JP2015109027A - Information processing device, method and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce overlooking of a fraudulent application.SOLUTION: In a computer, a database manages history information associating a first key including a family name key corresponding to a family name and an address key corresponding to address characters with a second key resulting from converting or deleting a part of the first key. An application first key is generated which is an application family name key and an application address key corresponding to the family name and the address characters included in application information and includes the application family name key and the application address key, and an application second key resulting from converting or deleting a part of the application first key is generated. The history information including the first key or the second key matched with the application first key or the application second key is extracted from the database.

Description

  The present invention relates to information processing technology, and more particularly to technology for identifying similar information.

  There is a so-called capture fraud in which a product or the like is ordered from a company that sells the product or the like via the Internet, telephone or FAX, and the price is not paid even though the delivered product is received. Capture fraudsters (referred to as “scammers”) slightly change the address notation for each order to make it difficult for businesses to recognize them as the same person. For example, a fraudster uses different address notations for each order: “Shinjuku-ku, Shinjuku-ku 1-2-3” and “Shinjuku-ku 1-2-3, Tokyo”. In order to absorb such differences in address notation, a technique is disclosed in which all address character strings are converted into numerical values, and whether the converted numerical values match or not is determined to be the same address. (Patent Document 1). Thereby, even if it is a case where a fraudulent person uses the notation of an address intentionally as mentioned above, patent document 1 determines with it being the address of the same person.

JP 2008-65494 A

  In Patent Document 1, a history of address character strings for a certain period is registered in a database, and when new address character string information is received, the number of address character strings that match the registered history (history) Number). However, this can still overlook applications from fraudsters. Therefore, an object of the present invention is to reduce oversight of applications from fraudsters.

An information processing apparatus according to an embodiment of the present invention includes a first key including a last name key corresponding to a last name and an address key corresponding to an address character, and a second key obtained by converting or deleting a part of the first key. And a database for managing history information
A first key generating unit for generating a first key including an application last name key and an application address key corresponding to a last name and an address character included in the application information, respectively,
A second key generation unit for generating an application second key obtained by converting or deleting a part of the application first key;
And an extraction unit that extracts history information having a first key or a second key that matches the application first key or the application second key from the database.

  According to the present invention, oversight of orders from fraudsters can be reduced.

It is a block diagram which shows an example of the function structure which an information processing apparatus has. It is a figure which shows an example of a NAK key and a similar NAK key. The structural example of log | history information DB (DataBase) is shown. It is a schematic diagram which shows an example of the procedure which calculates a fraud score. The structural example of an application determination system is shown.

  Since the conventional name identification technique is a technique for identifying a person, there is a technique that corrects the degree of unification of the notation, but basically it is determined that they are the same when they completely match. For this reason, when trying to apply it to prevent fraud in Internet mail order sales etc., it is difficult to detect a case where the notation is deliberately false so that the notation is inconsistent. It was. In order to solve the problem, in the present invention, fuzzy matching is performed using a similar key adapted to an illegal pattern specific to the notation change in the capture fraud. Hereinafter, an example of an embodiment according to the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram illustrating an example of a functional configuration of the information processing apparatus.
The information processing apparatus 1 is a so-called computer including a CPU (central processing unit), a memory, a storage device (all not shown), and the like, and executes a predetermined computer program using the CPU, the memory, the storage device, and the like. Thus, various functions to be described later are realized.

  The information processing apparatus 1 functions as an application information acquisition unit 10, a NAK key generation unit 11, a similar NAK key generation unit 12, an extraction unit 13, a fraud score calculation unit 14, a warning unit 15, and history information. DB200.

  The application information acquisition unit 10 acquires application information 100, which is information related to application for products or services. The application information 100 includes, for example, the applicant's name, address, telephone number, product name, quantity, amount, and the like. The application information 100 may further include a credit card number owned by the applicant, terminal identification information used by the applicant for application, and the like.

  The application information 100 may be input by the applicant from the Internet shopping WEB site, or may be information input by the operator through telephone or FAX from the applicant.

  The application information acquisition unit 10 may acquire the application information 100 for each application, or may acquire a plurality of application information 100 collectively. The application information acquisition unit 10 passes the application information 100 to be calculated for the fraud score 180 to the NAK key generation unit 11. The fraud score 180 is a value indicating the possibility of fraud in the application information 100.

  FIG. 2 is a diagram illustrating an example of a NAK key and a similar NAK key. Hereinafter, the NAK key generation unit 11 and the similar NAK key generation unit 12 will be described with reference to FIG.

  The NAK key generation unit 11 generates a NAK key 120 from the application information 100. The NAK key generation unit 11 generates a NAK key from the last name and address character strings included in the application information 100. The NAK key generation unit 11 generates a NAK key 120 by unifying the notation fluctuations of the last name kanji (character string) included in the application information 100. For example, the NAK key generation unit 11 generates the NAK key 120 by unifying the last characters “Watanabe”, “Watanabe”, and “Watanabe” into “Watanabe”.

  The NAK key generation unit 11 converts the address character string into a numerical value and generates the NAK key 120 in order to absorb the difference in the address notation method included in the application information 100. For example, the NAK key generation unit 11 converts the character strings of the addresses of “2-3405 Higashi Gouencho, Shinjuku-ku, Tokyo” and “2-3-405 Higashi Gouen-cho, Shinjuku-ku, Tokyo” to “1620813, NAK key 120 is generated by converting the value into “2, 3, 405”. Here, “1620813” is the postal code of “Higashi Goken-cho, Shinjuku-ku, Tokyo”.

  That is, as shown in FIG. 2, when the name of the application information 100 is “Ichiro Watanabe” and the address is “2-3405 Higashi Gokencho, Shinjuku-ku, Tokyo”, the NAK key generation unit 11 The NAK key 120 “Watanabe, 1620813, 2, 3, 405” is generated.

  The similar NAK key generation unit 12 generates a plurality of similar NAK keys 140 from the NAK key 120. The similar NAK key 140 is data obtained by converting or deleting a part of the NAK key 120.

  The similar NAK key generation unit 12 converts the NAK key 120 in consideration of the trick of the fraudster, and generates a similar NAK key 140. For example, the similar NAK key generation unit 12 generates a similar NAK key 140 by converting or deleting a part of the NAK key 120 within a range in which a product or the like can be delivered. For example, the similar NAK key generation unit 12 has a similar NAK key corresponding to a surname and an address that may be deliberately converted so that a fraudulent person can deliver a product or the like without being recognized as the same person. 140 is generated.

  For example, the similar NAK key generation unit 12 generates a similar NAK key 140 in which the surname is changed to a reading kana. The reason is as follows. Assume that a fraudster deliberately uses “Watanabe”, “Watanabe”, and “Watanabe” as the surname of the delivery destination. In this case, in the comparison of the normal NAK key 120, “Watanabe” and “Watanabe” are unified as “Watanabe” and are recognized as the same person, but “Watanabe” is recognized as another person. However, if a fraudster has written “WATANABE” in Roman letters on the nameplate, the deliveryman will look at the nameplate and check the fraudulent product, regardless of whether the surname is “Watanabe” or “Watanabe”. May be delivered to the person. Therefore, the similar NAK key generation unit 12 reads the surname keys “Watanabe” and “Watanabe” of the NAK key 120 so that information with the possibility of the same person can be extracted even for the input of such a fraudster. A similar NAK key 140 including the surname key converted to the pseudonym “WATANABE” is generated.

  For example, the similar NAK key generation unit 12 generates a similar NAK key 140 in which the address and / or room number is omitted. The reason is as follows. The scammers deliberately set the delivery address as “2-3405 Higashi Gouencho, Shinjuku-ku, Tokyo”, “2-3-405 Higashi Gouen-cho, Shinjuku-ku, Tokyo” Suppose you have used “Gokencho 2-3” properly. In this case, in comparison with the normal NAK key 120, both “2-3405 Higashi Gouencho, Shinjuku-ku, Tokyo” and “2-3-405 Higashi Gouen-cho, Shinjuku-ku, Tokyo” are both “1620813, "2,3,405" will be recognized as the same person, but "Higashi Gokencho 2-3, Shinjuku-ku, Tokyo" will be recognized as a different person because it will be converted to "1620813,2,3" . However, even if the delivery destination room number “405” is omitted, the delivery person may guess the name and the like and deliver the product to the fraudster. Therefore, the similar NAK key generation unit 12 uses the address key “1620813, 2, 3, 405” of the NAK key 120 so that information with the possibility of the same person can be extracted in response to such a fraudulent input. The similar NAK key 140 including the address key “1620813, 2, 3” from which the room number is omitted is generated.

  For example, the similar NAK key generation unit 12 generates a similar NAK key in which the address of the address and / or the room number is changed. The reason is as follows. It is assumed that a fraudster deliberately uses “3-4405 Higashi Gouencho, Shinjuku-ku, Tokyo” and “3-4465 Higashi Gouencho, Shinjuku-ku, Tokyo” as the delivery address. In this case, in the comparison of the normal NAK key 120, each address key is converted into “1620813, 2, 3, 405” and “1620813, 2, 3, 465”, so that it is recognized as a different person. However, even if “465” is written in the room number of the delivery destination of the product etc., the delivery person may assume that the error is “405” and deliver the product etc. to the fraudster. is there. Therefore, the similar NAK key generation unit 12 uses the address key “1620813, 2, 3, 405” of the NAK key so that information with the possibility of the same person can be extracted even for such a fraudster input. The similar NAK key 140 including the address key “1620813, 2, 3, 465” in which the number that is easily recognized as an error is changed is generated. Further, the similar NAK key generation unit 12 may generate the similar NAK key 140 including “1620813, 2, 3, 406” or the like in which the address and / or the room number are slightly changed.

  FIG. 3 shows a configuration example of the history information DB 200. Hereinafter, the history information DB 200 will be described with reference to FIG.

  The history information DB 200 manages past application information as history information. The history information associates the application ID 201, the NAK key 202, and the similar NAK key 203. The history information may further associate the application ID 201 with the telephone number 204 included in the application information 100, the purchase date 205 of the product, the amount 206 of the product, and the fraud factor 207. In the history information, the application ID 201 may be associated with the credit card number included in the application information 100 and the terminal identification number. The telephone number 204, the credit card number, and the terminal identification number may be used in the history information DB 200 to specify the history information of the same person.

  The application ID 201 is an ID for identifying the application information 100.

  The NAK key 202 is a NAK key generated from the application information 100 corresponding to the application ID 201, and the similar NAK key 202 is a similar NAK key generated from the NAK key 202.

  Here, for convenience of description, the NAK key is represented by uppercase letters “A”, “B”,. A plurality of similar NAK keys generated from the NAK key “A” are represented by “a1”, “a2”, “a3”,.

  The fraud coefficient 207 is a value related to the degree of possibility that the applicant related to the history information is a fraudster. For example, when the price of the product related to the history information is not paid, the fraud coefficient 207 of the history information is set high. The fraud coefficient 207 is used to calculate a fraud score 180 of the application information 100 described later.

  In FIG. 3, the history information 210a corresponding to the application ID 201 “1” includes the applicant's NAK key 202 of “A”, the similar NAK key 202 generated from the NAK key “A” of “a1”, “a2”, “A3”, the applicant's telephone number 204 is “03-xxxx-xxxx”, the purchase date 205 of the product etc. is “10/1”, the amount 206 of the product etc. is “1000 yen”, and the fraud factor 207 is “5”. Indicates that

  The extraction unit 13 extracts history information that may be the same delivery destination as the application information 100 from the history information DB 200. That is, the extraction unit 13 extracts history information having the NAK key 202 or the similar NAK key 204 that matches any of the NAK key 120 and the plurality of similar NAK keys 140 generated from the application information 100 from the history information DB 200, An extraction list 160 is generated. Thus, even if the application information 100 is intentionally converted so that the product or the like can be delivered and is not recognized as the same person, the extraction unit 13 may indicate the possibility of the same person. Certain history information can be extracted. That is, the extraction unit 13 extracts a matching character string from the history information DB and outputs the extracted character string itself.

  The fraud score calculation unit 14 calculates the fraud score 180 of the application information 100 based on the extraction list 160.

For example, the fraud score calculation unit 14 calculates the fraud score 180 based on the following conditions.
(Condition 1) The fraud score calculation unit 14 calculates the fraud score 180 of the application information 100 higher as the number of pieces of history information included in the extraction list 160 increases.
(Condition 2) The number of matched NAK keys and similar NAK keys (that is, the matched NAK key 162 in the extraction list 160 in FIG. 4) increases when the fraud score calculating unit 14 extracts history information from the history information DB 200. The fraud score 180 of the application information 100 is calculated high.
(Condition 3) The fraud score calculation unit 14 obtains the fraud score 180 when the NAK key 120 of the application information 100 matches the similar NAK key 203 of the history information DB 200, and the NAK key 120 of the application information 100 indicates the NAK of the history information DB 200. It is calculated higher than when the key 202 is matched. As described above, the scammers are more likely to deliberately write the last name and / or address input to the application information 100 so as not to be considered the same person. On the other hand, even if the general applicant and the last name and / or address input to the application information 100 may fluctuate, the general applicant is unlikely to make the error described above.
(Condition 4) The fraud score calculation unit 14 determines that the fraudulent score of the application information 100 becomes shorter as the time between the purchase date of the product included in the application information 100 and the purchase date of the product of the history information included in the extraction list 160 is shorter. Score 180 is calculated high. This is because a scammer is likely to apply for a large amount of goods in a short period of time while posing as another person.
(Condition 5) The fraud score calculation unit 14 calculates the fraud score 180 higher as the amount of merchandise included in the application information 100 is larger. This is because the larger the amount of merchandise, etc., the greater the loss if the application is from a fraudster.
(Condition 6) The fraud score calculation unit 14 calculates the fraud score 180 of the application information 100 higher as the fraud coefficient of the extracted history information is higher. This is because it is highly likely that the fraudster has not paid for the product etc. many times.

  The fraud score calculation unit 14 may calculate the fraud score 180 using any one of the above conditions 1 to 6, or may calculate the fraud score 180 by combining a plurality of conditions.

  When there is a high possibility that the application information 100 is fraud, the warning unit 15 warns a business operator or the like. If the fraud score 180 of the application information 100 calculated by the fraud score calculation unit 14 is greater than a predetermined threshold, the warning unit 15 warns the business operator that the application information 100 is likely to be fraud. Thereby, the business operator can reduce the possibility of shipping the product etc. in response to the application from the fraudster.

  FIG. 4 is a schematic diagram illustrating an example of a procedure for calculating the fraud score 180. Hereinafter, an example of processing of the extraction unit 13, the fraud score calculation unit 14, and the warning unit 15 will be described with reference to FIG.

  Here, it is assumed that a NAK key 101 “K” is generated from the application information 100 and three similar NAK keys 102 “k1”, “k2”, and “k3” are generated from the NAK key 101 “K”. Then, it is assumed that the NAK key 101 “K” matches the similar NAK key 203 “a1” shown in FIG. 3, and the similar NAK key 102 “k2” matches the similar NAK key 203 “c2” shown in FIG.

  The extraction unit 13 acquires the NAK key 101 “K” generated by the NAK key generation unit 11 from the application information 100 and the similar NAK keys 102 “k1”, “k2”, and “k3” generated by the similar NAK key 12. .

  The extraction unit 13 acquires history information having the NAK key 202 or the similar NAK key 203 that matches any of the NAK key 101 “K” and the similar NAK key 102 “k1”, “k2”, and “k3” from the history information DB 200. Extract. Since “K = a1” and “k2 = c2”, the extraction unit 13 applies the history information 210a of the application ID “1” including the similar NAK key 203 “a1” and the similar NAK key 202 “c2”. The history information 210c with ID “3” is extracted, and an extraction list 160 is generated.

  For example, the fraud score calculation unit 14 based on the above condition 1 calculates the fraud score 180 as “2” because the history information included in the extraction list 160 is two.

  For example, the fraud score calculation unit 14 based on the above conditions 1, 4, 5, and 6 calculates the fraud score 180 from the extraction list 160 as follows.

  For each piece of history information included in the extraction list 160, the fraud score calculating unit 14 determines a period Pi (i is a period between the purchase date 164 of the product or the like of the history information and the purchase date 104 of the product or the like of the application information 100. A positive integer indicating each of the plurality of extracted history information is calculated.

  Then, the fraud score calculation unit 14 calculates the fraud score “F” of the application information 100 based on the following Equation 1.

  F = C × Σi {(1 / Pi) × Ei} (Formula 1)

  Here, “Ei” is the fraud coefficient 166 of the history information included in the extraction list, and “C” is the amount 105 of the product related to the application information 100.

  According to Equation 1, the greater the number of history information items included in the extraction list 160, the higher the fraud score F (corresponding to condition 1).

  Further, the shorter the period (Pi) between the purchase date 104 of the product etc. of the application information 100 and the purchase date 164 of the product etc. of each history information included in the extraction list 160, the higher the fraud score F becomes (conditions) 4).

  In addition, the fraud score F increases as the amount 105 (C) of the product etc. of the application information 100 increases (corresponding to condition 5).

  Further, the fraud score F increases as the fraud coefficient 166 (Ei) of each piece of history information included in the extraction list 160 increases (corresponding to condition 6).

  When the fraud score F calculated by the fraud score calculation unit 14 is larger than a predetermined threshold T (F> T), the warning unit 15 warns the operator.

  In FIG. 4, when the NAK key 101 is normally compared, the NAK key 101 “K” of the application information 100 does not match either of the NAK keys 202 “A” and “C” of the history information DB 200. The history information 210a and 210c corresponding to “1” and “3” cannot be extracted. That is, when the NAK key 101 is normally compared, the applicant related to the application information 100 is recognized as a different person from the applicant related to the history information 210a and 210c registered in the history information DB 200. However, as described above, the similar NAK key 203 considering the trick of the fraudster is also registered in the history information DB 200 and compared with the NAK key 101 and the similar NAK key 102 generated from the application information 100. It is possible to extract history information 210a and 210c with the possibility of

  Then, the information (purchase date 164, fraud coefficient 1667, etc.) included in the history information 210a, 210c extracted as described above and the information (purchase date 104, amount 105, etc.) included in the application information 100 are determined according to predetermined conditions ( By applying to the conditions 1, 4, 5, 6, etc., it is possible to calculate the fraud score 180 indicating whether or not the application information 100 is likely to be an application from a fraudster. Then, by using the fraud score 180 calculated in this way, when there is a high possibility that the application information 100 is an application from a fraudster, it is possible to warn the operator of that fact.

  Note that the information processing apparatus 1 may calculate a high fraud score for the application information 100 when the history information DB 200 includes history information with a different surname despite the same address as the application information 100. . This is because there is a high possibility that a fraudster is performing a capture fraud pretending to be a plurality of different people living at the same address.

  In addition, when the purchase dates 164 of the plurality of history information included in the extraction list 160 are extremely concentrated in the period immediately before the purchase date 104 of the application information 100, the information processing apparatus 1 A high score may be calculated. This is because there is a high possibility that the scammers are carrying out a capture scam in the short term.

  Further, when the purchase dates 164 of the plurality of history information included in the extraction list 160 are extremely concentrated in a certain period, the information processing apparatus 1 may calculate a high fraud score for the application information 100. . This is because it is highly likely that a fraudster has performed a pre-exercise practice of fraud taking in that certain period, and the application information 100 is likely to be an application from that fraudster.

  The information processing apparatus 1 may calculate a high fraud score for the application information 100 when the product or the like applied for by the application information 100 is highly convertible. Alternatively, the information processing apparatus 1 may calculate a high fraud score for the application information 100 when many of the products related to the history information included in the extraction list 160 are highly convertible. This is because there is a high possibility that fraudsters will take in highly convertible products and acquire them as frauds.

  In addition, when there is no part of the content included in the application information 100, the information processing apparatus 1 may calculate a high fraud score for the application information 100. For example, the information processing apparatus 1 stores in advance data on room numbers that actually exist in an apartment house (a condominium, an apartment, etc.), and the application information includes room numbers that do not exist in the stored data , Calculate a high fraud score for the application information. Alternatively, the information processing apparatus 1 stores the data of the room numbers of the apartment houses used in past transactions, and if the application information includes a room number that does not exist in the stored data, the application Calculate high fraud score for information.

  Further, the information processing apparatus 1 may calculate the fraud score under different conditions depending on whether the content included in the application information 100 is an apartment house or an individual house. For example, when the content included in the application information relates to an individual house, the information processing apparatus 1 calculates a high fraud score due to the difference between “x number” of x chome x address x number. This is because, in the case of an individual house, there is a high possibility that the product will arrive even if “x” is incorrect. For example, if the content included in the application information is related to an apartment house, the information processing apparatus 1 may have an x chome x address x if the application information 100 describes the name of the apartment house (for example, an apartment name). The fraud score due to the difference of “No. x” of the issue is calculated high, and if the name of the apartment is not described, the fraud score due to the difference of “No. x” is calculated low. Because, in the case of an apartment house, if the name of the apartment house is described, there is a high possibility that goods etc. will arrive even if “x” is incorrect, and the name of the apartment house is not described and “x” This is because there is a low possibility that the product etc. will arrive if it is wrong.

  Further, the information processing apparatus 1 may generate a similar NAK key 140 related to number disguise as a typo from the NAK key 120. For example, if the NAK key 120 includes “Room 108”, the scammers may pretend to be a typo in the adjacent key, so the information processing apparatus 1 uses the similar NAK including “Room 107”. A similar NAK key 140 including the key 140 and “Room 109” is generated. For example, if the NAK key 120 includes “room 110”, the scammers may pretend to be a mistake in the type order. Therefore, the information processing apparatus 1 uses the similar NAK key 140 including “room 101”. Is generated.

  Further, the information processing apparatus 1 may generate a similar NAK key 140 related to number disguise disguised as a handwritten error from the NAK key 120. For example, when “4” (“9”) is included in the NAK key 120, the scammers may pretend to be a handwritten error of “9” (“4”). Generates a similar NAK key 140 containing “9”. Similarly, similar NAK keys 140 may be generated for “3” and “5”, “1” and “7”, “6” and “0”, and the like.

  Further, the information processing apparatus 1 may generate a similar NAK key 140 related to notation impersonation posing as a conversion error from the NAK key 120. For example, if the NAK key 120 includes “shelf”, the information processing apparatus 1 generates a similar NAK key 140 including “Tanaka” because a fraudster may pretend to be a conversion error. May be. Similarly, when the NAK key 120 includes “Yamada”, the information processing apparatus 1 may generate a similar NAK key 140 including “Yamada”.

  Hereinafter, a second embodiment having a system configuration different from that of the first embodiment will be described. In the following, blocks having the same functions as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

FIG. 5 shows a configuration example of the application determination system.
The application determination system includes an application reception device 2 and an application determination device 3, which can transmit and receive data through the communication network 4. The communication network 4 is an Internet network, for example. The application reception device 2 includes an application information acquisition unit 10, a NAK key generation unit 11, a similar NAK key generation unit 12, and a warning unit 15. The application determination device 3 includes an extraction unit 13, an unauthorized score calculation unit 14, and a history information DB 200.

  The application accepting apparatus 2 receives the application information 100 input from the PC 5 through the communication network 4 and performs a purchase process for a product or the like related to the application information 100. The application reception apparatus 2 is, for example, a WEB server that is operated by a sales company that operates Internet mail order sales. Therefore, a plurality of application reception devices 2 may be included in the application determination system. The functions of the application information acquisition unit 10, the NAK key generation unit 11, the similar NAK key generation unit 12, and the warning unit 15 may be provided as a program module or a hardware module. And the function of the application reception apparatus 2 may be implement | achieved by incorporating the program module or a hardware module in a WEB server.

  The application determination device 3 calculates a fraud score 180 for the NAK key 120 received from the application acceptance device 2 and transmits the fraud score to the application acceptance device 2. The application determination device 3 is operated by, for example, a service provider that provides a service for determining the possibility of fraud of the application information 100. Therefore, the application determination device 3 and the application reception device 2 may have a one-to-N relationship (N is an integer of 1 or more).

Next, processing of the application determination system will be described.
When receiving the application information 100 from the PC 5 of the user who purchases the product etc., the application receiving apparatus 2 generates a NAK key 120 and a similar NAK key 140 corresponding to the application information 100 as shown in the first embodiment. . Then, the application reception device 2 transmits the NAK key 120 and the similar NAK key 140 to the application determination device 3.

  The application determination device 3 that has received the NAK key 120 and the similar NAK key 140 calculates the fraud score 180 for the NAK key 120 as shown in the first embodiment. Then, the application determination device 3 transmits the fraud score 180 to the application reception device 2. At this time, the application determination device 2 registers the NAK key 120 and the similar NAK 140 key in the history information DB 200. This is for later use in other determinations.

  The application receiving apparatus 2 that has received the fraud score 180 warns that the fraud score 180 is greater than a predetermined threshold, and the application information 100 is likely to be a fraud. Thereby, the sales company can reduce the damage caused by the capture fraud.

  Further, the application reception device 2 transmits the NAK key 120 generated from the application information 100, not the application information 100 itself, to the application determination device 3. Therefore, the sales company can use the service that the application information 100 captures and determines the possibility of fraud without providing personal information to a third party.

  The above-described embodiments of the present invention are examples for explaining the present invention, and are not intended to limit the scope of the present invention only to those embodiments. Those skilled in the art can implement the present invention in various other modes without departing from the spirit of the present invention.

DESCRIPTION OF SYMBOLS 1 ... Information processing apparatus 10 ... Application information acquisition part 11 ... NAK key generation part 12 ... Similar NAK key generation part 13 ... Extraction part 14 ... Injustice score calculation part 15 ... Warning part

Claims (7)

History information that associates a first key including a last name key corresponding to a last name and an address key corresponding to an address character with a second key in which a part of the first key is converted or deleted, and the history information is A database to manage,
A first key generating unit that generates an application first character key and an application address key respectively corresponding to a first character and an address character included in the application information, the application first key including the application first character key and the application address key;
A second key generation unit for generating an application second key obtained by converting or deleting a part of the application first key;
An extraction unit for extracting history information having the first key or the second key that matches the application first key or the application second key from the database;
An information processing apparatus.
The information processing apparatus according to claim 1, wherein the second key generation unit generates the application second key obtained by reading the application last character key included in the application first key and converting the key into a pseudonym.
The said 2nd key production | generation part produces | generates the said application 2nd key which converted or deleted the part corresponding to the address or room number of the said application address key contained in the said application 1st key. The information processing apparatus described.
The fraud score calculation part which calculates the fraud score which shows the possibility of fraud of the said application information based on the said log | history information extracted by the said extraction part, It further has a fraud score calculation part. Information processing device.
The information processing apparatus according to claim 4, wherein the fraud score calculation unit calculates the fraud score according to the number of the history information extracted by the extraction unit.
History information that associates a first key including a last name key corresponding to a last name and an address key corresponding to an address character with a second key in which a part of the first key is converted or deleted, and the history information is An information processing method executed by a computer having a database to be managed,
An application first name key and an application address key corresponding to a first name and an address character included in the application information, respectively, and generating an application first key including the application last name key and the application address key;
Generating a second application key obtained by converting or deleting a part of the first application key;
Extracting history information having the first key or the second key that matches the application first key or the application second key from the database;
An information processing method for executing this by a computer.
History information that associates a first key including a last name key corresponding to a last name and an address key corresponding to an address character with a second key in which a part of the first key is converted or deleted, and the history information is An information processing method executed by a computer having a database to be managed,
An application first name key and an application address key corresponding to a first name and an address character included in the application information, respectively, and generating an application first key including the application last name key and the application address key;
Generating a second application key obtained by converting or deleting a part of the first application key;
Extracting history information having the first key or the second key that matches the application first key or the application second key from the database;
A computer program that causes a computer to execute.

Images