identifier of a voice phishing pattern corresponding to a set of keywords including received text data and a keyword corresponding to the received text data to a voice phishing probability analyzing unit Wherein the voice phishing pattern analyzing apparatus comprises:
The method according to claim 1,
Wherein the voice phishing probability analyzing unit comprises:

(Where n is the number of total keyword sets of the voice phishing pattern, and r is the number of times the keyword has been transmitted from the keyword analyzing unit 20)
Wherein the voice phishing probability value (P) is calculated by the following equation.
5. The method of claim 4,
Wherein the voice phishing probability analyzer increases r by 1 every time a keyword is delivered until the analysis of the text data is completed by setting r = 1 when the transmission of the keyword is started from the keyword analyzing unit, Pattern analyzer.
6. The method of claim 5,
Wherein the voice phishing probability analyzing unit calculates each voice phishing pattern probability value (P) for each voice phishing pattern using a plurality of identifiers of the voice phishing pattern using the formula.
3. The method of claim 2,
Wherein the keyword set list further includes a predetermined weight for each keyword set.
8. The method of claim 7,
Wherein the voice phishing probability analyzing unit comprises:

(Where n is the number of the total keyword sets of the voice phishing pattern, r is the number of times the keyword has been transmitted from the keyword analyzing unit 20, and w is a cumulative weight which is the sum of the weights of the keywords detected so far)
Wherein the voice phishing probability value (P) is calculated by the following equation.
9. A method for analyzing a voice phishing pattern performed in an apparatus according to any one of claims 1 to 8,
A first step in which the communication unit receives text data from an external request apparatus and transfers the text data to the keyword analysis unit;
A second step of discriminating an identifier of a voice phishing pattern corresponding to the received text data by referring to a keyword set list stored in a voice phishing pattern database and delivering a keyword corresponding to the text data to a voice phishing probability analyzing unit;
A third step of calculating a probability value for voice phishing based on the identifier and the keyword of the received voice phishing pattern;
The voice phishing pattern detecting unit receives the voice phishing pattern identifier and the voice phishing pattern probability value from the voice phishing probability analyzing unit, determines a voice phishing pattern based on the voice phishing pattern identifier and the voice phishing pattern probability value, A fourth step of transmitting a probability value to a communication unit;
A fifth step of communicating the voice phishing pattern information and the voice phishing probability value received from the voice phishing pattern detector to the requesting device
And analyzing the voice phishing pattern.
As a probability phonetic pattern analyzing apparatus,
A communication unit for receiving text data from an external requesting device and delivering the text data to the keyword analyzing unit, receiving the voice phishing pattern information and the voice phishing probability value transmitted from the voice phishing pattern detecting unit, and transmitting the received voice phishing pattern information and voice phishing probability value to the requesting device;
A keyword analyzing unit for referring to a keyword set list stored in a voice phishing pattern database and transmitting a keyword corresponding to text data received from the communication unit and a phishing index for the keyword to a voice phishing probability analyzing unit;
A voice phishing probability analyzing unit for calculating a voice phishing probability value based on the keyword and the phishing index received from the keyword analyzing unit;
A voice phishing pattern detection unit for receiving a keyword and a voice phishing probability value from the voice phishing probability analysis unit, determining a voice phishing pattern based thereon, and transmitting voice phishing pattern information and a voice phishing probability value for the determined voice phishing pattern to a communication unit; And
A voice phishing pattern database storing a keyword set list of voice phishing patterns
Wherein the voice phishing pattern analyzer comprises:
11. The method of claim 10,
Wherein the phishing index indicates a correlation of voice phishing for each keyword as a probability value.
12. The method of claim 11,
Wherein the voice phishing probability analyzing unit comprises:

(Where A to N are phishing indices of keywords)
Wherein the voice phishing probability value (P) is calculated by the following equation.
13. The method of claim 12,
Wherein the voice phishing pattern detecting unit determines a voice phishing pattern or voice phishing based on at least one of a sequence of occurrence of a keyword, continuity, a number of times, and a probability value.
A method for analyzing a voice phishing pattern performed in an apparatus for analyzing a voice phishing pattern according to any one of claims 10 to 13,
A first step in which the communication unit receives text data from an external request device and transfers the text data to the keyword analysis unit;
A second step of referring to the keyword aggregation list stored in the voice phishing pattern database and transmitting the keyword corresponding to the text data received from the communication unit and the phishing index for the keyword to the voice phishing probability analyzer;
A third step of calculating a voice phishing probability value based on the keyword and the phishing index received from the keyword analyzing unit;
A voice phishing pattern detection unit receives a keyword and a voice phishing probability value from the voice phishing probability analysis unit, determines a voice phishing pattern based on the keyword and the voice phishing probability value, and transmits voice phishing pattern information and a voice phishing probability value for the determined voice phishing pattern to a communication unit Step 4; And
A fifth step of receiving the voice phishing pattern information and the voice phishing probability value transmitted from the voice phishing pattern detecting unit by the communication unit and transmitting the received information to the requesting device
And analyzing the voice phishing pattern.

Images