Fraudulent receipt detection program, fraudulent receipt detection method, and information processing device

Abstract

To output information of a receipt which may be fraudulent.SOLUTION: A fraudulent receipt detection program causes a computer to execute processing for: acquiring an image of a receipt; extracting the telephone number of an issuer of the receipt from the image; determining whether the extracted telephone number is the telephone number of the predetermined first issuer; and if the extracted telephone number is determined to be the telephone number of the first issuer, performing first confirmation processing to check whether the receipt in the image contains any point which does not match the description rules of receipts issued by the first issuer and outputting the results of the first confirmation processing.SELECTED DRAWING: Figure 14

Description

[Technical field]

[0001] The present invention relates to a fraudulent receipt detection program, a fraudulent receipt detection method, and an information processing device. [Background technology]

[0002] In order to promote sales of a specific product, we may conduct campaigns in which only those who have purchased that product can participate or apply. In such campaigns, a condition for participation or application may be to send (submit) to the campaign organizer an image of the receipt issued at the store when the specific product was purchased.

[0003] If such a campaign were to be implemented, there is a possibility that people will create fraudulent receipts by altering or fabricating receipts and then use those fraudulent receipts to apply for the campaign.

[0004] Recently, techniques for detecting fraudulent receipts have been known, such as those described in Patent Documents 1 and 2. Patent Document 1 describes determining whether receipt information has already been registered in an application information database (whether it is duplicated) to determine whether the receipt has been counterfeited or tampered with, and calculating a total amount for each purchased item from the amount (unit price) of each purchased item and the number of purchases included in the receipt information, and determining whether the calculated total amount matches the total amount in the receipt information to determine whether the receipt has been counterfeited or tampered with. Patent Document 2 also describes determining whether an application has been fraudulently made using a receipt for which no actual purchasing action has been made by comparing the time of actual transaction processing with the time of transaction processing written on the receipt. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2015-079371 A [Patent Document 2] JP 2018-101189 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, even if the techniques described in Patent Documents 1 and 2 are used, there is a risk that fraudulent receipts cannot be detected.

[0007] In one aspect, the present invention has an object to provide a fraudulent receipt detection program, a fraudulent receipt detection method, and an information processing device that are capable of outputting information on possibly fraudulent receipts. [Means for solving the problem]

[0008] In one aspect, the fraudulent receipt detection program is a program that causes a computer to execute a process of acquiring an image of a receipt, extracting the telephone number of the issuer of the receipt from the image, determining whether the extracted telephone number is the telephone number of a predetermined first issuer, and if the result of the determination is that the extracted telephone number is the telephone number of the first issuer, executing a first confirmation process to check whether there are any points in the receipt included in the image that do not match the receipt writing rules of the first issuer, and outputting the results of the first confirmation process. Effect of the Invention

[0009] It is possible to output information about potentially fraudulent receipts. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram illustrating a schematic configuration of a campaign management system according to an embodiment. [Diagram 2] FIG. 2(a) is a diagram showing an example of the hardware configuration of a server, and FIG. 2(b) is a diagram showing an example of the hardware configuration of a user terminal. [Diagram 3] FIG. 3 is a diagram illustrating an example of a hardware configuration of the manager terminal. [Figure 4] FIG. 4 is a functional block diagram of the server. [Diagram 5] FIG. 5(a) is a diagram showing an example of a target store DB, and FIG. 5(b) is a diagram showing an example of a rule table. [Figure 6] FIG. 6 is a flowchart showing the fraudulent receipt check process executed by the server. [Figure 7] FIG. 7(a) is a diagram showing an example of a receipt image, and FIG. 7(b) is a diagram for explaining the process of extracting a phone number from the receipt image. [Figure 8] FIG. 8 is a flowchart showing detailed processing of step S18 in FIG. [Figure 9] FIG. 9 is a diagram for explaining the process of FIG. [Figure 10] FIG. 10 is a flowchart showing detailed processing of step S62 in FIG. [Figure 11] FIG. 11 is a flowchart showing detailed processing of step S22 in FIG. [Figure 12] FIG. 12 is a flowchart showing detailed processing of step S162 in FIG. [Figure 13] FIG. 13 is a diagram illustrating an example of the return value table. [Figure 14] FIG. 14 is a diagram (part 1) for explaining fraudulent receipts. [Figure 15] FIG. 15 is a diagram (part 2) for explaining fraudulent receipts. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of the campaign management system will be described in detail with reference to FIGS.

[0012] Fig. 1 is a diagram showing a schematic configuration of a campaign operation system 100 according to an embodiment. As shown in Fig. 1, the campaign operation system 100 according to the present embodiment includes a server 10 as an information processing device, a user terminal 60, and an administrator terminal 70. The server 10, the user terminal 60, and the administrator terminal 70 are connected to a network 80 such as the Internet, enabling information to be exchanged between the devices.

[0013] The server 10 accepts applications for the campaign from the user terminal 60. The campaign in this embodiment is a campaign to which people who purchase a specific product can apply. A person who purchases a specific product (campaign applicant) takes a picture of the receipt issued by the retailer when purchasing the specific product using his / her own terminal (user terminal 60), and transmits the photographed image (receipt image) along with his / her personal information (name, address, etc.) to the server 10.

[0014] The server 10 acquires the receipt image and the applicant's personal information, etc., sent from the user terminal 60, and manages them as applicant information. The server 10 also uses the applicant's information to carry out processing related to the campaign, such as holding a lottery. The server 10 also checks whether the receipt photographed in the receipt image is likely to be a fraudulent receipt, and if there is a possibility that it is a fraudulent receipt, outputs to the administrator terminal 70 a message indicating that the receipt image needs to be visually confirmed. Furthermore, if the server 10 receives input from the administrator terminal 70 indicating that the receipt is fraudulent, it executes processing such as excluding the applicant who sent the fraudulent receipt from the lottery.

[0015] FIG. 2(a) shows an example of the hardware configuration of the server 10. As shown in FIG. 2(a), the server 10 includes a CPU (Central Processing Unit) 90, a ROM (Read Only Memory) 92, a RAM (Random Access Memory) 94, a storage (e.g., an SSD (Solid State Drive) or an HDD (Hard Disk Drive)) 96, a network interface 97, and a portable storage medium drive 99. These components of the server 10 are connected to a bus 98. In the server 10, the CPU 90 executes a program (including a fraudulent receipt detection program) stored in the ROM 92 or the storage 96, or a program read by the portable storage medium drive 99 from the portable storage medium 91, thereby realizing the functions of the components shown in FIG. 4. The functions of the components shown in FIG. 4 may be realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). Details of the components shown in FIG. 4 will be described later.

[0016] The user terminal 60 is an information processing terminal such as a smartphone, and is a terminal owned and used by a user (applicant) who applies for the campaign. The user terminal 60 can access the campaign site, and the applicant can apply for the campaign by inputting and transmitting a receipt image and personal information of the applicant on the site. FIG. 2(b) shows an example of the hardware configuration of the user terminal 60. As shown in FIG. 2(b), the user terminal 60 includes a CPU 190, a ROM 192, a RAM 194, a storage 196, a network interface 197, a display unit 193, an input unit 195, and a camera 189. These components of the user terminal 60 are connected to a bus 198. The display unit 193 includes a liquid crystal display, and the input unit 195 includes a touch panel. The camera 189 captures an object and generates an image in response to a user's operation. In this embodiment, the camera 189 generates an image (receipt image) of a captured receipt. The user terminal 60 does not have to include the camera 189. In this case, the user terminal 60 only needs to be able to obtain receipt images captured by other devices via communication or the like.

[0017] The administrator terminal 70 is a terminal such as a PC (Personal Computer) used by the administrator of the campaign. If a receipt that may be fraudulent is found on the server 10, the administrator terminal 70 is notified that the administrator should visually check the receipt. Upon receiving this notification from the server 10, the administrator terminal 70 notifies the administrator via the display unit 293 (see FIG. 3). When the notification is received on the display unit 293, the administrator visually checks the receipt using the administrator terminal 70.

[0018] Fig. 3 shows an example of a hardware configuration of the manager terminal 70. As shown in Fig. 3, the manager terminal 70 includes a CPU 290, a ROM 292, a RAM 294, a storage 296, a network interface 297, a display unit 293, and an input unit 295. These components of the manager terminal 70 are connected to a bus 298.

[0019] (About Server 10 functions) Next, the functions of the server 10 will be described with reference to Fig. 4. Fig. 4 is a functional block diagram of the server 10.

[0020] 4, in the server 10, the CPU 90 executes a program to realize functions as an application information acquisition unit 20, a receipt image acquisition unit 22, an information extraction unit 24, a store identification unit 26 as a determination unit, a receipt analysis unit 28, and a notification unit 29. Note that Fig. 4 also illustrates an application information DB 30, a target store DB 32, a rule table 34 as a first memory unit and a second memory unit, and a return value table 36, which are stored in the RAM 94, storage 96, etc.

[0021] The application information acquisition unit 20 acquires receipt images, personal information of applicants, etc. transmitted from the user terminal 60, and stores them in the application information DB 30. The application information DB 30 stores receipt images and personal information of applicants, etc., in association with each other.

[0022] The receipt image acquisition unit 22 acquires receipt images to be checked for fraud from the receipt images stored in the application information DB 30. Receipt images to be checked for fraud are receipt images that have not yet been checked for fraud.

[0023] The information extraction unit 24 analyzes the receipt image acquired by the receipt image acquisition unit 22 using OCR (Optical Character Recognition / Reader) or the like, and extracts various receipt information from the receipt image. The various information includes "phone number (of the issuer of the receipt)", "total amount", "date and time", "payment information (payment amount (type), change)", "product information (product name, amount, unit price and quantity, JAN (Japanese Article Number) code, discount)", "tax information (total amount and tax amount for tax rates of 8% and 10%)", "receipt additional information (point card information, business number)", etc.

[0024] The store identification unit 26 obtains the telephone number information of the receipt issuer extracted by the information extraction unit 24, and references the target store DB 32 to identify whether the receipt issuer is a predetermined "target store" (first issuer). The target store DB 32 has a data structure as shown in FIG. 5(a). Specifically, the target store DB 32 stores the "distribution name" and "store name" of predetermined target stores, as well as the "telephone number" of each store. Note that a target store refers to a store for which rules regarding receipt entry are managed in a rule table 34, which will be described later.

[0025] Returning to FIG. 4, the receipt analysis unit 28 uses the information extracted by the information extraction unit 24 to analyze whether the receipt is likely to be a fraudulent receipt based on the rule table 34. If there is a possibility that the receipt is fraudulent, the receipt analysis unit 28 outputs a return value to the notification unit 29. Note that, although details will be described later, the analysis method differs depending on whether the receipt was issued by a target store or not. The rule table 34 is a table as shown in FIG. 5(b). Specifically, as shown in FIG. 5(b), the rule table 34 stores, for each distributor name, information on whether a subtotal is written on the receipt issued by each distributor (o / x), whether a JAN code is written on the receipt (o / x), where the date position is on the receipt, what the fraudulent keyword is, and how the campaign target product is to be displayed. The JAN code is product identification information, and is a code consisting of 13 or 8 digits that is assigned to each product.

[0026] The notification unit 29 notifies the manager terminal 70 based on the return value output from the receipt analysis unit 28. When making the notification, the notification unit 29 refers to the return value table 36. Details of the return value table 36 will be described later.

[0027] (Regarding Server 10 processing) Next, the processing of the server 10 will be described in detail with reference to the flowchart and other drawings as appropriate.

[0028] Fig. 6 is a flowchart showing the fraudulent receipt check process executed by the server 10. Before the process in Fig. 6 can begin, it is assumed that the application information acquisition unit 20 in Fig. 4 has acquired receipt images and personal information of applicants who have applied to the campaign, and stored them in the application information DB 30. The process in Fig. 6 checks whether each receipt image that is subject to fraud check (receipt images that have not yet been checked for fraud) stored in the application information DB 30 is possibly a fraudulent receipt.

[0029] When the process of FIG. 6 starts, first, in step S10, the receipt image acquisition unit 22 acquires from the application information DB 30 one receipt image to be checked for fraud.

[0030] Next, in step S12, the information extraction unit 24 analyzes the receipt image acquired by the receipt image acquisition unit 22 using OCR or the like, and extracts information on the telephone number of the issuer of the receipt from the receipt image. For example, assume that a receipt image such as that shown in FIG. 7(a) is the receipt image to be checked for fraud. The information extraction unit 24 acquires telephone number information, shown in a dashed line frame in FIG. 7(b), from the receipt image in FIG. 7(a). At this time, the information extraction unit 24 extracts, for example, a string of 10 (or 11) digits, excluding symbols such as "(", ")", and "-" that are present in the receipt image, and uses this as information on the telephone number of the issuer of the receipt.

[0031] Next, in step S14, the store identification unit 26 identifies a store from the phone number extracted by the information extraction unit 24. More specifically, the store identification unit 26 references the target store DB 32 to identify whether the receipt was issued by one of the target stores or a store other than the target store. The identification result by the store identification unit 26 is sent to the information extraction unit 24.

[0032] Next, in step S16, the information extraction unit 24 judges whether the store is a target store. If the judgment in step S16 is positive, the process proceeds to step S18, and if not, the process proceeds to step S22.

[0033] When the process proceeds to step S18, a target store receipt analysis process is executed. In step S18, the process follows the flowchart of FIG.

[0034] When the process of Figure 8 starts, first in step S50, the information extraction unit 24 extracts the total amount from the receipt image. For example, the information extraction unit 24 extracts the characters "TOTAL" from the receipt image using OCR, and extracts the string of numbers to the right of the characters "TOTAL" as the total amount. In the case of the receipt image of Figure 7(a), "2,028" is extracted as shown in Figure 9.

[0035] Next, in step S52, the information extraction unit 24 extracts the date and time (purchase date and time) from the receipt image. For example, the information extraction unit 24 uses OCR from the receipt image to extract a character string written in a predetermined date and time format. In the case of the receipt image of FIG. 7(a), "2023 June 16th (Sun) 11:11" is extracted as shown in FIG. 9.

[0036] Next, in step S54, the information extraction unit 24 extracts the payment amount and the change amount from the receipt image. In the case of the receipt image of Fig. 7(a), the information extraction unit 24 extracts "2,028" written to the right of the character string "Payment" as the payment amount, as shown in Fig. 9. Note that if a "Subtotal" is written on the receipt, the information extraction unit 24 also extracts the subtotal amount information.

[0037] Next, in step S56, the information extraction unit 24 extracts information about the purchased items from the receipt image. In the case of the receipt image of Fig. 7(a), the information extraction unit 24 extracts information about the product names, such as radish and onion, and the price of each item, as shown in Fig. 9. If a 13-digit or 8-digit numeric string (JAN code) is written near the product name in the receipt image, the JAN code is also extracted.

[0038] Next, in step S58, the information extraction unit 24 extracts tax information from the receipt image. In the case of the receipt image of Fig. 7(a), the information extraction unit 24 extracts information such as "Product prices include consumption tax, etc." as tax information, as shown in Fig. 9. Note that depending on the receipt, the total amount and tax amount for tax rates of 8% and 10% may be written, and if such information is present, it is extracted as tax information.

[0039] Next, in step S60, the information extraction unit 24 extracts receipt additional information from the receipt image. In the case of the receipt image of Fig. 7(a), the information extraction unit 24 extracts information related to the electronic money number, balance, and points as receipt additional information, for example, as shown in Fig. 9.

[0040] Next, in step S62, the receipt analysis unit 28 performs a fraudulent receipt check process. In step S62, the process follows the flowchart in FIG.

[0041] When the processing in Figure 10 begins, first in step S100, the receipt analysis unit 28 executes a process to check whether or not a subtotal is included. Specifically, the receipt analysis unit 28 references the rule table 34 (Figure 5(b)) to determine whether or not the issuer of the receipt has a rule that a subtotal should be included on the receipt. The receipt analysis unit 28 then checks whether or not a subtotal has been extracted from the receipt image, and whether or not the result of this check (whether or not a subtotal has been extracted) matches the rule regarding the entry of a subtotal identified from the rule table 34.

[0042] Next, in step S102, the receipt analysis unit 28 determines whether fraud has been detected. If the processing in step S100 confirms that the subtotal on the receipt does not match the rules, the determination in step S102 is positive. If the determination in step S102 is positive (fraud has been detected), the process proceeds to step S104.

[0043] When the process proceeds to step S104, the receipt analyzer 28 outputs a return value (1) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0044] On the other hand, if the determination in step S102 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S106. When proceeding to step S106, the receipt analysis unit 28 executes a process to check whether or not a JAN code is present. Specifically, the receipt analysis unit 28 references the rule table 34 to determine whether or not a rule has been established that requires the issuer of the receipt to write the JAN code on the receipt. The receipt analysis unit 28 then checks whether or not a JAN code has been extracted from the receipt image, and checks whether or not the confirmation result (whether or not a JAN code has been extracted) matches the rule determined from the rule table 34.

[0045] Next, in step S108, the receipt analysis unit 28 determines whether fraud has been detected. If the process in step S106 confirms that the JAN code on the receipt does not match the issuer's rules regarding the JAN code, then the determination in step S108 is positive. If the determination in step S108 is positive (fraud has been detected), the process proceeds to step S110.

[0046] When the process proceeds to step S110, the receipt analyzer 28 outputs a return value (2) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0047] On the other hand, if the determination in step S108 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S112. When proceeding to step S112, the receipt analysis unit 28 executes a date position check process. Specifically, the receipt analysis unit 28 references the rule table 34 to identify the rule regarding where the date should be written on the receipt by the issuer of the receipt. The receipt analysis unit 28 then checks the position where the date was extracted from the receipt image, and checks whether the check result (date position) matches the rule identified from the rule table 34.

[0048] Next, in step S114, the receipt analysis unit 28 determines whether fraud has been detected. If the process in step S112 confirms that the date position on the receipt does not match the date position rules of the issuer, the determination in step S114 is positive. If the determination in step S114 is positive (fraud has been detected), the process proceeds to step S116.

[0049] In step S116, the receipt analyzer 28 outputs a return value (3) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0050] On the other hand, if the determination in step S114 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S118. When proceeding to step S118, the receipt analysis unit 28 executes an unauthorized keyword check process. Specifically, the receipt analysis unit 28 references the rule table 34 to identify character strings that the issuer of the receipt has specified as unauthorized keywords. The receipt analysis unit 28 then checks whether any of the character strings extracted from the receipt image in steps S50 to S60 match the unauthorized keyword. For example, assume that the store that issued the receipt never writes the keyword "Extra 8%" on a receipt. In this case, since the unauthorized keyword "Extra 8%" is stored in the rule table 34, if the keyword "Extra 8%" was extracted in step S58, the receipt analysis unit 28 determines that any of the character strings extracted from the receipt image match the unauthorized keyword.

[0051] Next, in step S120, the receipt analysis unit 28 determines whether fraud has been detected. If the processing in step S118 confirms that the character string extracted from the receipt contains a fraudulent keyword, the determination in step S120 is positive. If the determination in step S120 is positive (fraud has been detected), the process proceeds to step S122.

[0052] In step S122, the receipt analyzer 28 outputs a return value (4) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0053] On the other hand, if the determination in step S120 is negative (if no fraud is detected), receipt analysis unit 28 proceeds to step S124. In step S124, receipt analysis unit 28 executes a training receipt check process. A training receipt is a receipt that is output when a store clerk operates a cash register in a mode for practicing cash register operation (training mode). For this reason, participation in or application for a campaign using a training receipt is fraudulent. In this step S124, it is confirmed whether the receipt included in the receipt image is a training receipt. Specifically, receipt analysis unit 28 determines that a receipt is a training receipt if the receipt image contains the words "training" or "practice."

[0054] Next, in step S126, the receipt analysis unit 28 determines whether fraud has been detected. If the receipt is determined to be a training receipt in the processing of step S124, the receipt analysis unit 28 determines that fraud has been detected. If the determination in step S126 is positive (fraud has been detected), the process proceeds to step S128.

[0055] In step S128, the receipt analyzer 28 outputs a return value (5) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0056] On the other hand, if the determination in step S126 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S130. When proceeding to step S130, the receipt analysis unit 28 executes a reissue receipt check process. When a reissue receipt is issued, there is a possibility that the purchaser has the first issued receipt and the reissued receipt in hand. Therefore, participation in or application to a campaign using a reissued receipt is treated as fraud. In this step S130, the receipt analysis unit 28 determines that the receipt is a reissued receipt if the receipt image has the words "reissue" or "reprint" written on it. Note that, depending on the store, the reissued receipt may not have the character string "reissue" or "reprint" written on it. In such a case, it is sufficient to confirm whether or not the receipt is a reissued receipt using another method. For example, if the positional relationship between the date and the phone number in the case of a reissued receipt is different from that in the case of a normal receipt, it is sufficient to determine whether or not the receipt is a reissued receipt based on the positional relationship between the date and the phone number. Also, for example, in the case of a reissued receipt, if the positional relationship between the total amount and the phone number is different from that of a normal receipt, it can be determined whether or not it is a reissued receipt based on the positional relationship between the total amount and the phone number. Also, for example, in the case of a reissued receipt, if the positional relationship between the total amount and the date is different from that of a normal receipt, it can be determined whether or not it is a reissued receipt based on the positional relationship between the total amount and the date. Note that the rules used to determine whether or not a receipt is a reissued receipt can be stored in the rule table 34 for each store. In this case, the receipt analysis unit 28 can refer to the rule table 34 when executing step S126.

[0057] Next, in step S132, the receipt analysis unit 28 determines whether fraud has been detected. If the receipt is determined to be a reissued receipt in the processing of step S130, the receipt analysis unit 28 determines that fraud has been detected. If the determination in step S132 is positive (fraud has been detected), the process proceeds to step S134.

[0058] In step S134, the receipt analyzer 28 outputs a return value (6) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0059] On the other hand, if the determination in step S132 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S136. In step S136, the receipt analysis unit 28 executes a maintenance receipt check process. A maintenance receipt is a receipt that is output at the time of maintenance. Participation in or application for a campaign using a maintenance receipt is considered to be fraudulent. In this step S136, the receipt analysis unit 28 determines that a receipt is a maintenance receipt if the receipt image contains the word "maintenance," for example.

[0060] Next, in step S138, the receipt analysis unit 28 determines whether fraud has been detected. If the receipt is determined to be a maintenance receipt in the processing of step S136, the receipt analysis unit 28 determines that fraud has been detected. If the determination in step S138 is positive (if fraud has been detected), the process proceeds to step S140.

[0061] When the process proceeds to step S140, the receipt analyzer 28 outputs a return value (7) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0062] On the other hand, if the determination in step S138 is negative (if no fraud is detected), the receipt analysis unit 28 proceeds to step S142. In step S142, the receipt analysis unit 28 executes an electronic journal receipt check process. An electronic journal receipt is a receipt that is issued during settlement processing and contains sales information and sales status. Participation or application to a campaign using an electronic journal receipt will be treated as fraud. In this step S142, the receipt analysis unit 28 determines that it is an electronic journal receipt if the receipt image contains words such as "electronic journal" or "child journal."

[0063] Next, in step S144, the receipt analysis unit 28 determines whether fraud has been detected. If the receipt analysis unit 28 determines in the processing of step S142 that the receipt is an electronic journal receipt, it determines that fraud has been detected. If the determination in step S144 is positive (if fraud has been detected), the process proceeds to step S146.

[0064] In step S146, the receipt analyzer 28 outputs a return value (8) to the notification unit 29, ends the process of FIG. 10, and proceeds to step S20 in FIG.

[0065] On the other hand, if the determination in step S144 is negative (if no fraud is detected), the process in FIG. 10 ends and the process proceeds to step S20 in FIG.

[0066] Returning to FIG. 6, when the process proceeds to step S20, the receipt analysis unit 28 executes a product description determination process. The receipt analysis unit 28 refers to the "target product description" column in the rule table 34 and checks whether the description of the target product of the campaign on the receipt matches the product description defined for each store. The receipt analysis unit 28 identifies the description of the target product from the information of the purchased product extracted in step S56 in FIG. 8. For example, if the official name of the target product is "The Minato Free Zero", the receipt analysis unit 28 identifies the description of the purchased product corresponding to the official name ("Minato Free Zero" in the case of FIG. 9) and checks whether the description matches the target product description for the store stored in the rule table 34. If the result of this check shows that the product descriptions do not match, the receipt analysis unit 28 outputs a return value (9) to the notification unit 29 and proceeds to step S24. On the other hand, if the product descriptions match, the receipt analysis unit 28 does not output the information to the notification unit 29 and proceeds to step S24.

[0067] Meanwhile, if the determination in step S16 in FIG. 6 is negative (if the store is not a target store), the process proceeds to step S22, where a non-target store receipt analysis process is executed. In this step S22, the process is executed according to the flowchart in FIG. 11. Note that the processes in steps S150, S152, S154, S156, S158, and S160 in FIG. 11 are the same as the processes in steps S50, S52, S54, S56, S58, and S60 in FIG. 8, so detailed explanations will be omitted. After the process in step S160 in FIG. 11 is executed, the process proceeds to step S162. In step S162, the process is executed according to the flowchart in FIG. 12. Note that the processes in steps S224 to S246 in FIG. 12 are the same as steps S124 to S146 in FIG. 10. In other words, if the store that issued the receipt is not the target store, steps S100 to S122 in Fig. 10 will not be executed. In the reissue receipt check process in step S230, the receipt analysis unit 28 will determine that the receipt is a reissue receipt if a specific character string (such as "reissue" or "reprint") is present in the receipt image.

[0068] The information extracted in steps S150 to S160 in Fig. 11 is not used in the fraudulent receipt check in step S162, but can be used in step S24 in Fig. 6, which will be described later. Some of the processes in steps S150 to S160 may be omitted.

[0069] After the process of step S22 in FIG. 6 has been performed as described above, the process proceeds to step S24.

[0070] After steps S18 and S20, or step S22, the process proceeds to step S24, where the receipt analysis unit 28 performs other check processes. These other check processes include check processes for known fraudulent receipts (such as the check processes described in Patent Documents 1 and 2). For example, the receipt analysis unit 28 checks whether the receipt image is the same as a previous receipt image (whether it is a duplicate receipt) using the phone number, total amount, and date and time. If the receipt analysis unit 28 determines that the receipt is fraudulent as a result of the check, it outputs a return value (10) to the notification unit 29.

[0071] Next, when the process proceeds to step S26, the notification unit 29 executes a process to notify the manager terminal 70 of the check results. Specifically, if a return value is output from the receipt analysis unit 28 in steps S18, S20, S22, or S24, the notification unit 29 notifies the manager terminal 70 that the receipt may be fraudulent and that the receipt image should be visually confirmed. At this time, the notification unit 29 references the return value table 36 and also displays information about the fraudulent portion corresponding to the return value output from the receipt analysis unit 28 (the portion that determined the receipt to be fraudulent). The return value table 36 is a table as shown in FIG. 13, in which return values ​​and fraudulent portions are stored in association with each other.

[0072] Next, in step S28, the receipt image acquisition unit 22 determines whether checking of all receipt images to be checked is complete. In other words, the receipt image acquisition unit 22 determines whether all receipt images stored in the application information DB 30 have been acquired. If the determination in step S28 is negative, the process returns to step S10, and the above-mentioned process is repeated. On the other hand, if the determination in step S28 is positive, the entire process in FIG. 6 ends.

[0073] In this embodiment, steps S100, S106, S112, and S118 are executed if the store that issued the receipt is the target store. Therefore, for example, as shown in FIGS. 14 and 15, even if a receipt image is created by combining (joining) the header and footer of a receipt from store A (receipt A) with part of a receipt from store B (receipt B), it is possible to detect the possibility of a fraudulent receipt. In the example of FIG. 14, the telephone number of the receipt subject to fraud check identifies it as a receipt from store A. Receipts from store A do not normally include a subtotal. On the other hand, the receipt subject to fraud check in FIG. 14 does include a subtotal. In this case, the manager terminal 70 is notified that the receipt may be fraudulent (S102: Yes, S104). Furthermore, even if the receipt being checked for fraud in FIG. 14 did not have a subtotal listed, the date position on the receipt being checked for fraud is different from the date position on Receipt A, so the administrator terminal 70 would be notified that the receipt may be fraudulent (S114: Yes, S116).

[0074] Also, in the example of FIG. 15, the receipt being checked for fraud is identified as being from store A based on the telephone number on the receipt. In this case as well, the receipt being checked for fraud in FIG. 15 lists a subtotal, so the manager's terminal 70 is notified that the receipt may be fraudulent (S102: Yes, S104). Even if the receipt being checked for fraud did not list a subtotal, the receipt being checked for fraud in FIG. 15 lists a JAN code, which is different from receipt A (store A does not list a JAN code), so the manager's terminal 70 is notified that the receipt may be fraudulent (S108: Yes, S110). Even if the receipt being checked for fraud in FIG. 14 or 15 does not contain the above fraud, if part of the receipt (part of receipt B) lists a fraudulent keyword corresponding to the issuer of receipt A (a keyword not used by the issuer of receipt A), a notification is issued that the receipt may be fraudulent (S120: Yes, S122).

[0075] Also, in the cases of Figures 14 and 15, the product description "Minato Free" on the receipt image being checked for fraud is different from "Minato Free Zero", so as a result of the product description determination process in step S20 of Figure 6, the administrator terminal 70 will be notified that the receipt may be fraudulent.

[0076] The administrator visually checks receipt images that may be fraudulent using the administrator terminal 70, and if it is determined to be fraudulent, inputs a message to the administrator terminal 70 stating that the person will be excluded from participating or applying for the campaign. The administrator may block people who have used fraudulent receipts, preventing them from participating or applying again.

[0077] In this embodiment, the receipt analysis unit 28 and notification unit 29 function as a processing unit that verifies whether a receipt that is the subject of a fraud check is a fraudulent receipt and notifies the manager terminal 70 of the confirmation result.

[0078] As described above in detail, in this embodiment, the receipt image acquisition unit 22 of the server 10 acquires a receipt image from the application information DB 30 (S10 in FIG. 6), and the information extraction unit 24 extracts the telephone number of the issuer of the receipt from the receipt image (S12 in FIG. 6). If the extracted telephone number is the telephone number of the target store (S16 in FIG. 6: Yes), the receipt analysis unit 28 checks whether there is anything in the receipt image that does not match the receipt entry rules of the target store (rule table 34) (S18 in FIG. 6, S62 in FIG. 8, S100 to S122 in FIG. 10). If the receipt analysis unit 28 determines that the receipt is possibly fraudulent, the notification unit 29 notifies the manager terminal 70 (S26 in FIG. 6). As a result, even if an entry into a campaign is made with a fraudulent receipt, such as one that combines the header and footer of one receipt with parts of another receipt other than the header and footer, the fraudulent receipt can be accurately detected and notified to the administrator terminal 70. Also, even if an entry is made by an applicant on their own, it can be accurately detected as a fraudulent receipt.

[0079] Furthermore, in this embodiment, the rule table 34 stores receipt subtotal entry rules, product identification information entry rules, date entry rules, and fraudulent keywords in association with each target store, and the receipt analysis unit 28 refers to the rule table 34 to determine whether a receipt is likely to be fraudulent (S100-S122 in FIG. 10). This makes it possible to accurately detect fraudulent receipts using the receipt entry rules that are uniquely set for each target store.

[0080] Furthermore, in this embodiment, the rule table 34 stores the description of the target product in association with each target store, and the receipt analysis unit 28 references the "target product description" column in the rule table 34 to determine whether the receipt is likely to be fraudulent (S20 in FIG. 6). This makes it possible to accurately detect fraudulent receipts using the product descriptions that are uniquely set for each target store.

[0081] Furthermore, in this embodiment, if the receipt analysis unit 28 determines that a receipt may be fraudulent, the notification unit 29 notifies the manager that the receipt image should be visually checked. This allows both the receipt analysis unit 28 and the manager to check the receipt image, making it possible to accurately identify fraudulent receipts.

[0082] Furthermore, in this embodiment, the receipt is checked to see if it is a receipt that has been legitimately issued by the issuer (S124 to S146 in FIG. 10, S224 to S246 in FIG. 12). This makes it possible to prevent applications for or participation in campaigns using receipts that have not been legitimately issued.

[0083] In the above embodiment, in Figs. 10 and 12, when fraud is detected once, no further fraud checks are performed and the process in Figs. 10 and 12 is terminated. However, this is not limited to the above. The receipt analysis unit 28 may perform all of the fraud checks in Figs. 10 and 12 and output all return values ​​corresponding to the detected fraud to the notification unit 29. The order of the fraud checks in Figs. 10 and 12 is an example, and the fraud checks may be performed in other orders. Also, some of the fraud checks in Figs. 10 and 12 may be omitted. Also, if other rules are stored in the rule table 34 in Fig. 5(b), the fraud checks based on those rules may be performed in the process in Fig. 10.

[0084] In the above embodiment, the fraud checks in steps S18, S20, S22, and S24 in FIG. 6 are performed. However, the present invention is not limited to this, and other fraud checks do not necessarily have to be performed as long as at least step S18 is performed.

[0085] In the above embodiment, the product description determination process in step S20 determines whether the receipt description of the campaign target product matches the description used by the store, but this is not limited to this. That is, in step S20, it may be determined whether the receipt description of a product other than the campaign target product matches the description used by the store.

[0086] The above processing functions can be realized by a computer. In that case, a program is provided that describes the processing contents of the functions that the processing device should have. The above processing functions are realized on the computer by executing the program on the computer. The program that describes the processing contents can be recorded on a computer-readable recording medium (excluding carrier waves).

[0087] When a program is distributed, it is sold in the form of a portable recording medium on which the program is recorded, such as a DVD (Digital Versatile Disc), a CD-ROM (Compact Disc Read Only Memory), etc. The program can also be stored in a storage device of a server computer and transferred from the server computer to other computers via a network.

[0088] A computer that executes a program stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. The computer then reads the program from its own storage device and executes processing in accordance with the program. The computer can also read the program directly from a portable recording medium and execute processing in accordance with that program. The computer can also execute processing in accordance with the received program each time a program is transferred from the server computer.

[0089] The above-described embodiment is a preferred example of the present invention, but the present invention is not limited to this embodiment and can be modified in various ways without departing from the scope of the present invention. [Explanation of symbols]

[0090] 10 Server (information processing device) 20 Application Information Acquisition Department 22 Receipt image acquisition unit 24 Information extraction part 26 Store Identification Department (Judgment Department) 28 Receipt analysis unit (part of the processing unit) 29 Notification Department (part of the Processing Department) 30 Application Information DB 32 Target store DB 34 Rule table (first storage unit, second storage unit) 36 Return Value Table 60 User terminals 70 Administrator terminal 100 Campaign Operation System

Claims

1. Get an image of the receipt extracting the telephone number of the issuer of the receipt from the image; determining whether the extracted telephone number is a telephone number of a predetermined first issuer; If it is determined that the extracted telephone number is the telephone number of the first issuer, a first confirmation process is executed to confirm whether the receipt included in the image has any points that do not match the writing rules for receipts of the first issuer, and the result of the first confirmation process is output. A fraudulent receipt detection program that causes a computer to execute a process.

2. At least one of a rule for writing a subtotal on a receipt, a rule for writing product identification information, a rule for writing a date, and an illegal keyword is stored in a first storage unit in association with the first issuer, 2. The fraudulent receipt detection program according to claim 1, wherein in the first confirmation process, the receipt included in the image is confirmed by referring to the first storage unit.

3. information on the description of each commodity on a receipt issued by the first issuer is stored in a second storage unit in association with the first issuer; 2. The fraudulent receipt detection program according to claim 1, wherein in the first confirmation process, the receipt included in the image is confirmed by referring to the second storage unit.

4. The fraudulent receipt detection program according to any one of claims 1 to 3, wherein if the result of the first confirmation process indicates that the receipt included in the image does not match the receipt writing rules of the first source, a message is output indicating that the image should be visually confirmed.

5. The fraudulent receipt detection program according to claim 1, further comprising: causing the computer to execute a second confirmation process to confirm whether the receipt included in the image is a receipt legitimately issued by the issuer; and outputting the results of the second confirmation process.

6. 6. The fraudulent receipt detection program according to claim 5, further comprising an output indicating that the image should be visually confirmed if the second confirmation process determines that the receipt included in the image is not a properly issued receipt.

7. 2. The fraudulent receipt detection program according to claim 1, wherein the image of the receipt is an image sent to a campaign organizer by a person who applies for a campaign offering a special benefit for purchasing a specific product.

8. Get an image of the receipt extracting the telephone number of the issuer of the receipt from the image; determining whether the extracted telephone number is a telephone number of a predetermined first issuer; If it is determined that the extracted telephone number is the telephone number of the first issuer, a first confirmation process is executed to confirm whether the receipt included in the image has any points that do not match the writing rules for receipts of the first issuer, and the result of the first confirmation process is output. A fraudulent receipt detection method, characterized in that the processing is executed by a computer.

9. A receipt image acquisition unit that acquires an image of a receipt; an information extraction unit that extracts the telephone number of the issuer of the receipt from the image; a determination unit that determines whether the extracted telephone number is a telephone number of a predetermined first issuer; a processing unit that, if it is determined that the extracted phone number is the phone number of the first issuer, executes a first confirmation process to check whether the receipt included in the image has any points that do not match the writing rules for receipts of the first issuer, and outputs the results of the first confirmation process; An information processing device comprising:

Images