Journal entry extraction device, journal entry extraction method, and program

Abstract

To provide a journal entry extraction device that easily extracts journal entries that are highly likely to be fraudulent. [Solution] According to an embodiment, a device for extracting journal entries that are likely to be fraudulent includes an extraction unit that extracts, from journal entries entered by an inputter and approved by an approver, journal entries entered by the same person as the inputter and approver, as journal entries that are likely to be fraudulent.

Description

[Technical field]

[0001] The present invention relates to a journal entry extraction device, a journal entry extraction method, and a program for extracting journal entries that are likely to be fraudulent. [Background technology]

[0002] Traditionally, in the field of accounting, accounting entries that do not incorporate internal checks and separation of duties have been considered to pose a high risk of fraud.

[0003] Specifically, this type of journal entry is a so-called "one-person job" journal entry (self-approved journal entry) in which the person who inputs the information and the person who approves it are the same person.

[0004] In addition, even if it is not a one-person job, transactions in which the person entering the data and the person approving the data switch positions and mutually approve the data (mutual approval transactions) are also considered to pose a high risk of fraud due to insufficient internal checks and separation of duties.

[0005] Furthermore, when approvals are made in a circular manner by closed members within a closed approval route (circular approval entries), the familiarity of the members can lead to insufficient internal checks and separation of duties, increasing the risk of fraud.

[0006] Traditionally, these high-risk accounting entries could be extracted using the following two methods.

[0007] The first technique is to extract journal entries with a high risk of fraud based on the ID of the person who inputs the data and the ID of the person who approves the data.

[0008] The second method involves extracting accounting entries that pose a high risk of fraud by checking documents such as business procedure manuals to see whether rules for the approval process for accounting entries have been established. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] JP 2022-171190 A Summary of the Invention [Problem to be solved by the invention]

[0010] However, these conventional methods have the following problems, and may not be able to extract accounting entries that pose a high risk of fraud.

[0011] In other words, with the first method, it is not particularly difficult to extract self-approved journal entries, but it is not easy to extract mutually approved journal entries or circularly approved journal entries.

[0012] In the second method, for example, if the subject of the investigation is a subsidiary, it is necessary to obtain documents such as operating procedure manuals to confirm the approval route for accounting entries. However, obtaining documents is not easy due to the fact that approval routes often change due to organizational changes and personnel transfers, obtaining documents from subsidiaries is time-consuming and laborious, the format of documents differs from subsidiary to subsidiary, and recent regulations on the transfer of personal information from each country (GDPR, China's three data laws, etc.) have been strengthened.

[0013] Even if the documents are obtained, the accounting entries must be verified based on documents in different formats for each subsidiary, and the documents for each point in time must be compared with the accounting entries for the corresponding period. This makes the document review process extremely cumbersome, and it takes a lot of time and effort to verify that operations are being carried out in accordance with the documents.

[0014] For this reason, it is difficult to extract journal entries that pose a high risk of fraud using the second method.

[0015] The present invention has been made in consideration of the above circumstances, and relates to a journal entry extraction device, a journal entry extraction method, and a program that easily extract journal entries that are likely to be fraudulent. [Means for solving the problem]

[0016] In order to achieve the above object, the present invention takes the following measures.

[0017] In other words, the first aspect of the invention is a device that extracts journal entries that are likely to be fraudulent, and is equipped with an extraction unit that extracts, from among journal entries entered by an inputter and approved by an approver, journal entries where the inputter and approver are the same person, as journal entries that are likely to be fraudulent.

[0018] A second aspect of the invention is a device for extracting journal entries that are likely to be fraudulent, and includes a first extraction unit that extracts, from a plurality of journal entries input by a person inputting data and approved by an approver, pairs of a person who inputs the journal entry as the inputter and a person who approves the journal entry as the approver, in which the roles of inputter and approver are switched depending on the journal entry, and a second extraction unit that extracts journal entries in which one person in the pair is the inputter and the other person in the pair is the approver, as journal entries that are likely to be fraudulent.

[0019] A third aspect of the invention is a device for extracting journal entries that are likely to be fraudulent, and includes a third extraction unit that extracts a group consisting of three persons, a first person, a second person, and a third person, from a plurality of journal entries input by a person inputting data and approved by an approver, where in one journal entry, the first person is the person inputting data and the second person is the approver, in another journal entry, the second person is the person inputting data and the third person is the approver, and in yet another journal entry, the third person is the person inputting data and the first person is the approver, and a fourth extraction unit that extracts journal entries that have been approved within the group extracted by the third extraction unit as journal entries that are likely to be fraudulent.

[0020] A fourth aspect of the invention is a device for extracting journal entries that are likely to be fraudulent, and includes a fifth extraction unit that extracts a group of four or more persons in charge from a plurality of journal entries input by a person inputting data and approved by an approver, where each person in charge is the person inputting data for some journal entries and the approver for other journal entries for which he or she is not the inputter, and a sixth extraction unit that extracts, as journal entries that are likely to be fraudulent, journal entries in which any person in the group extracted by the fifth extraction unit is the person inputting data and any of the remaining persons in the group is the approver.

[0021] A fifth aspect of the invention is a device of any of the first to fourth aspects, further comprising a display control unit that visualizes and displays on a display the inputter, the approver, the approval path from the inputter to the approver, and the number of journal entries that fall into the approval path from a plurality of journal entries.

[0022] A sixth aspect of the invention is a method implemented by a processor for extracting journal entries that are likely to be fraudulent, in which the processor extracts, from among journal entries entered by an inputter and approved by an approver, journal entries where the inputter and approver are the same person, as journal entries that are likely to be fraudulent.

[0023] The seventh aspect of the invention is a method implemented by a processor to extract journal entries that are likely to be fraudulent, in which the processor extracts, from a plurality of journal entries input by a person inputting data and approved by an approver, pairs of a person who inputs the journal entry as the inputter and a person who approves the journal entry as the approver, in which the roles of inputter and approver are switched depending on the journal entry, and extracts journal entries in which one person in the pair is the inputter and the other person in the pair is the approver, as journal entries that are likely to be fraudulent.

[0024] An eighth aspect of the invention is a method implemented by a processor for extracting journal entries that are likely to be fraudulent, in which the processor extracts groups consisting of three persons, a first person, a second person, and a third person, from a plurality of journal entries entered by a person entering data and approved by an approver, where in one journal entry, the first person is the person entering data and the second person is the approver, in another journal entry, the second person is the person entering data and the third person is the approver, and in yet another journal entry, the third person is the person entering data and the first person is the approver, and extracts journal entries that have been approved within the extracted groups as journal entries that are likely to be fraudulent.

[0025] A ninth aspect of the invention is a method implemented by a processor for extracting journal entries that are likely to be fraudulent, in which the processor extracts groups of four or more persons from a plurality of journal entries entered by a person inputting data and approved by an approver, where each person is the person inputting data for some journal entries and the approver for other journal entries for which he or she is not the inputter, and extracts, as journal entries that are likely to be fraudulent, any person in the extracted group who is the person inputting data and any of the remaining persons in the group who is the approver.

[0026] A tenth aspect of the invention is a method of any one of the sixth to ninth aspects, in which a processor visualizes and displays, from a plurality of journal entries, the inputter, the approver, the approval path from the inputter to the approver, and the number of journal entries that fall into the approval path.

[0027] The invention of an eleventh aspect is a program for causing a processor to realize a function of extracting, from among journal entries entered by an inputter and approved by an approver, journal entries entered by the same person as the inputter and approver, as journal entries that may be fraudulent.

[0028] The invention of a twelfth aspect is a program for causing a processor to realize the following functions: from multiple journal entries input by a person inputting data and approved by an approver, extracting pairs of a person who inputs the journal entry as an inputter and a person who approves the journal entry as an approver, in which the roles of inputter and approver are switched depending on the journal entry; and extracting journal entries in which one person in the pair is the inputter and the other person in the pair is the approver, as possibly being fraudulent journal entries.

[0029] The invention of a thirteenth aspect is a program for causing a processor to realize a function of extracting a group consisting of three persons, a first person, a second person, and a third person, from a plurality of journal entries input by a person inputting data and approved by an approver, where in one journal entry, the first person is the person inputting data and the second person is the approver, in another journal entry, the second person is the person inputting data and the third person is the approver, and in yet another journal entry, the third person is the person inputting data and the first person is the approver, and a function of extracting journal entries that have been approved within the extracted group as journal entries that may be fraudulent.

[0030] The invention of a fourteenth aspect is a program for causing a processor to realize the following functions: extracting groups consisting of four or more persons in charge from multiple journal entries entered by a person inputting data and approved by an approver, where each person in charge is the person inputting data for some journal entries and the approver for other journal entries for which he or she is not the person inputting data; and extracting journal entries in which one of the persons in the extracted group is the person inputting data and one of the remaining persons in the group is the approver, as journal entries that may be fraudulent.

[0031] The invention of a fifteenth aspect is a program as described in any one of claims 11 to 14 for causing a processor to realize a function of visualizing and displaying, from a plurality of journal entries, the inputter, the approver, the approval path from the inputter to the approver, and the number of journal entries corresponding to the approval path. Effect of the Invention

[0032] According to the present invention, it is possible to realize a journal entry extraction device, a journal entry extraction method, and a program that can easily find and extract journal entries that are likely to be fraudulent. [Brief description of the drawings]

[0033] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a journal entry extraction device to which a journal entry extraction method according to the first embodiment is applied. [Diagram 2] FIG. 2 is a diagram showing an example of a display screen 50 that displays a table 40 showing a summary of the journal entries recorded in the journal entry database 32. [Diagram 3] FIG. 3 is a diagram showing a search condition specification field 70 displayed on the display 16 when the condition search button 52 is selected. [Figure 4] FIG. 4 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a self-approved journalization. [Diagram 5] FIG. 5 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a mutually approved journal entry. [Figure 6] FIG. 6 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a circular approval journalization. [Figure 7] FIG. 7 is a diagram illustrating a method for easily creating an approval route diagram from a complex approval route. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0034] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The drawings are schematic or conceptual. In the present specification and each drawing, elements similar to those described in the previous drawings are given the same reference numerals, and detailed and overlapping descriptions are omitted as appropriate.

[0035] (First embodiment) First, a journal entry extraction device to which a journal entry extraction method according to the first embodiment is applied will be described.

[0036] FIG. 1 is a block diagram showing an example of the configuration of a journal entry extraction device to which a journal entry extraction method according to the first embodiment is applied.

[0037] FIG. 2 is a diagram showing an example of a display screen 50 that displays a table 40 showing a summary of the journal entries recorded in the journal entry database 32.

[0038] 1, the journal entry extraction device 10 according to the first embodiment is a device that extracts journal entries that are likely to be fraudulent, and includes a CPU 12, a recording medium reading unit 14, a display 16, an input unit 17 such as a mouse or keyboard, a printer 18, a memory 20, and a storage device 30, all connected to one another by a bus 11. Although not limited thereto, such a journal entry extraction device 10 can be realized by a PC or a tablet terminal.

[0039] The CPU 12 is a computer that controls the operation of each circuit component in accordance with a program stored in the memory 20. This allows the journal entry extraction device 10 to operate in cooperation with software and hardware, as described below.

[0040] The storage device 30 is, for example, a solid state drive (SSD) or a hard disk drive (HDD), and stores a journalization database 32 in which journalization information is recorded.

[0041] For journal entries, information such as debit item, credit item, amount, and date is entered by the person inputting the entry, and the entry is stored in the journal entry database 32. The journal entries stored in the journal entry database 32 are approved by an approver. In order to clarify which person is the person who inputs the entry and which is the person who approves it, the ID of the person who inputs the entry is recorded in the input person ID 43, and the ID of the person who approved the entry is recorded in the approver ID 44 (see FIG. 2).

[0042] The memory 20 stores a program that realizes an extraction unit 21 and a display control unit 22 .

[0043] These programs may be pre-stored in the memory 20, or may be read from an external recording medium 13 such as a memory card via the recording medium reading unit 14 and stored in the memory 20. These programs are designed not to be rewritable.

[0044] In addition to the non-rewritable area in which such programs are stored, the memory 20 also has a writable data area 29 secured therein as an area for storing rewritable data.

[0045] The extraction unit 21 extracts journal entries according to instructions from the user. This extraction includes extracting, from among multiple journal entries, journal entries in which the inputter and approver are the same person as journal entries that may be fraudulent, in order to extract the self-approved journal entries described above. This extraction can be performed based on an ID that is uniquely assigned to each person in advance. This will be explained using FIG. 2.

[0046] Such a display screen 50 is displayed from the display 16.

[0047] In the table 40 on the display screen 50, records R1 to R16 are displayed in a list, each of which includes a debit item 41, a credit item 42, an inputter ID 43, an approver ID 44, and a number of items 45, based on each journal entry stored in the journal entry database 32.

[0048] Note that these items are merely examples and are not limited to these, and items can be added or deleted as necessary. For example, although not shown, since fraudsters often commit fraud by using family companies as business partners, a combination of the inputter ID 43 and the business partner name may be displayed.

[0049] Since the combination of the person who inputs the journal entry and the person who approves it is standardized depending on the content of the transaction, as shown in table 40 in Figure 2, the journal entry approval route for each transaction content expressed as a combination of subjects can be shown from the information on the journal entry inputter ID 43 and the approver ID 44, making it possible to display and check the status of internal control.

[0050] In addition to table 40, display screen 50 also displays a display entire combination button 51, a condition search button 52, a file extraction button 53, a preview button 54, a print start button 55, an approval route diagram button 56, a company code input field 57, a company name input field 58, a relevant period input field 59, a display all approval routes button 60, a display only self-approved transactions button 61, a display only transactions that constitute closed approval routes button 62, a display all transaction details button 63, a display only cash expenditure transactions button 64, a display only transactions that include end-of-period inventory items button 65, and a display only transactions that include subcontract expenses, consulting fees, etc. button 66.

[0051] When the user operates the input unit 17 to designate one of the buttons 51 to 56, the extraction unit 21 and the display control unit 22 operate as follows.

[0052] When the user selects the Display entire combination button 51, the display control unit 22 displays the approval path (inputter ID, approver ID, business partner name) through the multiple "account item combinations" displayed on the display 16, if the user wishes to check the combinations of account items after absorbing the differences in the combinations of account items.

[0053] When the condition search button 52 is selected, the display control unit 22 causes the display 16 to display a search condition specification field 70 as shown in FIG.

[0054] FIG. 3 is a diagram showing a search condition specification field 70 displayed on the display 16 when the condition search button 52 is selected.

[0055] When the user inputs search conditions such as debit account item 71, credit account item 72, department 73, inputter ID 74, approver ID 75, and business partner name 76 in the search condition specification field 70 and selects the OK button 77, the extraction unit 21 narrows down the journal entry data to be analyzed by the debit account item 71, credit account item 72, department 73, inputter ID 74, approver ID 75, and business partner name 76 in accordance with the input search conditions. Then, the display control unit 22 causes the display 16 to display record R, which is a summary of the narrowed down journal entries. This enables the user to analyze the approval path of journal entries related to a specific business flow, department, etc.

[0056] In addition to an OK button 77, the search condition specification field 70 is also provided with a Cancel button 78 and a Condition Cancel button 79. The user can cancel the contents entered in the search condition specification field 70 by selecting the Cancel button 78, or can cancel the conditions entered in the search condition specification field 70 by selecting the Condition Cancel button 79.

[0057] Returning to FIG. 2, file cut button 53 is a button for receiving an instruction to cut out the display contents of table 40 into a file.

[0058] The preview button 54 is a button for previewing the print range. When the preview button 54 is selected, the display control unit 22 causes the display 16 to display a preview of the print range.

[0059] The print start button 55 is a button for receiving an instruction to start printing. The user can specify the print start button 55 to cause the printer 18 to start printing.

[0060] The approval route diagram button 56 is a button for displaying the approval route on the display 16. When the approval route diagram button 56 is specified, the extraction unit 21 extracts the approval route, and the display control unit 22 displays the approval route diagram on the display 16. The approval route diagram will be described in detail with reference to Figs. 4, 5, and 6.

[0061] When the user operates the input unit 17 to input conditions into the input fields 57 to 59, the extraction unit 21 and the display control unit 22 operate as follows.

[0062] The company code input field 57 is an input field for receiving designation of which company code's journal entry summary is to be displayed from table 40. When the user inputs a company code into the company code input field 57, the extraction unit 21 accesses the journal entry database 32 of the input company code, and the display control unit 22 displays record R of the journal entry summary of the accessed company code from table 40.

[0063] The company name input field 58 is an input field for accepting designation of the journal entry summary of which company name is to be displayed from the table 40. When the user inputs a company name into the company name input field 58, the extraction unit 21 accesses the journal entry database 32 of the input company name, and the display control unit 22 displays the record R of the journal entry summary of the accessed company code from the table 40.

[0064] The user can display the journal entry summary record R for the desired company from table 40 by inputting a company code in company code input field 57 or a company name in company name input field 58.

[0065] The relevant period input field 59 is an input field for receiving designation of the period of the summary to be displayed from the table 40. When the user designates a period from the relevant period input field 59, the extraction unit 21 extracts a summary of the journal entries for the designated period, and the display control unit 22 displays the record R of the extracted summary from the table 40.

[0066] In addition, when the user operates the input unit 17 to specify one of the buttons 60 to 62, the extraction unit 21 narrows down the journal entries according to the specified condition, and the display control unit 22 displays the result from the table 40.

[0067] When the display button 61 is selected by the user, the extraction unit 21 extracts journal entries for only self-approved transactions, and the display control unit 22 displays record R of the summary of the journal entries extracted by the extraction unit 21 from the table 40. In this way, by selecting the display button 61, the user can view the journal entries for self-approved transactions that are suspected of being fraudulent.

[0068] When the display button 62 is selected by the user, the extraction unit 21 extracts journal entries for only the transactions that constitute a closed approval path, and the display control unit 22 displays the record R of the summary of the journal entries extracted by the extraction unit 21 from the table 40. A closed approval path is, for example, the above-mentioned mutual approval or circular approval path. In this way, by selecting the display button 62, the user can view a group of journal entries that constitute a closed approval path in which internal control may be weakened.

[0069] When the user selects the display button 60, all approval paths are targeted, and the display control unit 22 causes the record R to be displayed on the display 16, regardless of the type of approval path. In this way, by selecting the display button 60, the user can view the journal entry, regardless of the type of approval path.

[0070] In addition, when the user operates the input unit 17 to specify one of the buttons 63 to 66, the extraction unit 21 narrows down the journal entries according to the specified condition, and the display control unit 22 displays the result from the table 40.

[0071] Display buttons 63 to 66 are buttons provided for extracting accounting entries limited to transactions of withdrawal of cash, inventory valuation, and payment of subcontract expenses and consulting fees, which are considered to be typical cases of fraudulent self-approved accounting entries.

[0072] When the display button 64 is selected by the user, the extraction unit 21 extracts journal entries for cash and deposit expenditure transactions only, and the display control unit 22 displays the records of the summaries of the journal entries extracted by the extraction unit 21 from the table 40. In this way, by selecting the display button 64, the user can view journal entries for cash and deposit expenditure transactions only.

[0073] When the user selects the display button 65, the extraction unit 21 extracts journal entries for only transactions that include end-of-period inventory items, and the display control unit 22 displays record R of the summary of the journal entries extracted by the extraction unit 21 from the table 40. In this way, by selecting the display button 65, the user can view the journal entries for transactions that include end-of-period inventory items.

[0074] When the display button 66 is selected by the user, the extraction unit 21 extracts the journal entries of the transactions including subcontract costs, consulting fees, etc., and the display control unit 22 displays the record R of the summary of the journal entries extracted by the extraction unit 21 from the table 40. In this way, the user can view the journal entries of the transactions including subcontract costs, consulting fees, etc., by selecting the display button 66.

[0075] When the user selects the display button 63, all transaction types are targeted, and the display control unit 22 causes the display 16 to display the record R regardless of the transaction type. In this way, by selecting the display button 63, the user can view the journal entry regardless of the transaction type.

[0076] Next, an example of the operation of the journal entry extraction device 10 according to the first embodiment configured as above will be described.

[0077] In this embodiment, a case will be described in which a self-approved journal entry is extracted from among a plurality of journal entries.

[0078] When extracting self-approved journal entries, the user inputs the same ID into the inputter ID 74 and the approver ID 75 in the search condition specification field 70 illustrated in FIG.

[0079] As a result, the extraction unit 21 extracts journal entries whose inputter ID and approver ID are the same from among the many journal entries stored in the journal entry database 32. For example, when "USER_A" is entered into the inputter ID 74 and approver ID 75 in the search condition specification field 70, the extraction unit 21 extracts journal entries in which a person with the ID "USER_A" is both the inputter and approver. Also, when "USER_B" is entered into the inputter ID 74 and approver ID 75 in the search condition specification field 70, the extraction unit 21 extracts journal entries in which a person with the ID "USER_B" is both the inputter and approver.

[0080] Record R4 shown in Figure 2 is an example of a record corresponding to a journal entry in which a person with the ID "USER_A" is both the inputter and approver, and record R10 is an example of a record corresponding to a journal entry in which a person with the ID "USER_B" is both the inputter and approver.

[0081] In this manner, in this embodiment, the extraction unit 21 can extract self-approved journal entries from among a plurality of journal entries, based on an ID that is uniquely assigned in advance to each person in charge.

[0082] 2 shows a portion of all records (from the first record R1 to the 16th record R16). Therefore, while records other than self-approved journal entries are also shown in FIG. 2, the extraction unit 21 also has a function of displaying only the extracted self-approved journal entries on the display 16. If the user wishes to use this function, the extraction unit 21 will display only the extracted self-approved journal entries on the display 16 by designating the radio button 61 labeled "Display only self-approved transactions" on the display screen shown in FIG. 2.

[0083] The display control unit 22 also visualizes the inputter, approver, approval path from the inputter to the approver, and the number of journal entries corresponding to the approval path for the self-approved journal entries extracted by the extraction unit 21, and displays them on the display 16. An example is shown in FIG.

[0084] FIG. 4 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a self-approved journalization.

[0085] The approval path diagrams illustrated in FIG. 4(a) and FIG. 4(b) can be weighted directed graphs in which the approval path from the inputter to the approver is represented by an arc that starts at the inputter and ends at the approver. The end point is indicated by an arrow. The arcs are displayed with different thicknesses depending on the number of journal entries that fall under the approval path. For example, the more journal entries that fall under the approval path, the thicker the arc is displayed. Also, a number indicating the corresponding number is displayed near each arc.

[0086] In the approval route diagram shown in Figure 4(a), the person with the ID "USER_F" is both the inputter and the approver, so the arc starts at "USER_F" and returns to "USER_F". The number of applicable journal entries, "15", is displayed near the arc.

[0087] In the approval route diagram shown in Figure 4(b), the person with the ID "USER_A" is both the inputter and the approver, so the arc starts at "USER_A" and returns to "USER_A." The number of applicable journal entries, "39," is displayed near the arc.

[0088] Since "39" is more than "15," the arcs are displayed thicker in FIG. 4(b) than in FIG. 4(a).

[0089] In addition, since approval paths that are used infrequently are irregular approvals and have a high possibility of fraud, only approval paths whose frequency is below a preset threshold value may be highlighted.

[0090] By expressing the approval route diagram as a weighted directed graph in this way, it becomes possible to visually easily grasp the approval route and status of a journal entry, which is otherwise difficult to visualize.

[0091] Second Embodiment Next, a journal entry extraction device to which the journal entry extraction method according to the second embodiment is applied will be described.

[0092] The journal entry extraction device to which the journal entry extraction method according to the second embodiment is applied is a modified example of the extraction unit 21 in the journal entry extraction device to which the journal entry extraction method according to the first embodiment is applied.

[0093] Therefore, in the following, configurations different from those in the first embodiment will be described, but other configurations are the same as those described in the first embodiment with reference to FIG. 1, and therefore description thereof will be omitted.

[0094] That is, in this embodiment, the extraction unit 21 performs a first extraction process and a second extraction process in order to extract the above-mentioned mutually approved journal entries.

[0095] The first extraction process extracts, from among the many journal entries stored in the journal entry database 32, pairs of a person who inputs a journal entry as an inputter and a person who approves the journal entry as an approver, whose roles as inputter and approver are switched depending on the journal entry.

[0096] The second extraction process extracts, as a possible fraudulent journal entry, a journal entry in which one person in a pair is the person who inputs the data and the other person in the pair is the person who approves the data.

[0097] These extractions can also be done based on the ID that is uniquely assigned to each person in advance. This will be explained using Figure 2.

[0098] The extraction unit 21 first recognizes pairs of a person who inputs a journal entry as an inputter, i.e., an inputter, and a person who approves the journal entry as an approver, i.e., an approver, through the first extraction process. This is done by extracting combinations of inputter ID 43 and approver ID 44. In the example shown in FIG. 2, the combinations of pairs of inputter ID 43 and approver ID 44 include "USER_B"-"USER_A", "USER_C"-"USER_A", "USER_D"-"USER_D", "USER_A"-"USER_A", "USER_A"-"USER_B", and "USER_B"-"USER_B". Note that a hyphen (-) indicates that what precedes and follows it constitutes a pair.

[0099] Among these combinations, "USER_B"-"USER_A" and "USER_A"-"USER_B" have the combinations of the input person ID 43 and the approver ID 44 reversed according to the journalization.

[0100] In this manner, the extraction unit 21 extracts the pair of "USER_A" and "USER_B" as a pair in which the positions of inputter and approver are switched according to the journalization through the first extraction process.

[0101] Next, the extraction unit 21 extracts, by the second extraction process, journal entries in which one of the pair of responsible persons is the inputter and the other of the pair is the approver, as journal entries that may be fraudulent. For example, referring to Fig. 2, journal entries corresponding to records R9, 11 to 13 in which "USER_A" is the inputter and "USER_B" is the approver, and journal entries corresponding to records R1, 5 to 7 in which "USER_B" is the inputter and "USER_A" is the approver are extracted as journal entries that may be fraudulent.

[0102] In this manner, in this embodiment, the extraction unit 21 can extract mutually approved journals from among a plurality of journals based on an ID uniquely assigned in advance to each person in charge.

[0103] Fig. 5 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a mutually approved journal entry. Fig. 5 can also be a weighted directed graph.

[0104] According to the approval path diagram illustrated in FIG. 5, it is possible to visually and easily grasp that a person in charge with the ID "USER_A" and a person in charge with the ID "USER_B" are in a mutual approval relationship.

[0105] (Third embodiment) Next, a journal entry extraction device to which a journal entry extraction method according to the third embodiment is applied will be described.

[0106] The journal entry extraction device to which the journal entry extraction method according to the third embodiment is applied is a modified example of the extraction unit 21 in the journal entry extraction device to which the journal entry extraction method according to the first embodiment is applied.

[0107] Therefore, in the following, configurations different from those in the first embodiment will be described, but other configurations are the same as those described in the first embodiment with reference to FIG. 1, and therefore description thereof will be omitted.

[0108] That is, in this embodiment, the extraction unit 21 performs a third extraction process and a fourth extraction process in order to extract the above-mentioned circulating approval journal entry.

[0109] The third extraction process extracts groups consisting of three persons (a first person, a second person, and a third person) from among the numerous journal entries stored in the journal entry database 32, where in one journal entry, the first person is the person who inputs the data and the second person is the person who approves it, in another journal entry, the second person is the person who inputs the data and the third person is the person who approves it, and in yet another journal entry, the third person is the person who inputs the data and the first person is the person who approves it.

[0110] To explain this in more detail with reference to Fig. 2, in record R14, the inputter ID is "USER_E" and the approver ID is "USER_F". In record R15, the inputter ID is "USER_G" and the approver ID is "USER_E". In record R16, the inputter ID is "USER_F" and the approver ID is "USER_G".

[0111] In this case, the three personnel ("USER_E", "USER_F", and "USER_G") correspond to a group in which, for one journal entry (the journal entry corresponding to record R14), the first personnel (USER_E) is the data inputter and the second personnel (USER_F) is the approver, for another journal entry (the journal entry corresponding to record R16), the second personnel (USER_F) is the data inputter and the third personnel (USER_G) is the approver, and for yet another journal entry (the journal entry corresponding to record R15), the third personnel (USER_G) is the data inputter and the first personnel (USER_E) is the approver.

[0112] The third extraction process extracts such groups from among the many journal entries stored in the journal entry database 32.

[0113] In the fourth extraction process, journal entries that have been approved within the group extracted by the third extraction process are extracted as journal entries that may be fraudulent entries.

[0114] Therefore, in the above example, the extraction unit 21 extracts, through the fourth extraction process, the journal entries corresponding to the records R14, R15, and R16 as possibly fraudulent journal entries.

[0115] In this manner, in this embodiment, the extraction unit 21 can extract cyclic approval journal entries from among a plurality of journal entries, based on the ID uniquely assigned in advance to each person in charge.

[0116] Fig. 6 is a diagram showing an approval path diagram, which is an example of a visualized display displayed on the display 16 by the display control unit 22 in the case of a circular approval journalization. Fig. 6 can also be a weighted directed graph.

[0117] According to the approval path diagram shown in FIG. 6, it is easy to visually understand that the three people, the person with the ID "USER_E", the person with the ID "USER_F", and the person with the ID "USER_G", are in a circular approval relationship.

[0118] In the above, the extraction of cyclic approval journal entries consisting of a group of three people has been described, but cyclic approval journal entries consisting of a group of four or more people can also be extracted in the same way.

[0119] When extracting cyclic approval journal entries consisting of groups of four or more people, the extraction unit 21 executes the fifth extraction process described below instead of the third extraction process, and executes the sixth extraction process described below instead of the fourth extraction process.

[0120] The fifth extraction process extracts, from among the numerous journal entries stored in the journal entry database 32, groups consisting of four or more persons, each of whom acts as an inputter for some journal entries and as an approver for other journal entries for which he or she is not the inputter.

[0121] The sixth extraction process extracts, as possibly fraudulent entries, journal entries in which one person in the group extracted by the fifth extraction process is the inputter and any one of the remaining people in the group is the approver.

[0122] By doing this, in this embodiment, the extraction unit 21 can also extract cyclical approval journal entries consisting of groups of four or more people from among multiple journal entries, based on a unique ID that is pre-assigned to each person in charge.

[0123] The display control unit 22 also displays the results as an approval path diagram on the display 16, making it possible to visually grasp the relationships between each of the staff members in a complex circular approval transaction carried out by a group of four or more staff members.

[0124] As described above, according to the present invention, it is possible to provide a journal entry extraction device, a journal entry extraction method, and a program that extract journal entries that are likely to be fraudulent by analyzing the approval path of the journal entries.

[0125] According to the journal entry extraction device, journal entry extraction method, and program of the present invention, specifically, it is possible to visualize and check approval paths that may have control deficiencies by narrowing down to only self-approved transactions and transactions with closed approval paths. This makes it possible to grasp the status of internal checks, judge the severity of risk, and select journal entries that should be checked with priority.

[0126] In addition, because the number of parties involved in a closed approval route is highly correlated with the effectiveness of control, by narrowing down approval routes to those consisting of a specified number of people or less, it is possible to extract approval routes with higher risk.

[0127] (Variations in the creation of approval flow diagrams) In the first, second, and third embodiments, the approval path diagrams are described using Figures 4, 5, and 6. However, the approval path diagrams shown in Figures 4, 5, and 6 are relatively simple examples, and the actual approval path may be more complicated.

[0128] Therefore, in this modified example, a method for easily creating an approval route diagram even when the approval route is complicated will be described.

[0129] FIG. 7 is a diagram illustrating a method for easily creating an approval route diagram from a complex approval route.

[0130] In this method, first, as shown in Fig. 7(a), the persons in charge (A to L) who are inputters or approvers are represented as vertices of a directed graph. In addition, each approval path between each vertex is represented as an arc of the directed graph. Furthermore, the number of transactions that follow each approval path is represented as the weight of the arc.

[0131] Generally, strongly connected component decomposition is used to express only the approval path of the closed loop of interest, but in this invention, the closed loop is emphasized by excluding from the display the agents whose in-degree or out-degree is a vertex of 0 (i.e., agents who do not appear as either approvers or inputters). In Fig. 7(a), the agents whose in-degree or out-degree is a vertex of 0 (zero) are I, J, K, and L. Therefore, the approval path diagram shown in Fig. 7(b) is obtained from the approval path diagram shown in Fig. 7(a) by excluding the agents I, J, K, and L from the display.

[0132] This process is repeated until there is no more person in charge whose vertex has an in-degree or out-degree of 0. In the approval path diagram shown in Figure 7(b), person in charge H corresponds to the person in charge whose vertex has an in-degree or out-degree of 0. Therefore, person in charge H is removed from the display, and the approval path diagram shown in Figure 7(c) is obtained from the approval path diagram shown in Figure 7(b).

[0133] The approval path diagram shown in Figure 7(c) can be modeled with two loops (the first loop consisting of A, B, and C, and the second loop consisting of E, F, and G) and a bridge (D) connecting these two loops.

[0134] If the bridge (D) is removed from the approval path diagram shown in Figure 7(c), it can be decomposed into two closed paths (strongly connected component decomposition), but the bridge (D) is left in as it corresponds to a person connecting the first closed path and the second closed path. Even without removing the bridge (D), problematic approval paths can be highlighted, so a sufficient effect can be obtained for practical purposes. In addition, by not removing the bridge (D), i.e., by not decomposing the first closed path and the second closed path, the calculation load can be reduced compared to the algorithm used for strongly connected component decomposition.

[0135] Although the best mode for carrying out the present invention has been described above with reference to the attached drawings, the present invention is not limited to such a configuration. Within the scope of the technical idea of ​​the invention defined in the claims, a person skilled in the art may conceive of various modifications and alterations, and it is understood that such modifications and alterations also belong to the technical scope of the present invention. [Industrial Applicability]

[0136] The path analysis technique applied in the journal entry extraction device, journal entry extraction method, and program of the present invention is not limited to application in the field of journal entry extraction, but can also be applied to technologies in other fields that require path analysis. [Explanation of symbols]

[0137] 10 Journal entry extraction device 11 Bus 12 CPU 13 External recording media 14 Recording medium reading unit 16 Display 17 Input section 18 Printers 20 Memory 21 Extraction part 22 Display control unit 29 Writable Data Area 30 Storage device 32 Journal entry database 40 Table 41 Debit account 42 Credit account 45 items 50 display screen 51 Combination full display button 52 Conditional search button 53 File extraction button 54 Preview button 55 Print start button 56 Approval flow chart button 57 Company code input field 58 Company name input field 59 Input field for applicable period 60 All approval route display button 61 Show only self-approved transactions button 62 Button to display only transactions that constitute a closed approval path 63 Transaction details display button 64 Display only cash and deposit transactions button 65 Display only transactions that include end-of-period inventory items button 66 Show only transactions including subcontracting and consulting fees button 70 Search criteria specification field 71 Debit Accounts 72 Credit Accounts 73 departments 76 Customer Name 77 OK button 78 Cancel button 79 Cancel condition button

Claims

1. A device for extracting journal entries that are likely to be fraudulent, comprising: An extraction unit that extracts, from among journal entries input by an inputter and approved by an approver, journal entries in which the inputter and the approver are the same person, as journal entries that may be fraudulent journal entries; A device comprising: a display control unit that visualizes and displays on a display the person who inputs the data, the person who approves the data, the approval path from the person who inputs the data to the person who approves the data, and the number of journal entries that fall under the approval path from a plurality of journal entries.

2. A device for extracting journal entries that are likely to be fraudulent, comprising: a first extraction unit that extracts pairs of a person who has input the journal entry as the inputter and a person who has approved the journal entry as the approver from a plurality of journal entries input by an inputter and approved by an approver, in which the roles of the inputter and the approver are switched depending on the journal entry; and a second extraction unit that extracts, as a possible fraudulent journal entry, a journal entry in which one person in the pair is the inputter and the other person in the pair is the approver.

3. A device for extracting journal entries that are likely to be fraudulent, comprising: a third extraction unit that extracts a group consisting of three persons, a first person in charge, a second person in charge, and a third person in charge, from a plurality of journal entries input by an inputter and approved by an approver, where in one journal entry, the first person in charge is the inputter and the second person in charge is the approver, in another journal entry, the second person in charge is the inputter and the third person in charge is the approver, and in yet another journal entry, the third person in charge is the inputter and the first person in charge is the approver; and a fourth extraction unit that extracts journal entries that have been approved within the group extracted by the third extraction unit as journal entries that may be fraudulent entries.

4. A device for extracting journal entries that are likely to be fraudulent, comprising: a fifth extraction unit that extracts a group consisting of four or more persons in charge from a plurality of journal entries input by an inputter and approved by an approver, each person in charge being an inputter for some journal entries and an approver for other journal entries for which he / she is not the inputter; and a sixth extraction unit that extracts, as a possible fraudulent entry, a journal entry in which any of the persons in the group extracted by the fifth extraction unit is the inputter and any of the remaining persons in the group is the approver.

5. The device according to any one of claims 2 to 4, further comprising a display control unit that visualizes and displays on a display the person who inputs the information, the person who approves the information, the approval path from the person who inputs the information to the person who approves the information, and the number of journal entries that fall under the approval path from a plurality of journal entries.

6. A method for extracting journal entries that are likely to be fraudulent, the method comprising: The processor, Extracting, from among the journal entries input by an inputter and approved by an approver, journal entries in which the inputter and the approver are the same person, as journal entries that may be fraudulent journal entries; A method for visually displaying, from a plurality of journal entries, the person who inputs the data, the person who approves the data, the approval path from the person who inputs the data to the person who approves the data, and the number of journal entries that fall under the approval path.

7. A method for extracting journal entries that are likely to be fraudulent, the method comprising: The processor, From a plurality of journal entries input by an inputter and approved by an approver, pairs of a person who has input the journal entry as the inputter and a person who has approved the journal entry as the approver are extracted, in which the roles of the inputter and the approver are switched depending on the journal entry; The method extracts, as a possible fraudulent entry, a journal entry in which one person in the pair is the inputter and the other person in the pair is the approver.

8. A method for extracting journal entries that are likely to be fraudulent, the method comprising: The processor, Extracting a group consisting of three persons, a first person in charge, a second person in charge, and a third person in charge, from a plurality of journal entries input by an inputter and approved by an approver, where in one journal entry, the first person in charge is the inputter and the second person in charge is the approver, in another journal entry, the second person in charge is the inputter and the third person in charge is the approver, and in yet another journal entry, the third person in charge is the inputter and the first person in charge is the approver; A method for extracting approved journal entries within the extracted group as journal entries that may be fraudulent entries.

9. A method for extracting journal entries that are likely to be fraudulent, the method comprising: The processor, Extracting a group consisting of four or more persons in charge from a plurality of journal entries input by a person inputting data and approved by a person approving data, each person in charge being an inputter for some journal entries and an approver for other journal entries for which he / she is not an inputter; The method extracts, as a possibly fraudulent journal entry, a journal entry in which any of the persons in the extracted group is the inputter and any of the remaining persons in the group is the approver.

10. The processor, 10. The method according to claim 7, further comprising visualizing and displaying, from a plurality of journal entries, the person who inputs the journal entry, the person who approves the journal entry, an approval path from the person who inputs the journal entry to the person who approves the journal entry, and the number of journal entries that fall under the approval path.

11. A function of extracting, from among journal entries input by an inputter and approved by an approver, journal entries in which the inputter and the approver are the same person, as journal entries that may be fraudulent; A program for enabling a processor to realize a function of visualizing and displaying, from multiple journal entries, the person who inputs the data, the person who approves the data, the approval path from the person who inputs the data to the person who approves the data, and the number of journal entries that fall into the approval path.

12. A function of extracting pairs of a person who has input a journal entry as an inputter and a person who has approved the journal entry as an approver from a plurality of journal entries input by an inputter and approved by an approver, in which the roles of the inputter and the approver are switched depending on the journal entry; A program for causing a processor to realize a function of extracting journal entries in which one person in the pair is the inputter and the other person in the pair is the approver, as journal entries that may be fraudulent.

13. A function of extracting a group consisting of three persons, a first person in charge, a second person in charge, and a third person in charge, from a plurality of journal entries input by an inputter and approved by an approver, where in one journal entry, the first person in charge is the inputter and the second person in charge is the approver, in another journal entry, the second person in charge is the inputter and the third person in charge is the approver, and in yet another journal entry, the third person in charge is the inputter and the first person in charge is the approver; A program for causing a processor to realize a function of extracting journal entries approved within the extracted group as journal entries that may be fraudulent.

14. A function for extracting a group consisting of four or more persons in charge from multiple journal entries input by a person inputting data and approved by a person approving data, where each person in charge is an inputter for some journal entries and an approver for other journal entries for which he or she is not an inputter; A program for causing a processor to realize a function of extracting, as a possibly fraudulent entry, any entry in which any of the persons in the extracted group is the inputter and any of the remaining persons in the group is the approver.

15. 15. The program according to claim 12, for causing a processor to realize a function of visualizing and displaying, from a plurality of journal entries, the person who inputs the information, the person who approves the information, the approval path from the person who inputs the information to the person who approves the information, and the number of journal entries that fall under the approval path.

Images