JP2021182359A - Expense inspection device, expense inspection method, and program - Google Patents

Abstract

To properly perform an inspection of an improper expense application, in contrast to the prior art.SOLUTION: An expense inspection device 1 is provided with: an expense receiving unit 121 that receives expense information specifying contents of an expense application; an inspection unit 131 that acquires an inspection result by performing an inspection concerning fraud in the expense information received by the expense receiving unit 121 by using external information other than expense information, or expense information other than the expense information received by the expense receiving unit 121; and a result output unit 141 that outputs the inspection result acquired by the inspection unit 131. The expense inspection device properly performs an inspection of an improper expense application.SELECTED DRAWING: Figure 3

Description

The present invention relates to an expense inspection device or the like that inspects an applied expense information for fraud and outputs an inspection result.

Conventionally, there has been an accounting system that can check the objectivity and consistency of transaction data between companies, between companies and individuals, or socially (see Patent Document 1).

Japanese Unexamined Patent Publication No. 2017-97669

However, in the prior art, the inspection could not be properly performed for an inappropriate expense application.

The expense inspection device of the first invention uses an expense reception unit that receives expense information that specifies the content of an expense application, external information other than the expense information, or expense information other than the expense information received by the expense reception unit. This is an expense inspection device including an inspection unit that inspects fraudulent expense information received by the expense reception unit and acquires inspection results, and a result output unit that outputs the inspection results acquired by the inspection unit.

With such a configuration, by using the expense information other than the external information or the applied expense information, the inspection can be appropriately performed for the inappropriate expense application.

Further, in the expense inspection device of the second invention, for the first invention, the expense reception unit receives the expense information in association with the applicant identifier that identifies the applicant of the expense, and the external information is the external information. It is information corresponding to each of two or more applicants, has an applicant set including two or more applicant information associated with the applicant identifier, and the inspection department corresponds to the expense information received by the expense reception department. Expense inspection to obtain the applicant information corresponding to the attached applicant identifier from the applicant set, use the applicant information to inspect the fraudulent expense information received by the expense reception department, and obtain the inspection result. It is a device.

With such a configuration, by using the applicant information, inspection can be appropriately performed for inappropriate expense application.

Further, in the expense inspection device of the third invention, for the second invention, the expense information has the usage date information for specifying the date when the expense was used, and the applicant information is identified by the applicant identifier. It has one or more attendance information including working day information that specifies the working day of the applicant, and the inspection department is the attendance information corresponding to the applicant identifier associated with the expense information, and the date of use. It is an expense inspection device that inspects fraudulent expense information received by the expense reception department using attendance information that includes the same work day information as the information, and acquires inspection results.

With such a configuration, by using the attendance information of the applicant, inspection can be appropriately performed for inappropriate expense application.

Further, in the expense inspection device of the fourth invention, for the second invention, the expense information has the usage date information for specifying the date when the expense was used, and the applicant information is identified by the applicant identifier. The applicant has one or more entry / exit information that specifies the date and time when the applicant entered / exited his / her place of work, and the inspection department corresponds to the applicant identifier associated with the expense information. It is information, and it is an expense inspection device that inspects for fraudulent expense information received by the expense reception department using the entrance / exit information corresponding to the date specified by the usage date information possessed by the expense information, and acquires the inspection result. be.

With this configuration, by using the applicant's entry / exit information, inspection can be performed appropriately for inappropriate expense applications.

Further, in the expense inspection device of the fifth invention, for the second invention, the expense information has the usage date information for specifying the date when the expense was used, and the applicant information is the business activity of the applicant. Applicants who have business activity information including business day information that specifies the day of the visit and one or more point identifiers that specify the visited point, and the inspection department is associated with the expense information received by the expense reception department. It is the business activity information corresponding to the identifier, and it has the same business day information as the usage date information of the expense information. It is an expense inspection device that conducts inspections and obtains inspection results.

With such a configuration, by using the business activity information of the applicant, it is possible to appropriately inspect an inappropriate expense application.

Further, in the expense inspection device of the sixth invention, for the second invention, the expense information has the usage date information for specifying the date when the expense was used, and the applicant information was used by the applicant. It has card details information that is information indicating the details of the credit card and includes information on the date of use, which is the date when the credit card was used, and the inspection department is associated with the expense information received by the expense reception department. Using the card statement information that corresponds to the applicant identifier and has the same usage date information as the expense information, the expense reception department inspects for fraudulent expense information. It is an expense inspection device that acquires inspection results.

With this configuration, by using the applicant's credit card statement information, inspection can be appropriately performed for inappropriate expense applications.

Further, in the expense inspection device of the seventh invention, for the second invention, the expense information has the usage date information for specifying the date when the expense was used, and the applicant information was used by the applicant. It is one or more business vehicle position information that specifies the position of the business vehicle, and has one or more business vehicle position information associated with the usage date information that specifies the date when the business vehicle was used. One or more business vehicle position information corresponding to the applicant identifier associated with the expense information received by the expense reception department, and is associated with the same usage date information as the usage date information possessed by the expense information 1. It is an expense inspection device that acquires the above-mentioned business vehicle position information, inspects the fraudulent expense information using the business vehicle position information, and acquires the inspection result.

With such a configuration, by using the business vehicle position information of the applicant, inspection can be appropriately performed for inappropriate expense application.

Further, in the expense inspection device of the eighth invention, for the second invention, the expense information corresponds to the expense type identifier indicating the expense type, and the applicant information is possessed by the card owned by the applicant. It also has a card usage expense storage unit that has card information and stores one or more expense type identifiers scheduled to be paid by card, and the inspection department responds to the expense information received by the expense reception department. Check whether the expense type identifier is stored in the card usage expense storage unit as the expense type identifier corresponding to the owned card information corresponding to the applicant identifier associated with the expense information received by the expense reception department. , It is an expense inspection device that acquires inspection results.

With such a configuration, by using the information of the card owned by the applicant, the inspection can be appropriately performed for the inappropriate expense application.

Further, in the expense inspection device of the ninth invention, with respect to the first invention, the external information includes the company regulation information regarding the regulation of the company to which the applicant belongs, and the inspection department uses the company regulation information. It is an expense inspection device that inspects whether the expense information received by the expense reception department conforms to company regulations and acquires the inspection results.

With this configuration, by using company-specified information, inspections can be appropriately performed for inappropriate expense applications.

Further, the expense inspection device of the tenth invention has, with respect to the first invention, the expense information includes the usage date information for specifying the date on which the expense was incurred, the departure identifier, the arrival identifier, and the charge information 1. It has the transportation cost information having the above section information, the external information includes the transportation means information including the calculation information for calculating the transportation cost, and the inspection department has the expense information received by the expense reception department and the transportation means. It is an expense inspection device that inspects for fraudulent transportation expenses using information and obtains inspection results.

With such a configuration, by using the transportation means information, inspection can be appropriately performed for inappropriate expense application.

Further, in the expense inspection device of the eleventh invention, with respect to the first invention, the external information has a store set which is one or more store information having a store identifier for identifying the store, and the expense information is , Has a store identifier that identifies the store that has spent the expense, and the inspection department uses the store identifier that the expense information received by the expense reception department and the store information that the store set has to inspect for fraudulent expense information. It is an expense inspection device that performs and acquires inspection results.

With such a configuration, by using a store set, inspection can be appropriately performed for inappropriate expense applications.

Further, the expense inspection device of the twelfth invention has, with respect to the eleventh invention, the store information has the amount information, and the expense information has the expense amount information for specifying the amount spent. The inspection department inspects the fraudulent expense information using the amount information paired with the store identifier of the expense information received by the expense reception department and the expense amount information of the expense information, and acquires the inspection result. It is an expense inspection device.

With such a configuration, inappropriate spending in the store can be appropriately detected.

Further, in the expense inspection device of the thirteenth invention, with respect to the first invention, the expense information has the other party identifier that identifies the other party who used the expense, and the external information is the use of the expense. It has one or more other party information about the candidate of the other party, and the inspection department uses the other party identifier of the expense information received by the expense reception department and one or more other party information of the external information, and the expense information. It is an expense inspection device that inspects fraud and obtains inspection results.

With such a configuration, it is possible to appropriately detect inappropriate spending for an inappropriate partner.

Further, the expense inspection device of the fourteenth invention has, for the first invention, expense information has an expense type identifier indicating travel expenses and usage date information, and external information has working day information. An application that has one or more attendance information including one or more, and one or more of one or more entry / exit information that specifies the date and time of entry / exit, and is associated with the expense information received by the expense reception department. An expense inspection device that acquires attendance information or entry / exit information that corresponds to the person identifier and date of use information, inspects for fraudulent travel expenses using attendance information or entry / exit information, and acquires inspection results. be.

With such a configuration, inappropriate business trip expenses can be appropriately detected.

Further, the expense inspection device of the fifteenth invention has, with respect to the first invention, the expense information includes the usage date information for specifying the date on which the expense was incurred, the departure identifier, the arrival identifier, and the charge information. The external information is a group of applicants having transportation cost information having one or more section information, transportation means information including calculation information for calculating transportation cost, or one or more applicant information, and an applicant. The information includes regular section information that identifies the applicant's regular section, or business activity information that includes business day information that identifies the applicant's business activity date and one or more point identifiers that identify the visited point. However, the inspection department acquires the applicant information corresponding to the applicant identifier associated with the expense information related to the transportation expenses received by the expense reception department, and uses the applicant information or the applicant information and the means of transportation information. It is an expense inspection device that inspects fraud related to transportation expense claims and obtains inspection results.

With such a configuration, inappropriate transportation costs can be appropriately detected.

Further, in the expense inspection device of the sixteenth invention, for the first invention, the inspection department has the same expense information received by the expense reception unit and the applicant identifier associated with the expense information. It is determined whether or not one or more of the past expense information stored in the expense information storage unit associated with the applicant identifier is close enough to satisfy a predetermined condition, and the inspection result is determined. It is an expense inspection device to be acquired.

With such a configuration, it is possible to appropriately detect double application, reuse of receipts, and the like.

Further, in the expense inspection device of the seventeenth invention, for the second invention, the expense information includes the usage date information for specifying the date of staying at the accommodation facility, and the applicant information includes the applicant's work. 1 The external information includes the company regulation information having the above-mentioned entrance / exit information and the overtime time information that specifies the overtime time that allows accommodation, and the inspection department corresponds to the applicant identifier associated with the expense information. The working day information or entrance / exit information that is attached, and the working time information that is paired with the working day information that matches the usage date information that the expense information has, or the working end time that the entrance / exit information has, is obtained, and the working time information or It is an expense inspection device that determines whether or not accommodation in an accommodation facility is permitted using the work end time and overtime work time information, and acquires the inspection result.

With such a configuration, it is possible to appropriately detect an application for an illegal accommodation fee.

Further, in the expense inspection device of the eighteenth invention, the expense information is associated with the applicant identifier for the first invention, and includes the usage date information for specifying the date when the expense is used. In addition, it has a store identifier that identifies the store that has spent the expense, and the inspection department inspects two or more expense information corresponding to one applicant identifier, and the number of expense information having the same usage date information is in advance. It is an expense inspection device that determines whether or not there are enough to satisfy the determined conditions and acquires the inspection results.

With such a configuration, it is possible to appropriately detect an application for an illegal second party expense.

Further, the expense inspection device of the nineteenth invention has two or more expenses stored in the expense information storage unit for storing two or more expense information corresponding to the applicant identifier for the first invention. It is further equipped with a statistical processing unit that performs statistical processing based on information and acquires statistical processing results, and the inspection unit inspects whether the statistical processing results satisfy predetermined conditions and acquires inspection results. It is an expense inspection device.

With this configuration, by using the statistical processing results based on two or more expense information, it is possible to appropriately inspect for inappropriate expense application.

Further, the expense inspection device of the twenty-eighth invention has, for the nineteenth invention, the expense information has a store identifier that identifies the store that has spent the expense, and the statistical processing unit is the expense information storage unit. Using the store identifier included in the two or more expense information corresponding to one applicant identifier, which is two or more expense information stored in, store frequency information regarding the frequency of use of the store identified by the store identifier can be obtained. The acquisition and inspection unit is an expense inspection device that inspects whether or not the store frequency information acquired by the statistical processing unit satisfies a predetermined condition and acquires the inspection result.

With such a configuration, it is possible to detect that the specific store is frequently used.

Further, the expense inspection device of the twenty-first invention has, for the nineteenth invention, the expense information has a counterparty identifier that identifies the counterparty who has spent the expense, and the statistical processing unit has the expense. Expenditure to the other party identified by the other party identifier using the other party identifier included in the two or more expense information corresponding to one applicant identifier, which is two or more expense information stored in the information storage unit. The inspection unit is an expense inspection device that acquires the entertainment frequency information related to the frequency of the above, inspects whether the entertainment frequency information acquired by the statistical processing unit satisfies a predetermined condition, and acquires the inspection result.

With such a configuration, it is possible to detect that the frequency of expenditure corresponding to a specific destination is high.

In addition, the expense inspection device of the 22nd invention is incorrect because it is the examination result of the expense information corresponding to one applicant identifier by the statistical processing unit for the 19th invention. The number of frauds, which is the number of times the corresponding examination result is acquired, is acquired in association with one applicant identifier, and the inspection unit further prepares the fraudulent number acquired by the statistical processing unit satisfying a predetermined condition. It is an expense inspection device that inspects whether or not it is present and obtains inspection results.

With such a configuration, it is possible to detect that there is a lot of fraud in a specific person.

In addition, the expense inspection device of the 23rd invention has the approver information that specifies that the company regulation information can be the approver of the expense for the 11th invention, and the expense information is. It has an approver identifier that identifies the approver of the expense, and the inspection department inspects whether the approver identified by the approver identifier is the approver specified by the approver information, and outputs the inspection result. It is an expense inspection device to be acquired.

With such a configuration, it can be detected that the approver for the expense application does not meet the requirements.

Further, in the expense inspection device of the twenty-fourth invention, for the first invention, the applicant information includes the working day information that specifies the working day of the applicant identified by the applicant identifier1. The above attendance information, one or more entry / exit information that specifies the date and time when the applicant entered / exited the place of work identified by the applicant identifier, and business day information that specifies the date of the applicant's business activities. One or more business activity information having one or more point identifiers that specify the visited point, one or more business vehicle position information that specifies the position of the business vehicle used by the applicant, and the date when the business vehicle was used. It has one or more business vehicle position information associated with the specified date of use information, and the inspection department responds to the applicant information by using the applicant information corresponding to one applicant identifier. It is determined whether or not the expense information exists in the expense information storage unit that stores one or more expense information associated with one applicant identifier, and the expense corresponding to the applicant information is applied. It is an expense inspection device that inspects for omissions in applications to see if it has been passed and obtains inspection results.

With such a configuration, it is possible to detect omission of application for expenses.

Further, the expense inspection device of the 25th invention has a receipt image for the expense information for any one of the first to the twenty-fourth inventions, and the inspection department displays the receipt image. An image inspection means for inspecting and acquiring an image inspection result is provided, and a result output unit is an expense inspection device that outputs an image inspection result.

With such a configuration, an appropriate inspection of the receipt image can be performed.

Further, in the expense inspection device of the invention of the present twenty-sixth, for the invention of twenty-five, the image inspection means acquires one or more character strings from the receipt image, and each of the one or more character strings is It is an expense inspection device that determines whether or not a predetermined condition is satisfied and acquires an image inspection result.

With such a configuration, it is possible to appropriately inspect the items described in the receipt image.

Further, the expense inspection device of the 27th invention has a time stamp on the receipt image for the 26th invention, and the image inspection means obtains the time stamp on the receipt image. Then, the receipt issuance date corresponding to the receipt image is acquired, and it is determined whether or not the date specified by the time stamp has passed a predetermined period or more with respect to the receipt issuance date. , It is an expense inspection device that acquires image inspection results.

With such a configuration, it is possible to inspect the receipt image for deficiencies in the time stamp.

Further, the expense inspection device of the twenty-eighth invention has one or more image inspection means from the receipt image received by the expense reception unit by using the character recognition technology for the twenty-fifth invention. The first blurring process, which determines that the receipt image is unclear when the character recognition score is obtained along with the character string and the score is low enough to meet a predetermined condition, or one or more unclear receipt images. The learning device that obtains the judgment result of whether the receipt image is unclear or clear by inputting the receipt image by the machine learning algorithm and the expense reception department accepts one or more receipt images that are clear and clear. It is an expense inspection device that acquires an image inspection result by performing one of the second blurring processes for acquiring a determination result by a machine learning algorithm using the receipt image.

With such a configuration, it is possible to inspect an unclear receipt image.

In addition, in the expense inspection device of the invention of the present 29, for the invention of 25, the external information includes the company regulation information regarding the regulation of the company to which the applicant belongs, and the company regulation information is not allowed. It has the payment method information that specifies the payment method, the image inspection means acquires the payment method information that is a character string from the receipt image, and whether the payment method information is the payment method information that the company regulation information has. It is an expense inspection device that inspects and acquires image inspection results.

With such a configuration, it is possible to inspect inappropriate payment means.

Further, in the expense inspection device of the thirty-third invention, for the twenty-fifth invention, the image inspection means is one or more receipt images whose amount is disguised by hand and one or more which is not disguised. Machine learning with a receipt image using a machine learning algorithm, using a learning device that obtains the judgment result of whether or not the amount has been disguised by inputting the receipt image, and the receipt image received by the expense reception department. It is an expense inspection device that performs prediction processing by the algorithm of the above and acquires an image inspection result indicating whether or not the amount of money described in the receipt image is disguised by hand.

With such a configuration, it is possible to inspect the amount of money by handwriting.

In addition, the expense inspection device of the thirty-first invention has external information or expense information corresponding to the organization identifier that identifies the organization for the first to thirty inventions, and the expense reception unit. Receives the expense information associated with the organization identifier from the expense management device, and the inspection unit is expense information other than the external information associated with the organization identifier or the expense information received by the expense reception unit. , The expense information associated with the organization identifier is used to inspect the expense information received by the expense reception unit for fraud, and the inspection result is acquired. It is an expense inspection device that sends to the expense management device. Here, it is preferable that the expense receiving unit receives the expense information associated with the organization identifier from each of the two or more expense management devices.

With such a configuration, it is possible to provide an expense application inspection function to the expense management device.

According to the expense inspection device according to the present invention, an inappropriate expense application can be appropriately inspected.

Conceptual diagram of the information system A in the first embodiment Block diagram of the information system A Block diagram of the expense inspection device 1 constituting the information system A A flowchart illustrating an operation example of the expense inspection device 1. A flowchart illustrating an example of the inspection process. A flowchart illustrating an example of the statistical inspection process. Flow chart explaining the processing example of the leak inspection Figure showing the same output example Figure showing the same output example Block diagram of other systems including the expense inspection device 1. Overview of the computer system Block diagram of the computer system

Hereinafter, embodiments of the expense inspection device and the like will be described with reference to the drawings. In addition, since the components with the same reference numerals perform the same operation in the embodiment, the description may be omitted again.

(Embodiment 1)
In the present embodiment, an information system including an expense inspection device that receives expense information, inspects fraud using external information or expense information applied for in the past, and outputs inspection results will be described. The external information is information other than expense information, and is, for example, an applicant set described later, a company regulation information described later, a restaurant set described later, a sex shop set described later, a partner set described later, and card use described later. Includes expense information and transportation information described later. In addition, the targets of fraud inspection are, for example, business trip expenses, transportation expenses, double application, inspection of reuse of receipts, accommodation expenses, and expenses of the second party.

Further, in the present embodiment, an information system including an expense inspection device that inspects fraud and outputs the inspection result by using statistical processing results using information corresponding to two or more expense information will be described. The statistical processing result is, for example, the frequency of use of a specific store, the frequency of providing profits to a specific destination, and the number of frauds of a specific person.

Further, in the present embodiment, an information system including an expense inspection device for inspecting whether or not the approver of the expense application meets the requirements will be described.

Further, in the present embodiment, an information system including an expense inspection device for inspecting a receipt image will be described. The inspection of the receipt image is described in, for example, inspection of deficiencies in the items described in the receipt image, inspection of deficiencies in the time stamp of the receipt image, inspection of the receipt image being unclear, and inspection of the receipt image. It is an inspection of improper payment methods and a fraudulent inspection of the handwritten amount of the receipt image.

Further, in the present embodiment, an information system including an expense inspection device for detecting an application omission of an expense to be applied for will be described using external information.

FIG. 1 is a conceptual diagram of the information system A in the present embodiment. The information system A includes an expense inspection device 1, one or more applicant terminals 2, and one or more external servers 3.

The expense inspection device 1 is an apparatus that inspects the applied expense information. The expense inspection device 1 is usually a so-called server, for example, a cloud server, an ASP server, or the like. The type of the expense inspection device 1 does not matter.

The applicant terminal 2 is a terminal used by the applicant. The applicant is a user who applies for expenses. The applicant terminal 2 is, for example, a so-called personal computer, a tablet terminal, a smartphone, or the like, and the type thereof does not matter.

The external server 3 is a device that stores external information described later and provides the external information to the expense inspection device 1. In addition, all or part of the external information described later may exist in the expense inspection device 1. Further, when all of the external information described later exists in the expense inspection device 1, the external server 3 is unnecessary. The external server 3 is usually a so-called server, for example, a cloud server, an ASP server, or the like. The type of the external server 3 does not matter.

The expense inspection device 1, the applicant terminal 2, and the external server 3 can usually communicate with each other via a network such as the Internet or a dedicated line.

FIG. 2 is a block diagram of the information system A according to the present embodiment. FIG. 3 is a block diagram of the expense inspection device 1 constituting the information system A.

The expense inspection device 1 includes a storage unit 11, a reception unit 12, a processing unit 13, and an output unit 14. The storage unit 11 is, for example, an expense information storage unit 111, an applicant set storage unit 112, a company regulation information storage unit 113, a restaurant set storage unit 114, a sex shop set storage unit 115, a destination set storage unit 116, and a card use. It includes an expense storage unit 117 and a transportation means information storage unit 118. The applicant set storage unit 112 includes a personal information storage unit 1121, an attendance information storage unit 1122, an entrance / exit information storage unit 1123, a sales activity information storage unit 1124, a card detail information storage unit 1125, and a business vehicle position information storage unit 1126. To prepare for. The transportation means information storage unit 118 includes a calculation information storage unit 1181 and a map information storage unit 1182. The reception unit 12 includes an expense reception unit 121. The processing unit 13 includes an inspection unit 131 and a statistical processing unit 132. The inspection unit 131 includes image inspection means 1311. The output unit 14 includes a result output unit 141.

The applicant terminal 2 includes a first storage unit 21, a first reception unit 22, a first processing unit 23, a first transmission unit 24, a first reception unit 25, and a first output unit 26.

Expense information storage unit 111, applicant collective storage unit 112, company regulation information storage unit 113, restaurant collective storage unit 114, sex shop collective storage unit 115, destination collective storage unit 116, card usage expense storage in the expense inspection device 1. The unit 117 and the moving means information storage unit 118 may exist in other external devices.

The external server 3 may store any one of the restaurant set storage unit 114, the sex shop set storage unit 115, the other party set storage unit 116, the card usage expense storage unit 117, and the transportation means information storage unit 118. Suitable.

The inspection unit 131 may perform inspection processing described later using various information of the storage unit 11, various information stored in the external server 3, or information transmitted from the external server 3. May be used to perform the inspection process described later.

The external server 3 receives, for example, an inquiry including information for identifying a departure station (an example of a departure identifier described later) and information for identifying an arrival station (an example of an arrival identifier described later), and charges one or more transportation expenses. It is a server that outputs. The external server 3 has a calculation information storage unit 1181 for storing the fare calculation information described later. The external server 3 is a Jordan (registered trademark) server (URL "https://www.jorudan.co.jp/") and a NAVITIME (registered trademark) server (URL "https://www.navitime"). .co.jp/ ").

The external server 3 is, for example, a server that receives an inquiry including a departure identifier and an arrival identifier and outputs a taxi fare. The external server 3 has a calculation information storage unit 1181 and a map information storage unit 1182 for storing taxi fare calculation information described later. Further, the external server 3 is a taxiste server (URL "https://www.taxisite.com/far/").

Various types of information are stored in the storage unit 11. The various types of information are, for example, expense information cases described later and external information described later.

Expense information is stored in the expense information storage unit 111. Usually, two or more expense information is stored in the expense information storage unit 111. Expense information is information that identifies the content of the applied expense. The expense information has, for example, an expenditure amount information, a destination identifier, a store identifier, a destination identifier, a date of use information, a payment instrument identifier, a subject, a description, and an approver identifier. Expenditure amount information is information that identifies the amount used as an expense. The destination identifier is information indicating the destination of the receipt, and is information for identifying the company or the like that pays the amount. The store identifier is information that identifies the store that used the expense. The store identifier is, for example, a store name, a store ID, a store URL, or a store telephone number. The destination identifier is information that identifies the destination using the expense. The other party identifier is information that identifies the other party who has provided benefits such as entertainment or souvenirs. The destination identifier is, for example, an organization name (for example, a company name or an organization name), a name, and a destination ID. The date of use information is information that identifies the date on which the expense was used. The payment instrument identifier is information that identifies the payment instrument. The payment instrument identifier is, for example, "cash", "credit card", "electronic money", "○○ PAY", or the like. A subject is information that identifies the classification of expenses. The subject may be called an expense type identifier. The subjects are, for example, business trip expenses, travel expenses, transportation expenses (or transportation expenses), accommodation expenses, business trip travel expenses, entertainment expenses, and conference expenses. The approver identifier is information that identifies the approver of the expense. The approver identifier is, for example, an applicant identifier or a job title identifier. The applicant identifier is information that identifies the applicant who applied for the expense, and is, for example, a name or an ID. The job title identifier is information that identifies the job title, and is, for example, a job title (for example, "manager" or "director") or ID. The expense information of the expense information storage unit 111 usually corresponds to the applicant identifier.

When the expense information is the expense information of the transportation expense, the expense information has, for example, the transportation expense information having one or more section information having the usage date information, the departure identifier, the arrival identifier, and the charge information. It is preferable that the expense information has an expense type identifier indicating the transportation expense. When the expense information is the expense information of the business trip expense, the expense information has, for example, the date of use information. It is preferable that the expense information has an expense type identifier indicating a business trip expense. When the expense information is the expense information of the accommodation expense, the expense information includes, for example, the usage date information specifying the date of staying at the accommodation facility. When the expense information is entertainment expenses and conference expenses, the expense information is associated with, for example, an applicant identifier, and has a date of use information and a store identifier that identifies the store where the expense was spent.

The expense information may have a receipt image. The receipt image is an image of the receipt. The receipt image is, for example, an image obtained by photographing a receipt. It goes without saying that the data structure of the receipt image, such as JPEG and GIF, does not matter.

The expense information includes, for example, participant information and number of people information. Participant information is information (eg, name) that identifies a person (eg, a person who eats or drinks) who participated in the use of expenses. The number information is the number of people who participated in the use of the expense.

The applicant set is stored in the applicant set storage unit 112. The applicant set is information unique to each applicant corresponding to each applicant. The applicant set is, for example, a set of personal information described later, a set of attendance information described later, a set of entrance / exit information described later, a set of sales activity information described later, a set of card details information described later, or a business described later. It is a set of vehicle position information.

One or more personal information is stored in the personal information storage unit 1121. Personal information is information about the applicant. The personal information is associated with the applicant identifier. Personal information includes, for example, the applicant's name, affiliation company name, affiliation company ID, affiliation department name, affiliation department ID, job title identifier that identifies the job title, regular section information that identifies the applicant's regular section, fraud count, etc. Have an email address. The number of frauds is the number of times the applicant has filed a fraudulent expense application or the number of times the applicant has been detected for fraud. The e-mail address may be any information that specifies the destination to which the information is sent, and may be an SNS ID or the like, and is widely understood.

It should be noted that the correspondence between the two pieces of information means that the other information can be obtained from one piece of information. For example, one piece of information and another piece of information are linked, and one piece of information and another piece of information are linked. Is stored in the same buffer, one piece of information contains another, and so on.

The attendance information storage unit 1122 stores one or more attendance information. Attendance information is information regarding the attendance of the applicant. Attendance information usually corresponds to the applicant identifier. Attendance information includes work day information. The working day information is information for identifying the working day of the applicant identified by the corresponding applicant identifier. The working day information is usually information indicating working days, but may be information indicating non-working days. When the working day information is information indicating a non-working day, the day other than the day indicated by the working day information is a working day. Attendance information usually has working hours information that identifies the hours worked. The working time information has, for example, a working start time and a working end time. The attendance information in the attendance information storage unit 1122 is, for example, information received from one external server 3 (attendance management server). Since the attendance management server (attendance management system) is a server that manages the attendance of employees and is a device related to publicly known technology, detailed description thereof will be omitted.

The entrance / exit information storage unit 1123 stores one or more entrance / exit information. The entry / exit information is information regarding the entry / exit of the applicant. Entry / exit means entry into the applicant's place of work (which may be called the place of work) and exit from the place of work. In addition, it may be said that entering the building means entering the office to start work on the working day, and leaving the building means leaving the office after the end of work on the working day. Entry / exit information usually corresponds to the applicant identifier. The entry / exit information includes, for example, working day information and working time information. The working time information has, for example, a working start time and a working end time. The working hours information has, for example, information for specifying the working hours. Entry / exit information is information composed of time card stamping, entry / exit input to the entry / exit system, and entry / exit information. The entry / exit information may be the same as the attendance information. The entrance / exit information in the entrance / exit information storage unit 1123 is, for example, information received from one external server 3 (entrance / exit management server). Since the entrance / exit management server (entrance / exit management system) is a device related to publicly known technology, detailed description thereof will be omitted.

One or two or more sales activity information is stored in the sales activity information storage unit 1124. Sales activity information is information related to sales activities. The business activity information is, for example, information in a daily business report. Sales activity information usually corresponds to the applicant identifier. The business activity information includes business day information and one or more point identifiers. The business activity information may have a character string or the like indicating the content of the business activity. Business day information is information that identifies the day of the applicant's business activities. The point identifier is information that identifies the visited point. The point identifier is, for example, (latitude, longitude), company name, place name, place name, station name, bus stop name, or the like. The data structure of sales activity information does not matter.

One or more card detail information is stored in the card detail information storage unit 1125. The card details information is information indicating the details of the credit card used by the applicant. The card statement information usually corresponds to the applicant identifier. The card detail information includes, for example, usage date information, a payee identifier that identifies a payee (for example, a store identifier), amount information, and a card identifier. The card identifier is information that identifies the card, and is, for example, a card number. The card detail information in the card detail information storage unit 1125 is, for example, information received from one external server 3 (card management company server).

The business vehicle position information storage unit 1126 stores one or more business vehicle position information. The business vehicle position information is information that identifies the position of the business vehicle used by the applicant. The business vehicle location information usually corresponds to the applicant identifier. The business vehicle position information is associated with the usage date information that identifies the date when the business vehicle was used. The business vehicle position information is, for example, (latitude, longitude), but may be a point identifier. The business vehicle position information corresponds to, for example, a business vehicle identifier that identifies the business vehicle.

The company regulation information storage unit 113 stores the company regulation information. Company regulation information is information indicating company regulations. Company regulation information is information regarding the regulations of the company to which the applicant belongs. The company regulation information includes, for example, overtime time information, one or more approver information, and one or more payment means information.

The overtime time information is information that specifies the overtime time that allows accommodation. The overtime time information is, for example, "23:00" and "after 24:00".

Approver information is information that identifies a potential approver of an expense. The approver information is, for example, an approver identifier (for example, an applicant identifier, a name) that identifies an approver, and a job title identifier (for example, a “manager” or “director”) that identifies a job title of an approver.

The payment method information is information that identifies an unacceptable payment method. The payment means information is, for example, "○○ PAY", "electronic money", and "credit card".

The restaurant set is stored in the restaurant set storage unit 114. The restaurant set has one or more restaurant information. The restaurant information is information about the restaurant. The restaurant information has, for example, a restaurant identifier that identifies the restaurant and price information. The restaurant information may include address information, telephone number information, and the like. The restaurant identifier is, for example, a restaurant name or an ID. The amount information is, for example, information for specifying an appropriate usage amount at a restaurant. The amount information is, for example, information on the amount of money used by one person at the restaurant. The amount information is, for example, information indicating an average amount of money used by one person at the restaurant. The amount information may be, for example, information that specifies an upper limit of the amount of money used by one person at the restaurant. The amount information may be, for example, information that specifies an upper limit of the amount of money to be used at one time.

The sex shop set storage unit 115 stores the sex shop set. The sex shop set has one or more sex shop information. The sex shop information has, for example, a sex shop identifier that identifies the sex shop. The sex shop information may include amount information, address information, telephone number information, and the like. The sex shop information is, for example, information that identifies a store that the applicant should not go to.

The destination set storage unit 116 stores the destination set. The destination set has one or more destination information. The other party information has, for example, a other party identifier and an affiliation identifier that identifies the affiliation of the other party. The other party information is, for example, information of a public employee. The other party information is, for example, the information of a member of the Diet. The other party information is, for example, information that identifies the other party to which the applicant should not provide profit. Profit provision is, for example, the act of giving entertainment or souvenirs.

The card usage expense storage unit 117 stores one or more expense type identifiers scheduled to be paid by the card. The expense type identifier may correspond to the card type identifier. The card type identifier is information that identifies the type of card. The card type identifier is, for example, "EX-IC card" or "corporate card". The expense type identifier is, for example, "Shinkansen" or "entertainment expense" indicating the fare of the Shinkansen.

The transportation means information storage unit 118 stores the transportation means information. The transportation means information is information for calculating the transportation cost when a transportation means such as a train, a bus, an airplane, or a taxi is used. The transportation means information includes the calculation information described later. The transportation information includes, for example, map information described later.

The calculation information is stored in the calculation information storage unit 1181. The calculation information is, for example, fare calculation information and taxi fare calculation information. The fare calculation information is, for example, information for calculating the fare of a means of transportation such as a train, a bus, or an airplane. The fare calculation information is, for example, information having one or more sets of a departure identifier, an arrival identifier, and a fare information. The departure identifier and the arrival identifier are, for example, a station name, an airport name, a bus stop name, a station ID, an airport ID, and a bus stop ID. The taxi fare calculation information is information for calculating the taxi fare. The taxi fare calculation information is, for example, an arithmetic expression for calculating a taxi fare by inputting a distance. The taxi fare calculation information is, for example, information having one or more pairs of distance and taxi fare.

Map information is stored in the map information storage unit 1182. The map information is, for example, map information used for navigation. The map information is, for example, KIWI format information.

The reception unit 12 receives various information and instructions. The reception here is usually reception from a terminal such as the applicant terminal 2. However, even if the concept includes acceptance of information input from an input device such as a keyboard, mouse, or touch panel, and acceptance of information read from a recording medium such as an optical disk, a magnetic disk, or a semiconductor memory. good.

The expense reception unit 121 accepts one or more expense information. The expense reception unit 121 usually receives expense information in association with the applicant identifier. The expense reception unit 121 receives, for example, expense information and an applicant identifier. The expense reception unit 121 may, for example, receive one or more expense information for each one or more applicants. The expense information is received, for example, from the applicant terminal, but may be received from an administrator terminal (not shown), read from a recording medium, or may be received from the applicant's input.

It is preferable that the expense receiving unit 121 receives the expense information having the receipt image. The expense reception unit 121 may separately accept the receipt image and the information other than the receipt image and constituting the expense information. However, the receipt image and the information other than the receipt image are accepted so as to correspond.

The processing unit 13 performs various processes. The various processes are, for example, processes performed by the inspection unit 131 and the statistical processing unit 132.

The inspection unit 131 inspects the expense information received by the expense reception unit 121 regarding fraud and acquires the inspection result. The inspection unit 131 may inspect the expense information stored in the expense information storage unit 111 regarding fraud and acquire the inspection result. The expense information stored in the expense information storage unit 111 is the expense information received by the expense reception unit 121. The inspection result is information related to the inspection result of the expense information. The test result includes, for example, information indicating whether or not the test result is fraudulent. The inspection result includes, for example, a part or all of the expense information that is fraudulent. The inspection result is, for example, information indicating that it is not fraudulent (valid). The inspection result has, for example, one or more information among the expense information identifier, the applicant identifier, the expense type identifier, the amount information, the inspection content identifier, the application date, and the like. The expense information identifier is information that identifies expense information, and is, for example, an ID. The inspection content identifier is information that identifies the inspection content. The inspection content identifier is usually stored in the storage unit 11 in association with the inspection process described later. The inspection content identifiers are, for example, "expenses other than work days", "illegal travel expenses", "illegal accommodation expenses", "accommodation expenses on non-accommodation days", "illegal transportation expenses", and "receipt division". The content of the inspection content identifier and the data structure are not limited.

The inspection unit 131, for example, uses external information to inspect the expense information received by the expense reception unit 121 regarding fraud, and acquires the inspection result. External information is information other than expense information. The external information used by the inspection unit 131 may be stored in the storage unit 11 of the expense inspection device 1 or may be stored in the external server 3.

For example, the inspection unit 131 inspects the expense information received by the expense reception unit 121 regarding fraud using the expense information other than the expense information received by the expense reception unit 121, and acquires the inspection result. The expense information other than the expense information received by the expense reception unit 121 is, for example, the expense information stored in the expense information storage unit 111, but may be the expense information stored in the external server 3.

For example, the inspection unit 131 acquires the applicant information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121 from the applicant set, and uses the applicant information to obtain the expense reception unit. The inspection regarding the fraud of the expense information received by 121 is performed, and the inspection result is acquired. The applicant set is, for example, information stored in the applicant set storage unit 112, but may be information stored in an external device (not shown). The applicant information is information corresponding to the applicant identifier, and is, for example, attendance information, entrance / exit information, sales activity information, card details information, and business vehicle location information.

The image inspection means 1311 inspects the receipt image of the expense information received by the expense reception unit 121, and acquires the image inspection result. The image inspection result is an example of the inspection result.

The image inspection means 1311 acquires one or more character strings from the receipt image, inspects whether the receipt image is valid or not using the one or more character strings, and acquires an image inspection result. The image inspection means 1311 uses any one or more of the character recognition process and the machine learning process in order to acquire one or more character strings from the receipt image.

In this specification, the algorithm used in the machine learning process does not matter. Machine learning is, for example, deep learning, decision tree, random forest, SVM, SVR and the like. Modules used in machine learning are, for example, TensorlFlow modules, TinySVM, and the like.
(1) Method by character recognition processing

The image inspection means 1311 performs character recognition processing on the receipt image, for example, and acquires one or more character strings. A process of performing character recognition processing and acquiring information constituting expense information from a receipt image is a known technique. The image inspection means 1311 determines which of the payment amount information, the store identifier, the other party's identifier, the date of use information, the subject, the description, and the payment means information is the acquired character string. Acquires a character string in association with information that identifies the type of information (which may be called a class, category, etc.). The technique for determining the class of the acquired character string is an information classification problem (clustering problem), which is a known technique.

The image inspection means 1311 performs machine learning learning processing, for example, by using two or more character strings belonging to each class as a positive example and one or more character strings not belonging to the class as a negative example for each class. , Construct a learner that performs binary classification as to whether or not it belongs to a class. Then, the image inspection means 1311 performs machine learning prediction processing on each of one or more character strings acquired from the receipt image by using two or more learning devices, and obtains prediction results and scores for each class. Acquire and determine the class with the highest score belonging to the class as the class to which it belongs.

Further, the image inspection means 1311 constitutes, for example, one learning device that performs learning processing as teacher data having two or more sets of character strings and classes, and performs multi-value classification to output the classes. Next, the image inspection means 1311 performs machine learning prediction processing on each of one or more character strings acquired from the receipt image by using a learning device, and determines the class to which each character string belongs.

The image inspection means 1311 uses, for example, a learning device for classifying a character string using a large number of teacher data which is a set of a character string and a class, and a module that performs learning processing among machine learning algorithms. And get it.
(2) Method by machine learning processing

The image inspection means 1311 uses, for example, two or more teacher data having one or more sets of a receipt image and a set of a character string and a class, and performs learning processing by a machine learning algorithm to acquire a learner. .. It should be noted that this learning device is a learning device that inputs a receipt image and outputs one or more sets of a character string and a class.

Next, the image inspection means 1311 gives the acquired learning device and the receipt image of the expense information received by the expense receiving unit 121 to the module of the machine learning algorithm that performs the prediction processing, and sets it as a character string. Acquire one or more pairs with the class.
(3) Method using both character recognition processing and machine learning processing

The image inspection means 1311 obtains a character string for each class by using both of the above (1) and (2), and if they do not match in the processes of (1) and (2) for each class, the character recognition process is performed. Using the score acquired in step 1 and the score acquired by machine learning processing, the higher character string is acquired so that the score satisfies the condition.

Further, using one or more character strings acquired by any of the above methods (1) to (3), the image inspection means 1311 determines whether each of the one or more character strings satisfies a predetermined condition. Judge whether or not, and acquire the image inspection result. The character string of 1 or more is, for example, any one of payment amount information, store identifier, destination identifier, date of use information, subject, description, and payment means information.

The image inspection means 1311 uses machine learning using teacher data having, for example, a negative example of an invalid one or more receipt images and a positive example of two or more valid receipt images. The learning process is performed by the algorithm of, and a learning device is constructed. Next, the image inspection means 1311 gives the receipt image of the expense information received by the expense reception unit 121 and the learning device to the module that performs the prediction processing of machine learning, performs the prediction processing, and is it illegal. Get the prediction result of whether or not. The learning device is, for example, a learning device for detecting fraudulent processing of the amount of money in the receipt, a learning device for inspecting the presence or absence of a signature on the receipt, and a learning device for detecting an unclear receipt image. It is a learning device. In addition, the learner is usually a learner for binary classification of whether or not it is illegal.

The statistical processing unit 132 is 1 or 2 or more expense information corresponding to the applicant identifier corresponding to the received expense information, and is based on 1 or 2 or more expense information stored in the expense information storage unit 111. Perform statistical processing and acquire statistical processing results. The statistical processing unit 132 is, for example, one or two or more expense information corresponding to the applicant identifier for each applicant identifier, and is based on one or two or more expense information stored in the expense information storage unit. Perform statistical processing and acquire statistical processing results. The timing at which the statistical processing unit 132 performs statistical processing does not matter. The statistical processing unit 132 performs statistical processing, for example, by accepting a user's instruction. The statistical processing unit 132 performs statistical processing, for example, by receiving expense information.

The statistical processing unit 132 is, for example, two or more expense information stored in the expense information storage unit 111, and uses the store identifier included in the two or more expense information corresponding to one applicant identifier to store frequency. Get information. The store frequency information is information regarding the usage frequency of the store identified by one store identifier. The store frequency information is information indicating the frequency of use. The store frequency information is, for example, information indicating the total number of times of use and information indicating the number of times of use in a unit period. The unit period is, for example, 1 month, 6 months, and 1 year. It is preferable for the statistical processing unit 132 to acquire store frequency information for each applicant identifier.

The statistical processing unit 132 is entertainment frequency information using two or more expense information stored in the expense information storage unit 111, and using the destination identifier included in the two or more expense information corresponding to one applicant identifier. To get. The entertainment frequency information is information on the frequency of spending on the other party identified by one of the other party identifiers. Expenditure to the other party is payment for providing profit to the other party, for example, payment of entertainment to the other party, payment of souvenirs to the other party, and the like. Entertainment frequency information is information indicating the frequency of profit provision. The entertainment frequency information is, for example, information indicating the total number of times profits have been provided and information indicating the number of times profits have been provided in a unit period. It is preferable for the statistical processing unit 132 to acquire entertainment frequency information for each applicant identifier.

The statistical processing unit 132 associates the number of frauds, which is the examination result for the expense information corresponding to one applicant identifier and the number of times the examination result corresponding to the fraud is obtained, with the one applicant identifier. get. It is preferable for the statistical processing unit 132 to acquire the number of frauds for each applicant identifier.

The result output unit 141 outputs the inspection result acquired by the inspection unit 131. The result output unit 141 outputs the image inspection result acquired by the image inspection means 1311.

The result output unit 141 outputs the statistical processing result acquired by the statistical processing unit 132. The result output unit 141 outputs the statistical processing result in association with the applicant identifier, for example.

Various types of information are stored in the first storage unit 21 constituting the applicant terminal 2. Various types of information are, for example, an applicant identifier and expense information.

The first reception unit 22 receives various instructions and information. Various instructions and information are, for example, expense information and answer information. The answer information is information indicating the answer to the test result.

The first processing unit 23 performs various processing. The various processes are, for example, a process of converting the instructions and information received by the first reception unit 22 into instructions and information of a data structure to be transmitted. The various processes are, for example, a process of converting the information received by the first receiving unit 25 into the information of the data structure to be output.

The first transmission unit 24 transmits various instructions and information. Various instructions and information are, for example, expense information, applicant identifiers, and response information.

The first receiving unit 25 receives various types of information. Various types of information are, for example, inspection results and image inspection results.

The first output unit 26 outputs various information. Various types of information are, for example, inspection results and image inspection results.

Here, the output means display on a display, projection using a projector, printing by a printer, sound output, transmission to an external device, storage in a recording medium, storage in another processing device, another program, or the like. It is a concept that includes delivery of processing results.
(Specific example of inspection process)

Hereinafter, an example of a specific inspection performed by the inspection unit 131 will be described below. In the following, the case of fraud will be mainly described, but if it is not fraudulent, the inspection unit 131 usually acquires the inspection result to the effect that it is “legitimate”. Further, if it is legitimate, the inspection unit 131 usually stores the expense information to be inspected in the expense information storage unit 111 in association with the applicant identifier.

The information contained in the expense information used by the inspection unit 131 for inspection may be a character string received by the expense reception unit 121 or information acquired from the receipt image received by the expense reception unit 121. The character string received by the expense reception unit 121 is, for example, information input by the applicant.
(1) Inspection using attendance information

The inspection unit 131 is, for example, attendance information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and is the same as the usage date information included in the expense information received by the expense reception unit 121. The attendance information including the work day information of is acquired from the attendance information storage unit 1122. Next, the inspection unit 131 uses the attendance information to inspect the expense information received by the expense reception unit 121 regarding fraud, and acquires the inspection result.
(1-1) Inspection of expenses other than work days

The inspection unit 131 is, for example, attendance information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and is the same as the usage date information included in the expense information received by the expense reception unit 121. The attendance information including the work day information of is acquired from the attendance information storage unit 1122. Next, the inspection unit 131 determines whether or not the attendance information is information on a day other than the work day. In the case of information on a day other than the work day, the inspection unit 131 acquires an inspection result indicating that the information is “illegal expense not on the work day”. In addition, if the attendance information is information on a day other than the work day, the inspection department 131 indicates that the attendance information includes "information indicating that it is a holiday" or that the attendance information does not include the work start time and the work end time. Judge from the facts.
(1-2) Inspection of fraudulent business trip expenses

The inspection unit 131 acquires the applicant identifier and the date of use information associated with the expense information received by the expense reception unit 121. Next, the inspection unit 131 acquires the attendance information corresponding to the acquired applicant identifier and the date of use information from the attendance information storage unit 1122. Next, when the inspection unit 131 determines that the acquired attendance information is not information indicating a business trip, the inspection unit 131 acquires an inspection result indicating "illegal business trip". If "attendance information is not information indicating a business trip", for example, (a) if there is no information indicating "business trip" in the attendance information, (b) start work in the attendance information. For example, when information on the time and end time of work exists.
(1-3) Inspection of accommodation expenses for fraud

The inspection unit 131 acquires the applicant identifier and the date of use information associated with the expense information received by the expense reception unit 121. Next, the inspection unit 131 acquires working time information from the acquired applicant identifier and working day information corresponding to the usage date information. Next, the inspection unit 131 uses the acquired working time information to determine whether or not accommodation at the accommodation facility is permitted, and acquires the inspection result.

More specifically, the inspection unit 131 compares the work end time of the working time information with the overtime time information indicating the overtime time at which accommodation is permitted, and determines whether or not the predetermined accommodation condition is satisfied. do. If the inspection unit 131 determines that the accommodation conditions are not satisfied, the inspection unit 131 acquires the inspection result of "the fact that the accommodation fee is illegally charged". The accommodation condition is a condition using the overtime time information and the work end time, for example, "overtime time information = <work end time" and "overtime time information <work end time".
(2) Inspection using entrance / exit information

The inspection unit 131 is entry / exit information corresponding to the applicant identifier associated with the expense information, and stores the entry / exit information corresponding to the date specified by the usage date information possessed by the expense information. Obtained from the department 1123, and using the entrance / exit information, the expense reception unit 121 inspects the fraudulent expense information and acquires the inspection result.
(2-1) Inspection of expenses other than work days

The inspection unit 131 acquires the applicant identifier and the date of use information associated with the expense information received by the expense reception unit 121. Next, the inspection unit 131 performs a process of acquiring the entrance / exit information corresponding to the acquired applicant identifier and usage date information from the entrance / exit information storage unit 1123. Next, the inspection unit 131 determines whether or not the entry / exit information is information on a day other than the work day. In the case of information on a day other than the work day, the inspection unit 131 acquires an inspection result indicating that the information is “illegal expense not on the work day”. In addition, if the entrance / exit information is information on a day other than the work day, the inspection department 131 indicates that the entrance / exit information includes "information indicating that it is a holiday", or the entrance / exit information includes the work start time and the work start time. Judgment is made based on the fact that the work end time is not included.
(2-2) Inspection of fraudulent business trip expenses

The inspection unit 131 acquires the applicant identifier and the date of use information associated with the expense information received by the expense reception unit 121. Next, the inspection unit 131 performs a process of acquiring the entrance / exit information corresponding to the acquired applicant identifier and usage date information from the entrance / exit information storage unit 1123. Next, when the inspection unit 131 determines that the result of the entry / exit information acquisition process is not information indicating a business trip, the inspection unit 131 acquires an inspection result indicating "illegal business trip". If "the result of the entrance / exit information acquisition process is not information indicating a business trip", for example, (a) if the entrance / exit information can be acquired, (b) the work that the acquired entrance / exit information has. This is the case where the information on the start time and the work end time is the time information on the start of work and the time information on the end of work in normal work. In addition, the time information of the start of work and the time information of the end of work of normal work are stored in, for example, company regulation information.
(2-3) Inspection of accommodation expenses for fraud

The inspection unit 131 acquires the applicant identifier and the date of use information associated with the expense information received by the expense reception unit 121. Next, the inspection unit 131 acquires the work end time from the entrance / exit information corresponding to the acquired applicant identifier and usage date information. Next, the inspection unit 131 determines whether or not accommodation at the accommodation facility is permitted using the acquired work end time, and acquires the inspection result.

More specifically, the inspection unit 131 compares the work end time with the overtime time information indicating the overtime time at which accommodation is permitted, and determines whether or not the predetermined accommodation condition is satisfied. If the inspection unit 131 determines that the accommodation conditions are not satisfied, the inspection unit 131 acquires the inspection result of "the fact that the accommodation fee is illegally charged". The overtime time information is stored in, for example, company regulation information.
(3) Inspection using sales activity information

The inspection unit 131 is sales activity information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and has the same business day information as the usage date information possessed by the expense information. Acquires one or more point identifiers possessed by the company from the business activity information storage unit 1124, and uses the one or more point identifiers to inspect the expense information received by the expense reception unit 121 regarding fraud and acquire the inspection result.
(3-1) Inspection of fraudulent transportation expenses (i)

The inspection unit 131 gives information indicating the location of the company and one or more point identifiers possessed by the business activity information to the transportation cost calculation module, and uses the calculation information for calculating the transportation cost to obtain one or two or more points. Get transportation expenses. It is preferable that the transportation cost acquired here is the lowest transportation cost. In addition, when acquiring two or more transportation expenses, it is a case where there are two or more travel routes. Further, since the technology for acquiring the transportation cost and the technology for acquiring the lowest transportation cost are publicly known technologies, detailed description thereof will be omitted. Further, the inspection unit 131 acquires the expenditure amount information from the expense information received by the expense reception unit 121.

Next, the inspection unit 131 determines whether or not the expenditure amount information satisfies a predetermined legitimate transportation cost condition by using the expenditure amount information and the acquired transportation cost. If the conditions for legitimate transportation expenses are not met, an inspection result stating that the transportation expenses claim is incorrect is obtained. The conditions for legitimate transportation expenses are, for example, "expenditure amount information = lowest transportation cost", "expenditure amount information matches any one or more transportation expenses", and "expenditure amount information <= 1 or more". One of the transportation expenses. "
(3-2) Inspection of fraudulent transportation expenses (ii)

The inspection unit 131 acquires transportation expense information having a departure identifier and an arrival identifier from the expense information received by the expense reception unit 121. In addition, the inspection unit 131 acquires one or more point identifiers possessed by the business activity information. Then, the inspection unit 131 determines whether or not the distance between the departure identifier and any of the point identifiers is close enough to satisfy a predetermined condition. Then, when the distance between the departure identifier and all the point identifiers does not satisfy a predetermined condition, the inspection unit 131 acquires an inspection result indicating that "the application for transportation expenses is invalid". The inspection unit 131 determines whether or not the distance between the arrival identifier and any of the point identifiers is close enough to satisfy a predetermined condition. Then, when the distance between the arrival identifier and all the point identifiers does not satisfy a predetermined condition, the inspection unit 131 acquires an inspection result indicating that "the application for transportation expenses is invalid".

The technique for calculating the distance between the departure identifier or the arrival identifier and each point identifier is a known technique.
(4) Inspection using card details information

The inspection unit 131 is card detail information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and has the same usage date information as the usage date information of the expense information. Is obtained from the card detail information storage unit 1125, and the expense information received by the expense reception unit 121 is inspected for fraud and the inspection result is acquired.
(4-1) Inspection of receipt division

The inspection unit 131 is card detail information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and has the same usage date information as the usage date information of the expense information. Is obtained from the card details information storage unit 1125. Then, when the inspection unit 131 can acquire two or more card details information and determines that the store identifiers of each of the two or more card details information are the same, it means that "receipt division is illegal". Get the test result of.
(4-2) Inspection of taxi use after in-house drinking party

The inspection unit 131 acquires the card detail information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121. In addition, the inspection unit 131 determines whether or not it is an expense for using a taxi by using the information possessed by the card detailed information. In the case of taxi usage expenses, the inspection unit 131 inputs one or more participant information (for example, the name of the participant) included in the expense information including the same usage date information as the usage date information of the card details information. get. Next, the inspection unit 131 determines whether or not all of the participant information of one or more corresponds to the personal information possessed by the personal information storage unit 1121 (employee database) (for example, whether or not the name of the participant is included). To judge. Then, when all of the participant information of one or more corresponds to the personal information, the inspection unit 131 acquires the inspection result indicating that "the taxi is used after the in-house drinking party".
(5) Inspection using business vehicle location information

The inspection unit 131 is one or more business vehicle position information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121, and has the same usage date information as the usage date information possessed by the expense information. The one or more associated business vehicle position information is acquired from the business vehicle position information storage unit 1126. Next, the inspection unit 131 inspects the fraudulent expense information using the business vehicle position information and acquires the inspection result.
(5-1) Inspection of illegal use of business vehicles (inspection of illegal points)

The inspection unit 131 determines whether or not the acquired position information of one or more business vehicles is close enough to meet a predetermined condition with a predetermined illegal point (for example, a tourist spot). If it is determined that the condition is close enough to satisfy a predetermined condition, the inspection unit 131 acquires an inspection result indicating that the vehicle is illegally used. In addition, the information of the predetermined illegal point is included in the company regulation information, for example.
(6) Inspection using card usage expense information

In the inspection unit 131, the expense type identifier corresponding to the expense information received by the expense reception unit 121 corresponds to the possession card information corresponding to the applicant identifier associated with the expense information received by the expense reception unit 121. As a type identifier, it is inspected whether or not it is stored in the card usage expense storage unit 117, and the inspection result is acquired. For example, it is assumed that the use of a credit card is stored in the card usage expense storage unit 117 for food and drink charges or entertainment expenses.
(7) Inspection using company regulation information

The inspection unit 131 uses the company regulation information stored in the company regulation information storage unit 113 to determine whether the expense information received by the expense reception unit 121 conforms to the company regulation specified by the company regulation information. Perform inspections and obtain inspection results. Specific examples of the inspection using the company regulation information are, for example, (2-2), (2-3), (13-1) (13-2) described later, and (14-3) (a) described later. ), (14-4) (a) and the like described later.
(8) Inspection using transportation information

The inspection unit 131 inspects the fraudulent transportation expenses by using the expense information received by the expense reception unit 121 and the transportation means information stored in the transportation means information storage unit 118, and acquires the inspection result.
(8-1) Inspection of fraudulent transportation costs for trains, buses, or planes

The inspection unit 131 acquires the applicant information corresponding to the applicant identifier associated with the expense information related to the transportation expenses received by the expense reception unit 121, and uses the applicant information or the applicant information and the transportation means information. Inspect for fraud related to transportation expense claims and obtain inspection results.

More specifically, the inspection unit 131 acquires transportation expense information having one or more section information having a departure identifier, an arrival identifier, and a fare information from the expense information received by the expense reception unit 121. Next, the departure identifier and the arrival identifier of each of the one or more section information are given to the transportation cost calculation module, and one or two or more transportation costs are acquired by using the calculation information for calculating the transportation cost. It is preferable that the transportation cost acquired here is the lowest transportation cost. In addition, when acquiring two or more transportation expenses, it is a case where there are two or more travel routes. Further, since the technology for acquiring the transportation cost and the technology for acquiring the lowest transportation cost are publicly known technologies, detailed description thereof will be omitted. Further, it is preferable to acquire the transportation cost by using the external server 3.

Further, the inspection unit 131 acquires the expenditure amount information from the expense information received by the expense reception unit 121.

Next, the inspection unit 131 determines whether or not the fare information satisfies a predetermined legitimate transportation cost condition by using the fare information and the transportation cost. If the conditions for legitimate transportation expenses are not met, an inspection result stating that the transportation expenses claim is incorrect is obtained. The conditions for legitimate transportation costs are, for example, "fee information = lowest transportation cost", "fee information matches any one or more transportation costs", and "fee information <= 1 or more". Transportation expenses ".
(8-2) Inspection of taxi fare fraud

The inspection unit 131 acquires the applicant information corresponding to the applicant identifier associated with the expense information related to the taxi fare received by the expense reception unit 121, and uses the applicant information or the applicant information and the transportation means information. Inspect for fraud related to transportation expense billing and obtain inspection results.

More specifically, the inspection unit 131 acquires transportation expense information having a departure identifier, an arrival identifier, and an expenditure amount information from the expense information received by the expense reception unit 121. Next, the distance traveled is acquired using the departure identifier, arrival identifier, and map information possessed by the transportation cost information. Next, the inspection unit 131 acquires the taxi fare by using the distance and the calculation information for calculating the taxi fare. Here, since the taxi fare acquisition technique is a publicly known technique, detailed description thereof will be omitted. Further, it is preferable to obtain the taxi fare using the external server 3.

Further, the inspection unit 131 acquires the expenditure amount information from the expense information received by the expense reception unit 121.

Next, the inspection unit 131 uses the expenditure amount information and the taxi fare to determine whether or not the expenditure amount information satisfies a predetermined legitimate taxi fare condition. If you do not meet the conditions for a valid taxi fare, you will get an inspection result that says "the transportation bill is incorrect". The conditions for legitimate transportation expenses are, for example, "expenditure amount information = taxi fare", "expenditure amount information-taxi fare <= threshold value", and "expenditure amount information-taxi fare <threshold value".
(8-3) Inspection of overlap with regular sections

The inspection unit 131 acquires the periodic section information corresponding to the applicant identifier associated with the expense information related to the transportation expenses received by the expense reception unit 121 from the personal information storage unit 1121. Next, the inspection unit 131 acquires transportation expense information having one or more section information having a departure identifier and an arrival identifier from the expense information received by the expense reception unit 121.

Next, the inspection unit 131 determines whether or not there is an overlap between the section specified by the periodic section information and the section specified by one or more section information, using the transportation means information. The inspection unit 131 uses, for example, a set of station names (station name group 1) in which station names in one or more sections of each section information are arranged in order, using transportation means information. Then, the inspection unit 131 acquires a set of station names (station name group 2) in which the station names of the sections specified by the regular section information are arranged in order, using the transportation means information. Then, the inspection unit 131 determines whether or not two or more consecutive same station names are included in the station name group 1 and the station name group 2. When two or more consecutive same station names are included, the inspection unit 131 acquires an inspection result indicating that the applied transportation cost includes a regular section.
(9) Inspection using store information

The inspection unit 131 inspects for fraudulent expense information using the store identifier of the expense information received by the expense reception unit 121 and the store information of the store set, and acquires the inspection result. The store set is, for example, a restaurant set of the restaurant set storage unit 114 or a restaurant set of the sex shop set storage unit 115, but may be a restaurant set or a sex shop set existing in an external device (not shown). good.
(9-1) Inspection of store fraud (use of sex shops, etc.)

The inspection unit 131 acquires, for example, the store identifier of the expense information received by the expense reception unit 121. Next, the inspection unit 131 inspects whether or not the acquired store identifier (sex shop identifier) is stored in the sex shop collective storage unit 115, and if it is stored, "at a sex shop that should not be visited". Obtain the inspection result indicating "injustice that the expense was used". In addition, it is preferable that the inspection result includes the sex shop information including the store identifier of the sex shop collective storage unit 115.
(9-2) Inspection of amount fraud (a) Inspection of the entire amount

The inspection unit 131 acquires, for example, the amount information paired with the store identifier of the expense information received by the expense reception unit 121 from the restaurant set storage unit 114. Next, the inspection unit 131 acquires, for example, the expenditure amount information contained in the expense information received by the expense reception unit 121. Next, the inspection unit 131 compares the acquired expenditure amount information with the acquired amount information, and determines whether or not the information indicates an amount high enough to satisfy a predetermined condition. Then, the inspection unit 131 acquires an inspection result indicating that "an illegal amount of expenses has been used" when the information indicates an amount high enough to satisfy a predetermined condition. The predetermined conditions are, for example, "expenditure amount information-amount information> = threshold value", "expenditure amount information-amount information> threshold value", "expenditure amount information / amount information> = threshold value", and "expenditure amount information / amount information". > Threshold ". Needless to say, each threshold value may be different.
(9-3) Inspection of whether or not you are using a restaurant for entertainment expenses

The inspection unit 131 acquires, for example, the expense type identifier of the expense information received by the expense reception unit 121. Next, the inspection unit 131 determines whether or not the expense type identifier corresponds to "entertainment expenses". In the case of "entertainment expenses", the inspection unit 131 acquires the store identifier of the expense information received by the expense reception unit 121. Next, the inspection unit 131 determines whether or not the store identifier is a restaurant identifier. If it is not the identifier of the restaurant, the inspection unit 131 acquires the inspection result indicating that the application is for illegal entertainment expenses.

The inspection unit 131 determines whether or not the store identifier is a restaurant identifier, for example, depending on whether or not the store identifier is stored in the restaurant set storage unit 114. Further, the inspection unit 131 performs a web search using, for example, a store identifier as a key, and depending on whether or not the acquired web page contains information indicating a restaurant (for example, a pub, Italian, a cooking name, etc.), the store Determine if the identifier is a restaurant identifier. Further, other methods may be used for clustering of store identifiers. That is, the algorithm for determining whether or not the store identifier is the identifier of the restaurant does not matter.
(B) Inspection of the amount of money per person

The inspection unit 131 acquires, for example, the amount information paired with the store identifier of the expense information received by the expense reception unit 121 from the restaurant set storage unit 114. Next, the inspection unit 131 acquires, for example, the expenditure amount information contained in the expense information received by the expense reception unit 121. Further, the inspection unit 131 acquires, for example, the number of people information possessed by the expense information received by the expense reception unit 121, or the number of people information acquired from the participant information possessed by the expense information. Next, the inspection unit 131 determines whether or not the amount of money per person is high enough to satisfy a predetermined condition by using the acquired expenditure amount information, the acquired amount information, and the number of people information. Then, when the amount of money per person is high enough to satisfy a predetermined condition, the inspection unit 131 acquires an inspection result indicating that "an illegal amount of expenses has been used". The predetermined conditions are, for example, "(expenditure amount information / number of people information) -amount information> = threshold value""(expenditure amount information / number of people information) -amount information> threshold value""(expenditure amount information / number of people information)". ) / Amount information> = threshold value "" (expenditure amount information / number of people information) / amount information> threshold value ". Needless to say, each threshold value may be different.
(10) Inspection using the other party's information

The inspection unit 131 inspects the fraudulent expense information using the counterparty identifier of the expense information received by the expense reception unit 121 and one or more counterparty information possessed by the external information, and acquires the inspection result. The other party information of one or more is, for example, a destination set. Further, the counterparty set is, for example, the counterparty set of the counterparty set storage unit 116.
(10-1) Inspection of profit provision to public employees

The inspection unit 131 acquires the counterparty identifier of the expense information received by the expense reception unit 121. Next, the inspection unit 131 determines whether or not the acquired destination identifier exists in the destination set storage unit 116 in which the information of the public employee is stored. If it exists, obtain an inspection result indicating "illegal provision of unfair profits to public officials".
(10-2) Inspection of profit provision to members of the Diet

The inspection unit 131 acquires the counterparty identifier of the expense information received by the expense reception unit 121. Next, the inspection unit 131 determines whether or not the acquired counterparty identifier exists in the counterparty set storage unit 116 in which the information of the members of the Diet is stored. If it exists, obtain an inspection result indicating "injustice to the effect that an illegal profit was provided to a member of the Diet".
(11) Inspection using past expense information (11-1) Inspection of double application

The inspection unit 131 is associated with the expense information received by the expense reception unit 121 and the same applicant identifier as the applicant identifier associated with the expense information, and is stored in the expense information storage unit 111 in the past. It is determined whether or not the expense information of one or more of the above is close enough to satisfy a predetermined condition, and the inspection result is acquired.

The fact that the two expense information "approximate enough to satisfy a predetermined condition" means, for example, that the expense information received by the expense reception unit 121 has the expense amount information, the store identifier, the partner identifier, the date of use information, the subject, and the like. Of the description and the approver identifier, (a) a predetermined number or more of information matches, (b) a predetermined 1 or 2 or more pieces of information all match, and the like. The predetermined 1 or 2 or more information is one or more of the expenditure amount information, the store identifier, and the destination identifier of the expense information received by the expense reception unit 121.
(11-2) Inspection of reuse of receipts

In the inspection unit 131, the receipt image of the expense information received by the expense reception unit 121 and the receipt image of the past expense information stored in the expense information storage unit 111 satisfy a predetermined condition. If it is judged that they are similar to each other, the inspection result indicating that "the receipt is illegally reused" is obtained. The receipt image of the past expense information used by the inspection unit 131 for inspection is associated with the same applicant identifier as the applicant identifier associated with the expense information received by the expense reception unit 121. It is preferable that the receipt image is (a receipt image that the applicant himself used in the past), but a receipt image associated with another applicant identifier may also be used. Specifically, for example, the reuse of the receipt is inspected by any of the following methods (a) and (b).
(A) When using image similarity

The inspection unit 131 has, for example, a degree of similarity between the receipt image of the expense information received by the expense reception unit 121 and the receipt image of each of the past one or more expense information stored in the expense information storage unit 111. If the similarity is equal to or greater than the threshold or greater than the threshold, the inspection result indicating that the receipt is illegally reused is acquired. The inspection result may include past expense information determined to be reused, a receipt image of past expense information determined to be reused, and the like.
(B) By machine learning

The inspection unit 131 is, for example, a positive example of one or more sets of two receipt images determined by a person to reuse receipts, and a negative example of one or more sets of two different receipt images. Using the teacher data possessed, a learner constructed by the learning process of machine learning is constructed. It should be noted that such a learning device is a learning device that outputs whether or not it is reused (whether or not it is the same).

Next, the inspection unit 131 uses the learning device, the receipt image of the expense information received by the expense reception unit 121, and the past receipt image of the expense information storage unit 111 to perform machine learning prediction processing. And when the output result that the receipt is reused is obtained, the inspection result that "the receipt is illegally reused" is acquired. It is preferable that the past receipt image of the expense information storage unit 111 is a receipt image corresponding to the applicant identifier corresponding to the expense information received by the expense reception unit 121.
(11-3) Inspection that the number of expense applications on the same day is large

The inspection unit 131 inspects two or more expense information corresponding to one applicant identifier, determines whether or not the number of expense information having the same usage date information is large enough to satisfy a predetermined condition. Get the test result. When the number of expense information is large enough to satisfy a predetermined condition, the inspection unit 131 acquires an inspection result indicating that the item is fraudulent.

The number of expense information having the same usage date information corresponding to one applicant identifier may be acquired by the statistical processing unit 132.
(11-4) Inspection of fraud due to second party spending

The inspection unit 131 inspects two or more expense information corresponding to one applicant identifier, and the number of expense information having the same usage date information and different restaurant identifiers is so large that a predetermined condition is satisfied. Judge whether or not, and obtain the inspection result. For example, if there is an expense application for two or more restaurants on the same day, it means that the expense is used up to the second party, and the inspection result that "it is fraudulent due to the second party expenditure" is obtained.
(12) Inspection using statistical processing results

The inspection unit 131 inspects whether or not the statistical processing result acquired by the statistical processing unit 132 satisfies a predetermined condition, and acquires the inspection result.
(12-1) Inspection using store frequency information

The inspection unit 131 inspects whether the store frequency information acquired by the statistical processing unit 132 is large enough to satisfy a predetermined condition, and acquires the inspection result. If the value of the store frequency information is large enough to satisfy the store frequency condition, which is a predetermined condition, the inspection unit 131 acquires an inspection result indicating that "one store is being used too much". The store frequency condition is an example of an illegal condition. The store frequency condition is, for example, "store frequency information> = threshold value" and "store frequency information> threshold value". Further, the store frequency condition is stored in, for example, the storage unit 11.
(12-2) Inspection using entertainment frequency information

The inspection unit 131 inspects whether or not the entertainment frequency information acquired by the statistical processing unit 132 satisfies a predetermined condition, and acquires the inspection result. If the value of the entertainment frequency information is large enough to satisfy the entertainment frequency condition, which is a predetermined condition, the inspection unit 131 states that "the number of entertainments, etc. of one partner is too large (there is a possibility of adhesion, etc.)". Obtain the inspection result to that effect. The entertainment frequency condition is an example of an illegal condition. The entertainment frequency condition is, for example, "entertainment frequency information> = threshold value" and "entertainment frequency information> threshold value". Further, the entertainment frequency condition is stored in, for example, the storage unit 11.
(12-3) Inspection of the number of frauds of one applicant

The inspection unit 131 inspects whether or not the fraud count acquired by the statistical processing unit 132 satisfies the fraud count condition, which is a predetermined condition, and acquires the inspection result that "the applicant has a large fraud count". .. The predetermined conditions are, for example, "the number of frauds so far is equal to or greater than the threshold value or is greater than the threshold value" and "the number of fraudulent acts in a unit period (for example, 1 year or 3 years) is equal to or greater than the threshold value or is greater than the threshold value". The fraudulent count condition is an example of the fraudulent condition. The fraud count condition is, for example, "number of frauds> = threshold value" and "number of frauds> threshold value". Further, the invalid number condition is stored in, for example, the storage unit 11.
(13) Detection of fraud by the approver of the expense application

The inspection unit 131 inspects whether or not the approver identified by the approver identifier of the received expense information is the approver specified by the approver information, and acquires the inspection result.
(13-1) When an approverable approver is managed

The inspection unit 131 determines, for example, whether or not the approver identifier of the received expense information is included in the approver information stored in the company regulation information of the company regulation information storage unit 113, and includes the approver identifier. If not, get the inspection result to the effect that you are an illegal approver.
(13-2) When approvable job titles are managed

For example, the inspection unit 131 acquires a job title identifier paired with the approver identifier of the received expense information from the personal information storage unit 1121, and stores the job title identifier in the company regulation information of the company regulation information storage unit 113. It is judged whether or not it is included in the approved approver information, and if it is not included, the inspection result indicating that the approver is "illegal approver" is acquired.
(14) Detection of omission of application

The inspection unit 131 uses the applicant information corresponding to one applicant identifier, and stores the expense information corresponding to the applicant information and one or more expense information associated with the one applicant identifier. It is determined whether or not it exists in the expense information storage unit 111, and the omission of the application for whether or not the expense corresponding to the applicant information has been applied is inspected, and the inspection result is acquired. The applicant information is, for example, attendance information, entrance / exit information, business activity information, card details information, and business vehicle location information.
(14-1) Detection of omission of application using attendance information (a) Omission of application for labor wage expenses

The inspection unit 131 acquires the attendance information corresponding to one applicant identifier from the attendance information storage unit 1122. Next, the inspection unit 131 determines whether or not the acquired attendance information is information corresponding to attendance that does not fall within the salary range. Next, when it is determined that the attendance information corresponds to the attendance that does not fall within the salary range, the inspection department 131 corresponds to one applicant identifier and uses the same as the working day information possessed by the attendance information. It is determined whether or not the expense information having the daily information and having the expense type identifier (for example, "labor wage") indicating the labor wage exists in the expense information storage unit 111. If it is determined that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the attendance information has not been applied for".

Regarding whether or not the acquired attendance information is an attendance that does not fall within the salary range, the inspection unit 131 acquires, for example, the day of the week information that the attendance information has, and the day of the week information is not a working day (for example, Saturday, Sunday). , Holidays), it is judged that the acquired attendance information is not within the range of salary. The inspection unit 131 acquires, for example, the applicant identifier corresponding to the attendance information, and the applicant type information paired with the applicant identifier is predetermined information (for example, "part-time job", "part"). If this is the case, it is determined that the acquired attendance information is not within the salary range. The applicant type information is the information possessed by the personal information of the personal information storage unit 1121.
(B) Omission of application for business trip expenses

The inspection unit 131 acquires the attendance information corresponding to one applicant identifier from the attendance information storage unit 1122. Next, the inspection unit 131 determines whether or not the acquired attendance information is information corresponding to a business trip. When the attendance information is information corresponding to a business trip, the inspection unit 131 is expense information corresponding to one applicant identifier and having the same usage date information as the working day information possessed by the attendance information, and is the expense information of the business trip. It is determined whether or not the expense information having the expense type identifier (for example, "travel expense" and "travel expense") of the expense to be applied for at the time exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the business trip has not been applied for".

The inspection unit 131 determines that, for example, when the attendance information has a business trip flag indicating that the attendance information is a business trip, the attendance information is information corresponding to the business trip. Further, the inspection unit 131 determines, for example, that the attendance information is information corresponding to a business trip when the attendance information does not have the start time information and the end time information. Further, the inspection unit 131 determines that, for example, when the attendance information has a point identifier indicating a business trip destination, the attendance information is information corresponding to the business trip.
(C) Missing application for overseas daily allowance

The inspection unit 131 acquires the attendance information corresponding to one applicant identifier from the attendance information storage unit 1122. Next, the inspection unit 131 determines whether or not the acquired attendance information is information corresponding to an overseas business trip. When the attendance information is information corresponding to an overseas business trip, the inspection department 131 is expense information corresponding to one applicant identifier and having the same usage date information as the working day information possessed by the attendance information, and is overseas. It is determined whether or not the expense information having the expense type identifier (for example, "overseas daily allowance" and "overseas travel allowance") of the expense to be applied for during the business trip exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the overseas business trip has not been applied for".

The inspection unit 131 determines that, for example, when the attendance information has an overseas business trip flag indicating that the attendance information is an overseas business trip, the attendance information corresponds to the overseas business trip. Further, the inspection unit 131 determines that, for example, when the attendance information has a point identifier indicating an overseas business trip destination, the attendance information is information corresponding to the overseas business trip.

Further, the inspection unit 131 acquires, for example, the specified amount information of the overseas business trip held by the company regulation information of the company regulation information storage unit 113, compares it with the amount information of the corresponding expense information, and if they do not match, "overseas". You may obtain the inspection result to the effect that "the application for daily allowance is fraudulent".

Further, the inspection unit 131 acquires, for example, the sales activity information corresponding to one applicant identifier from the sales activity information storage unit 1124. Then, when the inspection unit 131 determines that the point identifier possessed by the business activity information is the point identifier corresponding to the overseas business trip, the inspection unit 131 acquires the business day information possessed by the business activity information. Next, the inspection unit 131 is expense information corresponding to one applicant identifier and having the same usage date information as the acquired business day information, and is an expense having an expense type identifier of the expense to be applied for during an overseas business trip. It is determined whether or not the information exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the overseas business trip has not been applied for".
(14-2) Detection of omission of application using entry / exit information (a) Omission of application for labor wage expenses

The inspection unit 131 acquires the entrance / exit information corresponding to one applicant identifier from the entrance / exit information storage unit 1123. Next, the inspection unit 131 determines whether or not the acquired entrance / exit information corresponds to work that does not fall within the salary range. Next, if it is determined that the entry / exit information corresponds to work that does not fall within the range of wages, the inspection department 131 corresponds to one applicant identifier and the work day information contained in the entry / exit information. It is determined whether or not the expense information having the same usage date information as the above and having the expense type identifier (for example, "labor wage") indicating the labor wage exists in the expense information storage unit 111. If it is determined that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the entrance / exit information has not been applied for".

Regarding whether or not the acquired entrance / exit information is attendance that does not fall within the salary range, the inspection department 131 acquires, for example, the day of the week information that the entrance / exit information has, and the day of the week information is not a working day (for example). , Saturdays, Sundays, national holidays), it is judged that the acquired entrance / exit information is not within the salary range. The inspection unit 131 acquires, for example, the applicant identifier corresponding to the entrance / exit information, and the applicant type information to be paired with the applicant identifier is predetermined information (for example, "part-time job", "part"). If this is the case, it is determined that the work does not fall within the salary range of the acquired entrance / exit information.
(B) Omission of application for business trip expenses

The inspection unit 131 acquires the entrance / exit information corresponding to one applicant identifier from the entrance / exit information storage unit 1123. Next, the inspection unit 131 determines whether or not the acquired entrance / exit information is information corresponding to a business trip. When the entry / exit information is information corresponding to a business trip, the inspection department 131 corresponds to one applicant identifier and uses the expense information having the same usage date information as the work day information of the entry / exit information. Yes, it is determined whether or not the expense information having the expense type identifier (for example, "travel expense" and "travel expense transportation expense") of the expense to be applied for during the business trip exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the business trip has not been applied for".

The inspection unit 131 determines, for example, that the entrance / exit information is information corresponding to a business trip when the entrance / exit information does not have the start time information and the end time information. The inspection unit 131, for example, has attendance information such that the entrance / exit information does not have the start time information and the end time information, corresponds to one applicant identifier, and has the same work day information as the use date information ( If it is determined that the information indicates that the employee is not absent), it is determined that the entry / exit information corresponds to the business trip.
(14-3) Detection of omission of application using sales activity information (a) Omission of application for business trip expenses

The inspection unit 131 acquires the sales activity information corresponding to one applicant identifier from the sales activity information storage unit 1124. Next, the inspection unit 131 determines whether or not the acquired sales activity information is information corresponding to a business trip. When the business activity information is information corresponding to a business trip, the inspection unit 131 is expense information corresponding to one applicant identifier and having the same usage date information as the business day information possessed by the business activity information. It is determined whether or not the expense information having the expense type identifier (for example, "travel expense" and "travel expense transportation expense") of the expense to be applied for during the business trip exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the business trip has not been applied for".

The inspection unit 131 uses, for example, the point identifier of the sales activity information and the predetermined position information of the company to determine the distance between the point and the company identified by the point identifier in the map information storage unit 1182. When it is acquired using map information and it is determined that the distance is greater than or equal to the distance in which the travel expenses are incurred, it is determined that the sales activity information is the information corresponding to the business trip. Information on the minimum distance at which business trip expenses are incurred is stored in, for example, the company regulation information storage unit 113.

Further, the inspection unit 131 determines, for example, that the sales activity information is information corresponding to a business trip when the sales activity information has a point identifier.

The algorithm for determining that the sales activity information corresponds to the business trip does not matter. The business activity information may have a flag indicating whether or not it is a business trip.
(B) Missing application for transportation expenses

The inspection unit 131 acquires one or more point identifiers of the business activity information corresponding to one applicant identifier from the business activity information storage unit 1124. Next, the inspection unit 131 determines whether or not transportation costs are incurred by using the location information of the company stored in advance and the acquired one or more point identifiers. When it is determined that transportation expenses will be incurred, the inspection department 131 is expense information having the same usage date information as the business day information possessed by the business activity information, and the expense information having the expense type identifier of "travel expense transportation expenses" is the expense. It is determined whether or not it exists in the information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result to the effect that "the expense for transportation expenses has not been applied for".

The inspection unit 131 determines whether or not transportation costs are incurred by using, for example, the calculation information of the calculation information storage unit 1181. Specifically, the inspection unit 131, for example, gives the location information and the point identifier of the company to the calculation module for calculating the transportation cost, and determines whether or not the information on the transportation cost is output. Further, the inspection unit 131 acquires, for example, the distance between the company's location information and the point identifier using map information, and determines whether or not transportation costs are incurred depending on whether or not the distance is equal to or greater than the threshold value. to decide. The calculation module is a known technique.
(14-4) Detection of omission of application using business vehicle location information (a) Omission of application for business trip expenses

The inspection unit 131 acquires one applicant identifier and one or more business vehicle position information corresponding to one usage date information from the business vehicle position information storage unit 1126. Next, the inspection unit 131 determines whether or not the acquired one or more business vehicle position information is information corresponding to a business trip. When the business vehicle position information is information corresponding to a business trip, the inspection unit 131 corresponds to one applicant identifier and has the same usage date information as one usage date information corresponding to the business vehicle position information. It is determined whether or not the expense information having the expense type identifier (for example, "business trip expense") of the expense to be applied for during the business trip exists in the expense information storage unit 111. When the inspection unit 131 determines that the expense information does not exist in the expense information storage unit 111, the inspection unit 131 acquires the inspection result indicating that "the expense corresponding to the business trip has not been applied for".

In addition, the inspection unit 131 maps the maximum value of the distance between the position indicated by the business vehicle position information and the company by using, for example, one or more business vehicle position information and the company position information stored in advance. When it is acquired using the map information of the information storage unit 1182 and it is determined that the maximum value of the distance is a distance equal to or greater than the distance in which the travel expenses are incurred, it is determined that the sales activity information is the information corresponding to the business trip. do. Information on the minimum distance at which business trip expenses are incurred is stored in, for example, the company regulation information storage unit 113.

Further, the inspection unit 131 determines that the business activity information is the information corresponding to the business trip when, for example, there is a business trip flag associated with each one or more business vehicle position information. The business trip flag is a flag indicating that the business trip is a business trip.

The output unit 14 outputs various information. The various types of information are, for example, inspection results, image inspection results, and statistical processing results. Here, the output is usually transmission to the applicant terminal 2, but may be transmission to another device such as an accounting terminal, storage in the expense inspection device 1, display on the display, or the like.
(15) Receipt image inspection performed by image inspection means 1311

The image inspection means 1311 performs, for example, one or more of the character recognition process and the machine learning process (one of the above methods (1) to (3), etc.) to acquire one or more character strings. It is determined whether or not each of the one or more character strings satisfies a predetermined condition, and the image inspection result is acquired. The character string of 1 or more is, for example, any one of payment amount information, store identifier, destination identifier, date of use information, subject, description, and payment means information.

Examples of the inspection performed by the image inspection means 1311 are the following (15-1) to (15-9). The image inspection means 1311 inspects at least one of (15-1) to (15-9), and obtains an image inspection result. In addition, (15-1) to (15-9) are examples of receipt image inspection.
(15-1) Inspection of inconsistency between receipt image and character string of expense information

The image inspection means 1311 acquires one or more sets of a character string and a class from the receipt image of the expense information received by the expense reception unit 121. Next, the inspection unit 131 corresponds to the class of the group for each group, and the character string (for example, payment amount information, store identifier, destination identifier, date of use) of the expense information received by the expense reception unit 121. Obtain information, subject, description, or payment method information). Next, the inspection unit 131 compares the character string of one class acquired from the receipt image with the character string of the same class of the expense information received by the expense reception unit 121 for each group, and is determined in advance. Determine if you have a relationship. If there is no predetermined relationship, the inspection unit 131 acquires an inspection result indicating fraud.

It should be noted that the predetermined relationships are the same, but they are similar enough to satisfy the predetermined conditions (for example, the degree of similarity is equal to or higher than the threshold value, and only the difference between "Co., Ltd." and "Co., Ltd." is described. Only the difference, etc.).
(15-2) Inspection of monetary processing (a) When character recognition processing is used

The image inspection means 1311 obtains a character string and a score in the area of the amount of money from the receipt image by, for example, a character recognition technique, and determines that the score is low enough to satisfy a predetermined condition, the receipt. It is judged that the amount of money has been processed. The image inspection means 1311 acquires, for example, all the characters of the receipt image by the character recognition technology, detects the character "¥" from the set of the characters, and uses the number string next to the character "¥" as the payment amount. Get as information.
(B) When using machine learning processing

For example, the image inspection means 1311 performs learning processing by a machine learning algorithm using one or more receipt images whose amount is disguised by hand and one or more receipt images that are not disguised as teacher data, and receives the receipt. It constitutes a learning device that acquires a test result as to whether or not the amount of money has been disguised by inputting a calligraphy image.

Next, the image inspection means 1311 performs prediction processing by a machine learning algorithm using, for example, the receipt image received by the expense reception unit 121 and the learning device, and the amount of money described in the receipt image is determined. Obtain an image inspection result indicating whether or not the image was disguised by hand.
(15-3) Inspection of improper tax classification

The image inspection means 1311 acquires two or more character strings from the receipt image. Next, the image inspection means 1311 determines whether or not the two or more character strings match a predetermined fraudulent pattern. When the fraudulent pattern is matched, the image inspection means 1311 acquires an inspection result indicating that the fraudulent pattern is used.

The fraudulent pattern is, for example, to include the character string "souvenir" in the character string corresponding to the class "summary" and "10%" in the character string corresponding to the class "tax classification". .. Such fraud patterns are tax classification fraud. The image inspection means 1311, for example, uses the acquired character string to include "souvenir" in the character string corresponding to the class "summary" and corresponds to the class "tax classification". If it is determined that "10%" is included in the character string, the image inspection result indicating that the character string is "illegal tax classification" is acquired.
(15-4) Inspection of address deficiencies

The image inspection means 1311 acquires one or more character strings from the receipt image in association with the class.

Next, when the image inspection means 1311 determines that the character string corresponding to the class "address" does not match the company name stored in the storage unit 11, the image inspection result indicating that the address of the receipt is inadequate. To get. Further, for example, in the image inspection means 1311, the character string corresponding to the class "address" is an inappropriate character string such as "superior" stored in the storage unit 11 by using the acquired character string. If it is determined that the receipt is incomplete, the image inspection result is obtained.
(15-5) Inspection of receipts for signatures

The image inspection means 1311 constructs a learning device by a machine learning algorithm, for example, using teacher data in which one or more receipts with a signature are a positive example and one or more receipts without a signature are a negative example. do. Then, the image inspection means 1311 performs a prediction process using the received receipt image and the learning device, and acquires information indicating the presence or absence of a signature. Then, when the image inspection means 1311 determines that there is no signature, the image inspection means 1311 acquires an image inspection result indicating a defect that there is no signature on the receipt.
(15-6) Inspection of unclearness of receipt image (a) When character recognition processing is used (first unclearing processing)

The image inspection means 1311 uses character recognition technology to obtain a character recognition score together with one or more character strings from the receipt image received by the expense reception unit 121, and the score is low enough to satisfy a predetermined condition. In this case, the first blurring process for determining that the receipt image is blurry is performed. It should be noted that a technique of performing character recognition, acquiring a character string, and outputting a character recognition score (which may be called likelihood) is a known technique.
(B) When machine learning processing is used (second blurring processing)

The image inspection means 1311 performs learning processing on one or more unclear receipt images and one or more clear receipt images by a machine learning algorithm to form a learning device. The learning device is a learning device that receives a receipt image as an input and acquires a determination result of whether the receipt image is unclear or clear.

Next, the image inspection means 1311 performs prediction processing by a machine learning algorithm using the receipt image received by the expense reception unit 121 and the learning device, and determines whether the receipt image is unclear or clear. Get the test result.
(15-7) Inspection of payment method

The image inspection means 1311 acquires the payment means information (for example, "○○ PAY") which is a character string from the receipt image, and inspects whether the payment means information is the payment means information possessed by the company regulation information. And get the image inspection result. The company regulation information exists in the company regulation information storage unit 113.
(15-8) Inspection of the presence or absence of square marks on handwritten receipts (a) When using machine learning processing

The image inspection means 1311 is, for example, handwritten, with one or more receipt images with a square stamp as a positive example and one or more receipt images without a square stamp as a negative example for machine learning. The learning process is performed by the algorithm to configure the learning device. The learning device is a learning device that inputs a receipt image and outputs whether or not a square mark is pressed.

Next, the image inspection means 1311 performs prediction processing by a machine learning algorithm using the received receipt image and the learning device, and acquires an image inspection result indicating whether or not the square mark is stamped.
(B) When using image recognition processing

The image inspection means 1311 performs character recognition processing on the received receipt image, and determines whether or not handwritten characters exist (for example, whether or not the score is below or below the threshold value). When it is determined that the handwritten character exists, the image inspection means 1311 determines, for example, by image recognition whether or not there is a square mark having a pixel value in the red region and having a substantially rectangular outline. do. If it is determined that the square mark does not exist, the image inspection means 1311 acquires an image inspection result indicating that the square mark is not pressed.
(15-9) Timestamp inspection

The image inspection means 1311 acquires the date information indicated by the time stamp of the received receipt image. The day information is information indicating the day when the time stamp is added to the receipt image.

Next, the image inspection means 1311 acquires the receipt issuance date corresponding to the received receipt image.

Next, the image inspection means 1311 determines whether or not the date indicated by the acquired date information has passed a predetermined period or longer with respect to the receipt issuance date, and acquires the image inspection result. .. The predetermined period or longer is, for example, "3 days or longer" or "4 days or longer".

The image inspection means 1311 usually acquires the receipt issuance date from the receipt image, but may acquire the receipt issuance date, which is a character string included in the expense information received by the expense reception unit 121. Further, the technique of adding a time stamp to the receipt image and the technique of acquiring the time stamp of the receipt image are known techniques.

Storage unit 11, expense information storage unit 111, applicant collective storage unit 112, company regulation information storage unit 113, restaurant collective storage unit 114, customs store collective storage unit 115, destination collective storage unit 116, card usage expense storage unit 117, transportation information storage unit 118, personal information storage unit 1121, attendance information storage unit 1122, entrance / exit information storage unit 1123, sales activity information storage unit 1124, card detail information storage unit 1125, business vehicle position information storage unit 1126. , The calculation information storage unit 1181, the map information storage unit 1182, and the first storage unit 21 are preferably a non-volatile recording medium, but can also be realized by a volatile recording medium.

The process of storing information in the storage unit 11 or the like does not matter. For example, the information may be stored in the storage unit 11 or the like via the recording medium, or the information transmitted via the communication line or the like may be stored in the storage unit 11 or the like. Alternatively, the information input via the input device may be stored in the storage unit 11 or the like.

The reception unit 12, the expense reception unit 121, and the first reception unit 25 are usually realized by wireless or wired communication means, but may be realized by means for receiving broadcasts.

The processing unit 13, the inspection unit 131, the statistical processing unit 132, the image inspection means 1311, and the first processing unit 23 can usually be realized from a processor, a memory, or the like. The processing procedure of the processing unit 13 is usually realized by software, and the software is recorded in a recording medium such as a ROM. However, it may be realized by hardware (dedicated circuit). The type of processor may be MPU, CPU, GPU, or the like.

The output unit 14, the result output unit 141, and the first transmission unit 24 are usually realized by wireless or wired communication means, but may be realized by broadcasting means.

The first reception unit 22 can be realized by a device driver for input means such as a touch panel or a keyboard, control software for a menu screen, or the like.

The first output unit 26 may or may not include an output device such as a display or a speaker. The first output unit 26 can be realized by the driver software of the output device, the driver software of the output device, the output device, or the like.
To prepare for.

Next, an operation example of the expense inspection device 1 will be described with reference to the flowchart of FIG.

(Step S401) The expense reception unit 121 determines whether or not the expense information corresponding to the application form identifier has been received from the applicant terminal 2. If the expense information is received, the process goes to step S402, and if the expense information is not received, the process goes to step S406.

(Step S402) The processing unit 13 performs inspection processing on the expense information received in step S401. The inspection process is the process of inspecting for fraudulent expense applications. An example of the inspection process will be described with reference to the flowchart of FIG.

(Step S403) The inspection unit 131 determines whether or not the inspection result acquired in step S402 is “valid”. If it is "legitimate", it goes to step S404, and if it is not "legitimate", it goes to step S405.

(Step S404) The processing unit 13 or the storage unit (not shown) stores the expense information received in step S401 in the expense information storage unit 111 in association with the application form identifier. Return to step S401. In step S404, the result output unit 141 may transmit the inspection result to the effect that it is "legitimate" to the applicant terminal 2 that has transmitted the expense information.

(Step S405) The result output unit 141 transmits the inspection result acquired in step S402 to the applicant terminal 2 that has transmitted the expense information. Return to step S401. In step S405, the inspection unit 131 may increment the number of frauds corresponding to the applicant identifier corresponding to the expense information by 1. Further, in step S405, the processing unit 13 or the storage unit (not shown) may store the expense information received in step S401 in the expense information storage unit 111 in association with the application form identifier.

(Step S406) The processing unit 13 determines whether or not it is time to perform a statistical inspection. If it is the timing to perform the statistical inspection, it goes to step S407, and if it is not the timing to perform the statistical inspection, it goes to step S414.

The processing unit 13 acquires a time from a clock (not shown), and if the time is after a predetermined time, it determines that it is time to perform a statistical inspection. Further, the processing unit 13 determines that it is the timing to perform the statistical inspection when the instruction of the statistical inspection is received. The statistical inspection is a fraudulent inspection using the statistical processing result of expense information. Statistical tests are preferably performed, for example, on a regular basis (eg, once a month).

(Step S407) The processing unit 13 substitutes 1 for the counter i.

(Step S408) The processing unit 13 determines whether or not the i-th applicant identifier exists in the expense information storage unit 111. If the i-th applicant identifier exists, the process goes to step S409, and if the i-th applicant identifier does not exist, the process goes to step S412. Whether or not the i-th applicant identifier exists is whether or not the i-th applicant to be inspected exists.

(Step S409) The processing unit 13 performs statistical processing on the expense information corresponding to the i-th applicant identifier, and uses the statistical processing result to inspect the expense application for fraud. An example of such statistical inspection processing will be described with reference to the flowchart of FIG.

(Step S410) The processing unit 13 outputs the inspection result acquired in step S409 in association with the i-th applicant identifier. The processing unit 13 stores the inspection result in the storage unit 11, for example, in association with the i-th applicant identifier. Further, the processing unit 13 transmits, for example, the inspection result associated with the i-th applicant identifier to a terminal (not shown).

(Step S411) The processing unit 13 increments the counter i by 1. Return to step S408.

(Step S412) The statistical processing unit 132 configures the statistical processing result to be output.

(Step S413) The result output unit 141 outputs the statistical processing result configured in step S412. Return to step S401. The output of the statistical processing result is, for example, storing the statistical processing result in the storage unit 11 and transmitting the statistical processing result to a terminal (not shown).

(Step S414) The processing unit 13 determines whether or not it is time to perform a declaration omission inspection. If it is the timing to perform the declaration omission inspection, the process goes to step S415, and if it is not the timing to perform the declaration omission inspection, the process returns to step S401.

The processing unit 13 acquires the time from a clock (not shown), and if the time is after a predetermined time, it determines that it is the timing to perform the declaration omission inspection. Further, the processing unit 13 determines that it is time to perform the declaration omission inspection when the instruction of the declaration omission inspection is received. In addition, the omission inspection is an inspection for omission of expense application. It is preferable that the declaration omission inspection is performed, for example, on a regular basis (for example, once a month).

(Step S415) The processing unit 13 performs a leak inspection of the expense application. An example of leak inspection processing will be described with reference to the flowchart of FIG. 7.

(Step S416) The result output unit 141 outputs the inspection result of the leak inspection acquired in step S415. Return to step S401. The output of the inspection result here is, for example, storage of the inspection result in the storage unit 11, transmission to the applicant identified by the applicant identifier corresponding to the omission, transmission to a terminal (not shown), and the like.

In the flowchart of FIG. 4, the process ends when the power is turned off or an interrupt for the end of the process is interrupted.

Next, an example of the inspection process in step S402 will be described with reference to the flowchart of FIG.

(Step S501) The inspection unit 131 determines whether or not the receipt image is included in the received expense information. If the receipt image is included, the process goes to step S502, and if the receipt image is not included, the process goes to step S505.

(Step S502) The image inspection means 1311 acquires one or two or more sets of a character string and a class from the receipt image. Since such processing has been described above, detailed description thereof will be omitted here.

(Step S503) The image inspection means 1311 performs a receipt image inspection and acquires an inspection result.

The receipt image inspection is, for example, one or more inspections of (15-1) to (15-9) described above.

(Step S504) The inspection unit 131 performs a common inspection and acquires an inspection result. The common inspection is an inspection that does not depend on the expense type identifier (may be a subject) of expense information.

Common inspections include, for example, (1-1) (2-1) inspection of expenses other than work days, (4-1) inspection of receipt division, (5-1) inspection of unauthorized use of business vehicles, and so on. (6) Inspection using card usage expense information, (10) Inspection using destination information, (11-1) Inspection of double application, (11-2) Inspection of reuse of receipts, (11-) 3) Inspection of the large number of expense applications on the same day, and (13) Detection of fraud by the approver of the expense application.

(Step S505) The inspection unit 131 acquires the expense type identifier from the received expense information.

(Step S506) The inspection unit 131 determines whether or not the expense type identifier acquired in step S505 is "travel expense (travel expense or travel expense)". If it is "transportation expense", it goes to step S507, and if it is not "transportation expense", it goes to step S508.

(Step S507) The inspection unit 131 performs an entertainment expense inspection process and acquires an inspection result. Go to step S514.

The transportation expense inspection process is a process for inspections specific to transportation expenses. The transportation expense inspection process is, for example, the above-mentioned (3-1) transportation expense fraud inspection (i), (3-2) transportation expense fraud inspection (ii), (4-2) after the in-house drinking party. Inspection of taxi use, (8-1) inspection of fraudulent transportation expenses for trains, buses, or planes, (8-2) inspection of fraudulent taxi fare, (8-3) inspection of duplication with regular sections. ..

(Step S508) The inspection unit 131 determines whether or not the expense type identifier acquired in step S505 is “entertainment expense (may be entertainment expense)”. If it is "entertainment expenses", it goes to step S509, and if it is not "entertainment expenses", it goes to step S510.

(Step S509) The inspection unit 131 performs an entertainment expense inspection process and acquires an inspection result. Go to step S514.

The entertainment expense inspection process is a process for inspection peculiar to the entertainment expense. The entertainment expense inspection process is, for example, (9-1) inspection of store fraud (use of sex shops, etc.), (9-2) inspection of amount fraud, and (9-3) whether entertainment expenses use restaurants. It is an inspection of whether or not, (10) an inspection using the other party's information, and (11-4) an inspection of fraud due to the expenditure of the second party.

(Step S510) The inspection unit 131 determines whether or not the expense type identifier acquired in step S505 is “accommodation expense”. If it is "accommodation fee", it goes to step S511, and if it is not "accommodation fee", it goes to step S512.

(Step S511) The inspection unit 131 performs an accommodation expense inspection process and acquires an inspection result. Go to step S514.

The accommodation expense inspection process is a process for inspections specific to accommodation expenses. The accommodation expense inspection process is, for example, (1-3) an inspection for fraudulent accommodation expenses and (2-3) an inspection for fraudulent accommodation expenses.

(Step S512) The inspection unit 131 determines whether or not the expense type identifier acquired in step S505 is a “business trip expense”. If it is "business trip expense", it goes to step S513, and if it is not "business trip expense", it goes to step S512.

(Step S513) The inspection unit 131 performs a business trip expense inspection process and acquires an inspection result. Go to step S514.

The business trip expense inspection process is a process for inspection peculiar to the business trip expense. The business trip expense inspection process is, for example, (1-2) inspection of business trip expenses fraud and (2-2) inspection of business trip expenses fraud.

(Step S514) The inspection unit 131 constitutes an output inspection result by using the acquired inspection result. Return to higher-level processing.

For example, when one or more fraudulent inspection results are obtained, the inspection unit 131 constitutes an inspection result including all the inspection results in the case of fraud. Further, when an incorrect inspection result is not obtained, the inspection unit 131 constitutes an inspection result to the effect that it is “legitimate”.

Next, an example of the statistical inspection process in step S409 will be described with reference to the flowchart of FIG.

(Step S601) The statistical processing unit 132 substitutes 1 for the counter i.

(Step S602) The statistical processing unit 132 is one or more expense information corresponding to the applicant identifier to be inspected (the i-th applicant identifier in step S408), and the expense information storage unit 111 has one or more expenses. It is determined whether or not the i-th store identifier exists in the store identifier in the information. If the i-th store identifier exists, the process goes to step S603, and if the i-th store identifier does not exist, the process goes to step S607.

The i-th store identifier is the i-th store identifier in which the store identifier in one or more expense information corresponding to the applicant identifier to be inspected is uniquely processed. Further, the period of the expense information to be inspected may be predetermined or may be the entire period.

(Step S603) The statistical processing unit 132 acquires the number of expense information including the i-th store identifier corresponding to the applicant identifier to be inspected. The statistical processing unit 132 acquires store frequency information using the number. The store frequency information may be the same value as the number, or may be a value obtained by dividing the number by the period. The store frequency information is usually a value calculated by an increasing function with the number as a parameter.

(Step S604) The inspection unit 131 determines whether or not the store frequency information acquired in step S603 is large enough to satisfy a predetermined store frequency condition. If the store frequency condition is satisfied, the process goes to step S605, and if the store frequency condition is not satisfied, the process goes to step S606.

(Step S605) The inspection unit 131 acquires the inspection result indicating that "one store is being used too much" and the inspection result having the i-th store identifier in association with the applicant identifier to be inspected. The inspection unit 131 may increment the number of frauds corresponding to the applicant identifier to be inspected by one.

(Step S606) The statistical processing unit 132 increments the counter i. Return to step S602.

(Step S607) The statistical processing unit 132 substitutes 1 for the counter i.

(Step S608) The statistical processing unit 132 is one or two or more expense information corresponding to the applicant identifier to be inspected, and is the i-th store identifier in the one or more expense information of the expense information storage unit 111. Determine if the destination identifier exists. If the i-th destination identifier exists, the process goes to step S609, and if the i-th destination identifier does not exist, the process goes to step S613.

The i-th counterparty identifier is the i-th counterparty identifier in which the counterparty identifier in one or more expense information corresponding to the applicant identifier to be inspected is uniquely processed. Further, the period of the expense information to be inspected may be predetermined or may be the entire period.

(Step S609) The statistical processing unit 132 acquires the number of expense information including the i-th destination identifier corresponding to the applicant identifier to be inspected. The statistical processing unit 132 acquires the destination frequency information using the number. The destination frequency information may be the same value as the number, or may be a value obtained by dividing the number by the period. The destination frequency information is usually a value calculated by an increasing function with the number as a parameter.

(Step S610) The inspection unit 131 determines whether or not the partner frequency information acquired in step S603 is large enough to satisfy a predetermined partner frequency condition. If the other party frequency condition is satisfied, the process goes to step S611, and if the other party frequency condition is not satisfied, the process goes to step S612.

(Step S611) The inspection unit 131 associates the inspection result to the effect that "the number of entertainments of one counterparty is too large" and the inspection result having the i-th counterparty identifier with the applicant identifier to be inspected. To get. The inspection unit 131 may increment the number of frauds corresponding to the applicant identifier to be inspected by one.

(Step S612) The statistical processing unit 132 increments the counter i. Return to step S608.

(Step S613) The inspection unit 131 acquires the number of frauds corresponding to the applicant identifier to be inspected from the personal information storage unit 1121.

(Step S614) The inspection unit 131 determines whether or not the fraud count acquired in step S613 satisfies the fraud count condition. If the fraud count condition is satisfied, the process goes to step S615, and if the fraud count condition is not satisfied, the process goes to step S616.

(Step S615) The inspection unit 131 acquires an inspection result indicating that “the number of frauds of the applicant is large” in association with the applicant identifier to be inspected. The inspection result may be only the applicant identifier with a large number of frauds.

(Step S616) The inspection unit 131 determines whether or not the fraudulent inspection result could be obtained as a result of the statistical inspection. If the inspection result is acquired, the process returns to the higher-level processing, and if the inspection result is not acquired, the process proceeds to step S617.

(Step S617) The inspection unit 131 substitutes the variable “inspection result” to the effect that it is “valid”. Return to higher-level processing.

Next, a process example of the leak inspection in step S415 will be described with reference to the flowchart of FIG. 7.

(Step S701) The inspection unit 131 substitutes 1 for the counter i. The counter i is a counter of the applicant to be inspected.

(Step S702) The inspection unit 131 determines whether or not the i-th applicant identifier exists in the storage unit 11. If the i-th applicant identifier exists, the process proceeds to step S703, and if the i-th applicant identifier does not exist, the process returns to higher-level processing. The i-th applicant identifier is the identifier of the i-th applicant to be inspected.

(Step S703) The inspection unit 131 substitutes 1 for the counter j. The counter j is a counter for the day to be inspected.

(Step S704) The inspection unit 131 determines whether or not there is day information indicating the jth day of the inspection target. If the day information indicating the j-th day exists, the process goes to step S705, and if it does not exist, the process goes to step S722. It is assumed that the leak inspection date is predetermined and the information for specifying the leak inspection date is stored in a buffer (not shown).

(Step S705) The inspection unit 131 determines whether or not the attendance information storage unit 1122 has attendance information corresponding to the i-th applicant identifier and having the same work day information as the day information indicating the j-th day. to decide. If it exists, it goes to step S706, and if it does not exist, it goes to step S709.

(Step S706) The inspection unit 131 acquires time information from the attendance information storage unit 1122, which corresponds to the i-th applicant identifier and has the same work day information as the day information indicating the j-th day. The inspection unit 131 uses the acquired attendance information to inspect for omissions in expense applications. Such an inspection is, for example, the process described in "(14-1) Detection of application omission using attendance information".

(Step S707) The inspection unit 131 determines whether or not there is a leak as a result of the inspection in step S706. If there is a leak, go to step S708, and if there is no leak, go to step S709.

(Step S708) The inspection unit 131 acquires the inspection result indicating that there was a leak and the content of the leak in association with the i-th applicant identifier and the j-th day information.

(Step S709) The inspection unit 131 has entry / exit information in the entrance / exit information storage unit 1123, which corresponds to the i-th applicant identifier and has the same working day information as the day information indicating the j-th day. Judge whether or not. If it exists, it goes to step S710, and if it does not exist, it goes to step S713.

(Step S710) The inspection unit 131 acquires entrance / exit information from the entrance / exit information storage unit 1123, which corresponds to the i-th applicant identifier and has the same working day information as the day information indicating the j-th day. .. The inspection unit 131 inspects the omission of the expense application by using the acquired entrance / exit information. Such an inspection is, for example, the process described in "(14-2) Detection of application omission using entrance / exit information".

(Step S711) The inspection unit 131 determines whether or not there is a leak as a result of the inspection in step S710. If there is a leak, go to step S712, and if there is no leak, go to step S713.

(Step S712) The inspection unit 131 acquires the inspection result indicating that there was a leak and the content of the leak in association with the i-th applicant identifier and the j-th day information.

(Step S713) Whether or not the inspection unit 131 has business activity information in the business activity information storage unit 1124, which corresponds to the i-th applicant identifier and has the same business day information as the day information indicating the j-th day. Judge. If it exists, it goes to step S714, and if it does not exist, it goes to step S717.

(Step S714) The inspection unit 131 acquires business activity information from the business activity information storage unit 1124, which corresponds to the i-th applicant identifier and has the same business day information as the day information indicating the j-th day. The inspection unit 131 uses the acquired sales activity information to inspect for omissions in expense applications. Such an inspection is, for example, the process described in "(14-3) Detection of application omission using business activity information".

(Step S715) The inspection unit 131 determines whether or not there is a leak as a result of the inspection in step S714. If there is a leak, go to step S716, and if there is no leak, go to step S717.

(Step S716) The inspection unit 131 acquires the inspection result indicating that there was a leak and the content of the leak in association with the i-th applicant identifier and the j-th day information.

(Step S717) In the inspection unit 131, the business vehicle position information corresponding to the i-th applicant identifier and having the same business day information as the day information indicating the j-th day exists in the business vehicle position information storage unit 1126. Judge whether or not. If it exists, it goes to step S718, and if it does not exist, it goes to step S721.

(Step S718) The inspection unit 131 stores one or more business vehicle position information corresponding to the i-th applicant identifier and having the same business day information as the day information indicating the j-th day, in the business vehicle position information storage unit 1126. Get from. The inspection unit 131 inspects the omission of the expense application by using the acquired position information of one or more business vehicles. Such an inspection is, for example, the process described in "(14-4) Detection of application omission using business vehicle position information".

(Step S719) The inspection unit 131 determines whether or not there is a leak as a result of the inspection in step S718. If there is a leak, go to step S720, and if there is no leak, go to step S721.

(Step S720) The inspection unit 131 acquires the inspection result indicating that there was a leak and the content of the leak in association with the i-th applicant identifier and the j-th day information.

(Step S721) The inspection unit 131 increments the counter j by 1. Return to step S704.

(Step S722) The inspection unit 131 constitutes an inspection result corresponding to the i-th applicant identifier. The inspection unit 131 temporarily stores the inspection results in association with the i-th applicant identifier, for example. The inspection result has, for example, the result of inspection of one or more leaks. If there is no leak, the inspection result is information indicating "no leak".

(Step S723) The inspection unit 131 increments the counter i by 1. Return to step S702.

Next, an operation example of the applicant terminal 2 will be described. The first reception unit 22 of the applicant terminal 2 accepts the input of expense information from the user. Next, the first processing unit 23 reads the applicant identifier stored in the first storage unit 21. Then, the first processing unit 23 constitutes transmission information having the received expense information and the read applicant identifier. Next, the first transmission unit 24 transmits the transmission information having the expense information and the applicant identifier to the expense inspection device 1.

Next, the first receiving unit 25 receives the inspection result from the expense inspection device 1. The first processing unit 23 has a structure for outputting the received inspection result. The first output unit 26 outputs the inspection result.
(Specific example of statistical processing results)

Hereinafter, an example of the statistical processing result acquired by the statistical processing unit 132 of the expense inspection device 1 and output by the result output unit 141 in the present embodiment will be described.

(Specific example 1)
The statistical processing unit 132 is the expense information of the expense information storage unit 111, and statistically processes the expense information of each applicant for a certain period (for example, one month in December 2019) for each applicant, and the applicant. It is assumed that the store frequency information for each identifier and each store identifier is acquired. Further, it is assumed that the statistical processing unit 132 has acquired the number of expense information for each applicant identifier.

Then, the statistical processing unit 132 plots the store frequency information for each applicant identifier and each store identifier on a graph whose vertical axis is the store frequency information and whose horizontal axis is the number of expense information for each applicant identifier. Suppose you get the result.

Further, it is assumed that the statistical processing unit 132 determines that the risk is high when the store frequency condition "store frequency information> = 3" of the storage unit 11 is satisfied. It can be said that high risk is fraudulent. Then, the statistical processing unit 132 acquires the applicant identifier of the applicant determined to have a high risk, the store identifier of the store to be used, and the store frequency information.

Then, the statistical processing unit 132 constitutes the statistical processing result output as a result of the above processing.

Next, the result output unit 141 outputs the statistical processing result (801, 802). An example of such an output is shown in FIG.

In FIG. 8, 801 is a statistical processing result in which store frequency information is plotted on a graph. The vertical axis of 801 is the maximum number of cases (store frequency information) for each store by person. The horizontal axis of 801 is the number of applications by person.

Further, 802 is a table including records having an applicant identifier (applicant), a store name, and store frequency information (number of cases) corresponding to the store frequency information determined to be high risk.

(Specific example 2)
Specific example 2 is an example of acquiring information when there is a stamp even though the person is on a business trip.

For example, when the statistical processing unit 132 indicates that the attendance information is a business trip for each applicant identifier, or when the expense information of the business trip expense or the transportation expense is applied, the working day information possessed by the attendance information. Alternatively, it is determined that the day corresponding to the usage date information of the expense information is the day of the business trip.

Next, the statistical processing unit 132 is entry / exit information including the same working day information as the working day information of the attendance information or the usage day information of the expense information, and the work start time and work end time are set. Obtain the frequency of existence (with stamping) (for example, the number of times in one month of a certain month). Such information is frequently stamped during business trips.

Further, in the statistical processing unit 132, for example, for each applicant identifier, the working hours indicated by each attendance information for one month and the working hours indicated by the entrance / exit information including the same working day information as the working day information possessed by the attendance information. Get the average value of the difference from time. Such information is the discrepancy between attendance information and stamping.

Then, the statistical processing unit 132 plots the dots of the applicant identifier on a graph in which the frequency of stamping during a business trip is on the horizontal axis and the deviation between attendance information and stamping is on the vertical axis for each applicant identifier. Configure the statistical processing results that are graphs.

Next, the result output unit 141 outputs the statistical processing result. An example of such an output is shown in FIG. In FIG. 9, the applicant corresponding to the dots in the area of 901 is an applicant who is likely to be fraudulent.

As described above, according to the present embodiment, by using the expense information other than the external information or the applied expense information, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, by using the applicant information, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, by using the attendance information of the applicant, an inspection can be appropriately performed for an inappropriate expense application.

Further, according to the present embodiment, by using the entry / exit information of the applicant, an inspection can be appropriately performed for an inappropriate expense application.

Further, according to the present embodiment, by using the business activity information of the applicant, an inspection can be appropriately performed for an inappropriate expense application.

Further, according to the present embodiment, by using the credit card detailed information of the applicant, an inspection can be appropriately performed for an inappropriate expense application.

Further, according to the present embodiment, by using the business vehicle position information of the applicant, an inspection can be appropriately performed for an inappropriate expense application.

Further, according to the present embodiment, by using the information of the card owned by the applicant, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, by using the company regulation information, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, by using the transportation means information, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, by using the store set, the inspection can be appropriately performed for the inappropriate expense application.

Further, according to the present embodiment, it is possible to appropriately detect inappropriate spending on an inappropriate partner.

Further, according to the present embodiment, it is possible to appropriately detect inappropriate travel expenses, inappropriate transportation expenses, double application, reuse of receipts, illegal accommodation expenses, illegal second party expenses, and the like.

Further, according to the present embodiment, by using the statistical processing result based on two or more expense information, the inspection regarding the inappropriate expense application can be appropriately performed.

Further, according to the present embodiment, it is possible to detect that the specific store is frequently used.

Further, according to the present embodiment, it is possible to detect that the frequency of expenditure corresponding to a specific destination is high.

Further, according to the present embodiment, it is possible to detect that there are many frauds of a specific person.

Further, according to the present embodiment, it is possible to detect that the approver for the expense application does not meet the requirements.

Further, according to the present embodiment, it is possible to detect omission of application for expenses.

Further, according to the present embodiment, an appropriate inspection of the receipt image can be performed.

Further, according to the present embodiment, it is possible to appropriately inspect the items described in the receipt image.

Further, according to the present embodiment, it is possible to inspect the receipt image for deficiencies in the time stamp.

Further, according to the present embodiment, it is possible to inspect an unclear receipt image.

Further, according to the present embodiment, it is possible to inspect inappropriate payment means.

Further, according to the present embodiment, it is possible to inspect the amount of money by handwriting.

The expense inspection device 1 in the present embodiment receives expense information from one or more external expense management devices 4, inspects the expense information, and transmits the inspection result to the expense management device. It is preferable to function as a device. In such a case, the expense inspection device 1 has, for example, the structure of FIG. That is, in such a case, the applicant collective storage unit 112 of the expense inspection device 1, the company regulation information storage unit 113, the card usage expense storage unit 117, the personal information storage unit 1121, the attendance information storage unit 1122, and the entrance / exit information storage unit 1123. , The business activity information storage unit 1124, the card detail information storage unit 1125, and the business vehicle position information storage unit 1126 exist for each organization identifier, for example. That is, the applicant set, company regulation information, card usage expense information, personal information, attendance information, entrance / exit information, sales activity information, card details information, and business vehicle position information are associated with the organization identifier.

Further, the expense information received by the expense inspection device 1 from the expense management device 4 is associated with the organization identifier. Then, when the expense inspection device 1 receives the expense information from the expense management device 4, the inspection unit 131 performs an inspection using the applicant set or the like associated with the organization identifier corresponding to the expense information. Get the test result. Then, the result output unit 141 transmits the inspection result to the expense management device 4. The organization identifier is information that identifies the organization, and is, for example, a company name or a company ID. The organization is, for example, a company, an organization, or a department.

Further, the expense management device 4 that has received the expense information from the applicant terminal 2 transmits the expense information to the expense inspection device 1 in order to inspect the expense information. Then, in response to the transmission of the expense information, the inspection result is received from the expense inspection device 1. Then, the expense management device 4 accumulates the expense information only when the inspection result is valid. Further, the expense management device 4 transmits the inspection result to the applicant terminal 2 when the inspection result is invalid.

With the above system configuration including the expense inspection device 1 and one or more expense management devices 4, the expense inspection device 1 can exert a function as an expense information inspection server. Further, the expense management device 4 can obtain an expense inspection function by using the expense inspection device 1 in addition to the conventional expense information management function.

Further, the expense inspection device 1 can receive the expense information of 2 or more from each expense management device 4 of 1 or 2 or more, and can also perform the above-mentioned statistical processing by including the expense information storage unit 111 for each organization. Become.

The processing in this embodiment may be realized by software. Then, this software may be distributed by software download or the like. Further, this software may be recorded on a recording medium such as a CD-ROM and disseminated. It should be noted that this also applies to other embodiments herein. The software that realizes the expense inspection device in this embodiment is the following program. That is, this program uses a computer with an expense reception unit that receives expense information that specifies the content of an expense application, and external information other than the expense information, or expense information other than the expense information received by the expense reception unit. This is a program for functioning as an inspection unit that inspects fraudulent expenses information received by the expense reception unit and acquires inspection results, and a result output unit that outputs the inspection results acquired by the inspection unit. be.

Further, FIG. 11 shows the appearance of a computer that executes the program described in the present specification to realize the expense inspection device 1 and the like in various embodiments described above. The above-described embodiment can be realized by computer hardware and a computer program executed on the computer hardware. FIG. 11 is an overview view of the computer system 300, and FIG. 12 is a block diagram of the system 300.

In FIG. 11, the computer system 300 includes a computer 301 including a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.

In FIG. 12, in addition to the CD-ROM drive 3012, the computer 301 includes an MPU 3013, a bus 3014 connected to the CD-ROM drive 3012 and the like, a ROM 3015 for storing a program such as a boot-up program, and an MPU 3013. It includes a RAM 3016 that is connected and for temporarily storing instructions of an application program and providing a temporary storage space, and a hard disk 3017 for storing an application program, a system program, and data. Although not shown here, the computer 301 may further include a network card that provides a connection to the LAN.

The program for causing the computer system 300 to execute the functions of the expense inspection device 1 and the like according to the above-described embodiment may be stored in the CD-ROM 3101, inserted into the CD-ROM drive 3012, and further transferred to the hard disk 3017. .. Alternatively, the program may be transmitted to the computer 301 via a network (not shown) and stored in the hard disk 3017. The program is loaded into RAM 3016 at run time. The program may be loaded directly from the CD-ROM3101 or the network.

The program does not necessarily have to include an operating system (OS), a third-party program, or the like that causes the computer 301 to execute the functions of the expense inspection device 1 or the like according to the above-described embodiment. The program only needs to include a part of instructions that call the appropriate function (module) in a controlled manner and obtain the desired result. It is well known how the computer system 300 works, and detailed description thereof will be omitted.

In the above program, in the step of transmitting information and the step of receiving information, the processing performed by the hardware, for example, the processing performed by the modem or the interface card in the transmission step (only performed by the hardware). Processing) is not included.

Further, the number of computers that execute the above program may be singular or plural. That is, centralized processing may be performed, or distributed processing may be performed. That is, the expense inspection device 1 may be a stand-alone device or may be composed of two or more devices.

Further, it goes without saying that in each of the above embodiments, the two or more communication means existing in one device may be physically realized by one medium.

Further, in each of the above embodiments, each process may be realized by centralized processing by a single device, or may be realized by distributed processing by a plurality of devices.

It goes without saying that the present invention is not limited to the above embodiments, and various modifications can be made, and these are also included in the scope of the present invention.

As described above, the expense inspection device according to the present invention has an effect that inspection can be appropriately performed in response to an inappropriate expense application, and is useful as an expense inspection device or the like.

1 Expense inspection device 2 Applicant terminal 3 External server 4 Expense management device 11 Storage unit 12 Reception unit 13 Processing unit 14 Output unit 21 First storage unit 22 First reception unit 23 First processing unit 24 First transmission unit 25 First Receiving unit 26 First output unit 111 Expense information storage unit 112 Applicant set storage unit 113 Company regulation information storage unit 114 Restaurant set storage unit 115 Customs store set storage unit 116 Destination set storage unit 117 Card usage cost storage unit 118 Move Means information storage unit 121 Expense reception unit 131 Inspection department 132 Statistical processing unit 141 Result output unit 1121 Personal information storage unit 1122 Attendance information storage unit 1123 Entrance / exit information storage unit 1124 Sales activity information storage unit 1125 Card details information storage unit 1126 Sales Vehicle position information storage unit 1181 Calculation information storage unit 1182 Map information storage unit 1311 Image inspection means

Claims (23)

An expense reception unit that accepts expense information that identifies the content of an expense application in association with an applicant identifier that identifies the applicant for the expense.
(1) Business activity information, which is information in a daily business report including business day information that specifies the business activity date of the applicant and one or more point identifiers that specify the visited point.
(2) Card statement information, which is information indicating the details of the credit card used by the applicant and includes information on the date of use, which is the date when the credit card was used.
(3) Owned card information regarding the card owned by the applicant,
(4) Overtime time information that specifies the overtime time that allows accommodation, one or two or more approver information that specifies that the person can be an approver of expenses, and one or two or more payment method information that specifies unacceptable payment methods. Company regulation information that has one or more of these information,
(5) A set of sex shops with information on one or more sex shops,
(6) A restaurant set having one or more restaurant information including a store identifier that identifies a restaurant and amount information.
(7) A set of counterparties having one or more counterparty information that identifies the counterparty to whom the applicant should not provide profits.
(8) Regular section information that specifies the regular section of the applicant,
(9) Judgment result of whether or not the expense information received by the expense reception unit and the past expense information associated with the same applicant identifier as the applicant identifier are close to each other.
(10) The number of expense information corresponding to the applicant identifier, including the same usage date information as the usage date information included in the expense information received by the expense reception department.
(11) Frequency of expenditure to the other party identified by the other party identifier, which has the other party identifier possessed by the expense information received by the expense reception unit and is acquired by using the expense information corresponding to the applicant identifier. Entertainment frequency information about,
(12) The number of frauds, which is the examination result for the expense information corresponding to one applicant identifier, and the number of times the examination result corresponding to the fraud is obtained.
(13) One or more character strings acquired from the receipt image included in the expense information received by the expense reception department.
(14) The time stamp of the receipt image included in the expense information received by the expense reception department,
(15) Score, which is the result of character recognition processing for the receipt image included in the expense information received by the expense reception unit.
(16) Approval means information acquired from the receipt image included in the expense information received by the expense reception department,
(17) Using a machine learning algorithm for one or more unclear receipt images and one or more clear receipt images, the receipt image is input to determine whether the receipt image is unclear or clear. Learner to get,
Using one or more of the information, the inspection department that inspects the fraudulent expenses information received by the expense reception department and obtains the inspection results.
An expense inspection device including a result output unit that outputs the inspection result acquired by the inspection unit. The expense information has usage date information that identifies the date on which the expense was used.
The inspection unit is business activity information which is information of a business daily report corresponding to an applicant identifier associated with the expense information received by the expense reception unit, and is the same as the usage date information possessed by the expense information. The expense inspection device according to claim 1, wherein an inspection regarding fraud of the expense information received by the expense reception unit is performed using one or more point identifiers possessed by the business activity information having business day information, and the inspection result is acquired. The expense information has usage date information that identifies the date on which the expense was used.
The inspection unit
It is the card detail information indicating the details of the credit card used by the applicant, and is the card details information corresponding to the applicant identifier associated with the expense information received by the expense reception unit, and the expense information is The expense inspection device according to claim 1, wherein an inspection regarding fraud of the expense information received by the expense reception unit is performed using the card detailed information having the same usage date information as the possession date information, and the inspection result is acquired. The expense information corresponds to an expense type identifier indicating the type of expense,
Further provided with a card usage expense storage unit for storing one or more expense type identifiers scheduled to be paid with the card.
The inspection unit
The expense type identifier corresponding to the expense information received by the expense reception unit corresponds to the owned card information related to the card corresponding to the applicant identifier associated with the expense information received by the expense reception unit. The expense inspection device according to claim 1, which inspects whether or not the card is stored in the card usage expense storage unit as an identifier and acquires the inspection result. The inspection unit
One of overtime time information that specifies overtime hours that allow accommodation, one or more approver information that identifies potential expense approvers, and one or more payment method information that specifies unacceptable payment methods. The expense inspection device according to claim 1, which inspects whether or not the expense information received by the expense reception unit conforms to the company regulations using the company-specified information having the above information, and obtains the inspection result. .. The expense information has a store identifier that identifies the store that has spent the expense.
The inspection unit
The expense according to claim 1, in which the expense information is inspected for fraud using the store identifier of the expense information received by the expense reception unit and the sex shop information of the sex shop set, and the inspection result is obtained. Inspection equipment. The expense information has expenditure amount information that specifies the amount spent.
The inspection unit
The amount information paired with the store identifier of the expense information received by the expense reception unit is acquired from the restaurant set having one or more restaurant information including the store identifier and the amount information, and the amount information and the amount information. The expense inspection device according to claim 11, wherein an inspection regarding fraud of the expense information is performed using the expense amount information possessed by the expense information, and an inspection result is acquired. The expense information has a counterparty identifier that identifies the counterparty who used the expense.
The inspection unit
Using the counterparty identifier of the expense information received by the expense receiving unit and the counterparty set having one or more counterparty information that identifies the counterparty to whom the applicant should not provide profits. The expense inspection device according to claim 1, wherein the expense information is inspected for fraud and the inspection result is acquired. The expense information includes usage date information for specifying the date on which the expense was incurred, and transportation expense information having one or more section information having a departure identifier, an arrival identifier, and a charge information.
The inspection unit
With the information of the regular section that specifies the regular section of the applicant, or the information of the business daily report including the business day information that specifies the day of the business activity of the applicant and one or more point identifiers that specify the visited point. The expense inspection device according to claim 1, wherein the expense information is inspected for fraud related to the claim of the transportation expense by using the business activity information, and the inspection result is acquired. The inspection unit
Any of the past one or more stored in the expense information storage unit associated with the expense information received by the expense reception unit and the same applicant identifier as the applicant identifier associated with the expense information. The expense inspection device according to claim 1, wherein it is determined whether or not the expense information is close enough to satisfy a predetermined condition, and the inspection result is acquired. The expense information is associated with an applicant identifier, includes usage date information that identifies the date on which the expense was used, and has a store identifier that identifies the store that has spent the expense.
The inspection unit
Inspect two or more expense information corresponding to one applicant identifier, determine whether the number of expense information having the same usage date information is large enough to satisfy a predetermined condition, and acquire the inspection result. The expense inspection device according to claim 1. The expense information has a counterparty identifier that identifies the counterparty who has spent the expense.
The other party identified by the other party identifier using the other party identifier included in the two or more expense information corresponding to one applicant identifier, which is two or more expense information stored in the expense information storage unit. Further equipped with a statistical processing unit that acquires entertainment frequency information regarding the frequency of expenses for
The inspection unit
The expense inspection device according to claim 1, which inspects whether or not the entertainment frequency information acquired by the statistical processing unit satisfies a predetermined condition, and acquires the inspection result. The statistical processing unit
The number of frauds, which is the examination result for the expense information corresponding to one applicant identifier and the number of times the examination result corresponding to the fraud is acquired, is acquired in association with the one applicant identifier.
Further equipped
The inspection unit
The expense inspection device according to claim 1, wherein the statistical processing unit inspects whether or not the number of frauds acquired satisfies a predetermined condition, and acquires the inspection result. The expense information has an approver identifier that identifies the approver of the expense.
The inspection unit
The expense inspection device according to claim 1, which inspects whether or not the approver identified by the approver identifier is the approver specified by the approver information possessed by the company-specified information, and obtains the inspection result. .. The inspection unit
One or more attendance information including working day information that identifies the day when the applicant identified by the applicant identifier, and the date and time when the applicant identified by the applicant identifier entered and exited his / her place of work. One or more entry / exit information to specify, one or more business activity information having business day information to specify the business day of the applicant and one or more point identifiers to specify the visited point, used by the applicant. One or more business vehicle position information that specifies the position of the business vehicle, and one or more business vehicle position information associated with the usage date information that specifies the date when the business vehicle was used. , It is determined whether or not the expense information received by the expense reception unit exists in the expense information storage unit in which one or more expense information associated with the one applicant identifier is stored, and the application is made. The expense inspection device according to claim 1, which inspects for omission of application whether or not an application for expenses corresponding to personal information has been made, and acquires the inspection result. The expense information has a receipt image and
The inspection unit
An image inspection means for acquiring one or more character strings from the receipt image, determining whether or not each of the one or more character strings satisfies a predetermined condition, and acquiring an image inspection result is provided.
The result output unit is
The expense inspection device according to claim 1, which outputs the image inspection result. The expense information has a receipt image with a time stamp and
The inspection unit
The time stamp of the receipt image is acquired, the receipt issuance date corresponding to the receipt image is acquired, and the date specified by the time stamp is predetermined with respect to the receipt issuance date. It is equipped with an image inspection means for determining whether or not the above period has passed and acquiring the image inspection result.
The result output unit is
The expense inspection device according to claim 1, which outputs the image inspection result. The inspection unit
Using the character recognition technology, a character recognition score is obtained from the receipt image received by the expense reception department together with one or more character strings, and if the score is low enough to satisfy a predetermined condition, the receipt. The first blurring process to determine that the image is unclear, or the receipt image is unclear by inputting the receipt image with one or more unclear receipt images and one or more clear receipt images by machine learning algorithm. Either of the second blurring process that acquires the judgment result by the machine learning algorithm using the learning device that acquires the judgment result of whether it is clear or clear, and the receipt image received by the expense reception unit. An image inspection means for acquiring an image inspection result by performing the above processing is provided.
The result output unit is
The expense inspection device according to claim 1, which outputs the image inspection result. The expense information has a receipt image and
The inspection unit
The payment method information, which is a character string, is acquired from the receipt image, and it is inspected whether the payment method information is the payment method information possessed by the company-specified information including the payment method information that identifies an unacceptable payment method. And equipped with image inspection means to acquire image inspection results,
The result output unit is
The expense inspection device according to claim 1, which outputs the image inspection result. The expense information has a receipt image and
The inspection unit
Obtain the judgment result of whether or not the amount of money is disguised by inputting the receipt image with one or more receipt images whose amount is disguised by hand and one or more receipt images that are not disguised by machine learning algorithm. Using the learning device to be used and the receipt image received by the expense reception unit, prediction processing is performed by a machine learning algorithm, and whether or not the amount of money described in the receipt image is disguised by hand. Equipped with an image inspection means to acquire the indicated image inspection result,
The result output unit is
The expense inspection device according to claim 1, which outputs the image inspection result. The expense information corresponds to an organization identifier that identifies the organization.
The expense reception department
The expense information associated with the organization identifier is received from the expense management device, and the expense information is received.
The inspection unit uses the expense information associated with the organization identifier to inspect the expense information received by the expense reception unit for fraud, and obtains the inspection result.
The result output unit is
The expense inspection device according to any one of claims 1 to 20, wherein the inspection result acquired by the inspection unit is transmitted to the expense management device. It is an expense inspection method using a computer, which is realized by an expense reception unit, an inspection unit, and a result output unit.
An expense reception step in which the expense reception unit receives expense information that identifies the content of an expense application in association with an applicant identifier that identifies an applicant for the expense.
The inspection unit
(1) Business activity information, which is information in a daily business report including business day information that specifies the business activity date of the applicant and one or more point identifiers that specify the visited point.
(2) Card statement information, which is information indicating the details of the credit card used by the applicant and includes information on the date of use, which is the date when the credit card was used.
(3) Owned card information regarding the card owned by the applicant,
(4) Overtime time information that specifies the overtime time that allows accommodation, one or two or more approver information that specifies that the person can be an approver of expenses, and one or two or more payment method information that specifies unacceptable payment methods. Company regulation information that has one or more of these information,
(5) A set of sex shops with information on one or more sex shops,
(6) A restaurant set having one or more restaurant information including a store identifier that identifies a restaurant and amount information.
(7) A set of counterparties having one or more counterparty information that identifies the counterparty to whom the applicant should not provide profits.
(8) Regular section information that specifies the regular section of the applicant,
(9) Judgment result of whether or not the expense information received by the expense reception unit and the past expense information associated with the same applicant identifier as the applicant identifier are close to each other.
(10) The number of expense information corresponding to the applicant identifier, including the same usage date information as the usage date information included in the expense information received by the expense reception department.
(11) Frequency of expenditure to the other party identified by the other party identifier, which has the other party identifier possessed by the expense information received by the expense reception unit and is acquired by using the expense information corresponding to the applicant identifier. Entertainment frequency information about,
(12) The number of frauds, which is the examination result for the expense information corresponding to one applicant identifier, and the number of times the examination result corresponding to the fraud is obtained.
(13) One or more character strings acquired from the receipt image included in the expense information received by the expense reception department.
(14) The time stamp of the receipt image included in the expense information received by the expense reception department,
(15) Score, which is the result of character recognition processing for the receipt image included in the expense information received by the expense reception unit.
(16) Approval means information acquired from the receipt image included in the expense information received by the expense reception department,
(17) Using a machine learning algorithm for one or more unclear receipt images and one or more clear receipt images, the receipt image is input to determine whether the receipt image is unclear or clear. Learner to get,
Using one or more of the information, the inspection step of inspecting the fraudulent expense information received by the expense reception department and acquiring the inspection result, and
An expense inspection method in which the result output unit includes a result output step for outputting the inspection result acquired in the inspection step. Computer,
An expense reception unit that accepts expense information that identifies the content of an expense application in association with an applicant identifier that identifies the applicant for the expense.
(1) Business activity information, which is information in a daily business report including business day information that specifies the business activity date of the applicant and one or more point identifiers that specify the visited point.
(2) Card statement information, which is information indicating the details of the credit card used by the applicant and includes information on the date of use, which is the date when the credit card was used.
(3) Owned card information regarding the card owned by the applicant,
(4) Overtime time information that specifies the overtime time that allows accommodation, one or two or more approver information that specifies that the person can be an approver of expenses, and one or two or more payment method information that specifies unacceptable payment methods. Company regulation information that has one or more of these information,
(5) A set of sex shops with information on one or more sex shops,
(6) A restaurant set having one or more restaurant information including a store identifier that identifies a restaurant and amount information.
(7) A set of counterparties having one or more counterparty information that identifies the counterparty to whom the applicant should not provide profits.
(8) Regular section information that specifies the regular section of the applicant,
(9) Judgment result of whether or not the expense information received by the expense reception unit and the past expense information associated with the same applicant identifier as the applicant identifier are close to each other.
(10) The number of expense information corresponding to the applicant identifier, including the same usage date information as the usage date information included in the expense information received by the expense reception department.
(11) Frequency of expenditure to the other party identified by the other party identifier, which has the other party identifier possessed by the expense information received by the expense reception unit and is acquired by using the expense information corresponding to the applicant identifier. Entertainment frequency information about,
(12) The number of frauds, which is the examination result for the expense information corresponding to one applicant identifier, and the number of times the examination result corresponding to the fraud is obtained.
(13) One or more character strings acquired from the receipt image included in the expense information received by the expense reception department.
(14) The time stamp of the receipt image included in the expense information received by the expense reception department,
(15) Score, which is the result of character recognition processing for the receipt image included in the expense information received by the expense reception unit.
(16) Approval means information acquired from the receipt image included in the expense information received by the expense reception department,
(17) Using a machine learning algorithm for one or more unclear receipt images and one or more clear receipt images, the receipt image is input to determine whether the receipt image is unclear or clear. Learner to get,
Using one or more of the information, the inspection department that inspects the fraudulent expenses information received by the expense reception department and obtains the inspection results.
A program for functioning as a result output unit that outputs the inspection result acquired by the inspection unit.

Images