Accounting information processing system, accounting information processing method, and accounting information processing program
The system uses AI-OCR and machine learning to automate the determination and calculation of prepaid expenses, addressing the complexity and error-prone nature of future billing in existing systems, enabling efficient and accurate accounting.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- FAST ACCOUNTING INC
- Filing Date
- 2024-11-07
- Publication Date
- 2026-05-19
AI Technical Summary
Calculating and inputting prepaid expenses from invoices is complicated and prone to errors due to the future billing nature of such expenses, which existing systems struggle to handle efficiently.
An accounting information processing system utilizing AI-OCR and machine learning to extract and determine prepaid expenses based on billing periods and dates, coupled with an expense calculation unit to allocate amounts to specific accounting periods.
Facilitates easy and accurate calculation of prepaid expenses, reducing user burden and ensuring precise accounting operations.
Smart Images
Figure 2026082203000001_ABST
Abstract
Description
Technical Field
[0004]
[0001] The present invention relates to an accounting information processing system, an accounting information processing method, and an accounting information processing program.
Background Art
[0002] When an invoice arrives at a company, an accounting staff or the like inputs information such as the name of the invoice issuer, the amount billed, the billing date, the payment due date, etc. described in the invoice and the accounting items into an accounting information processing system. In order to streamline the input work, a technique has also been proposed to automatically input invoice information by applying OCR to the invoice image (Non-Patent Document 1).
Prior Art Documents
Non-Patent Documents
[0003]
Non-Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] When the invoice describes a fee for a service to be received in the future, the accounting staff inputs the accounting item of prepaid expenses for the fee for the service to be received in the future. However, the work of calculating prepaid expenses from the description of the invoice is complicated and prone to errors.
[0005] An object of the present invention is to provide an accounting information processing system, an accounting information processing method, and an accounting information processing program that can easily and reliably perform input processing related to prepaid expenses.
Means for Solving the Problems
[0007] Furthermore, the system may include a machine learning model that is trained to output the text information of a document when the image of the document is input, and a document text information output unit that outputs the text information of the document when the image of the document is input to the machine learning model, and the document text information storage unit may store the text information of the document output by the document text information output unit.
[0008] Furthermore, the system may include an accounting period storage unit that stores information about the accounting period of the organization to which the user belongs, the document text information includes information about the invoice amount, and an expense calculation unit that calculates the expense amount for each accounting period based on the accounting period information and the invoice amount information stored in the accounting period storage unit, and the output unit that outputs the expense amount for each accounting period calculated by the expense calculation unit.
[0009] An accounting information processing method according to an embodiment of the present invention includes: a document text information storage step, which is performed by a computer to store document text information including information about the billing period and the billing date described in the document; a prepaid expense determination step, which determines whether the claim relating to the document is a claim relating to prepaid expenses based on the information about the billing period and the billing date; and an output step, which outputs the determination result of the prepaid expense determination unit.
[0010] An accounting information processing program according to an embodiment of the present invention is a program that causes a computer to execute a document text information storage step, which stores document text information including information about the billing period and the billing date described in the document; a prepaid expense determination step, which determines whether the claim relating to the document is a claim relating to prepaid expenses based on the information about the billing period and the billing date; and an output step, which outputs the determination result of the prepaid expense determination unit. [Effects of the Invention]
[0011] According to the present invention, it is possible to provide an accounting information processing system, an accounting information processing method, and an accounting information processing program that enable easy and reliable input of prepaid expenses. [Brief explanation of the drawing]
[0012] [Figure 1] This figure shows the configuration of an accounting information processing system according to an embodiment of the present invention. [Figure 2] This diagram shows the computer configuration used in this accounting information processing system. [Figure 3] This diagram shows a flowchart illustrating an example of information processing performed by this accounting information processing system. [Figure 4] This is a diagram showing an invoice image from Example 1. [Figure 5] This figure shows the text information recognized by AI-OCR in Example 1. [Figure 6] This figure shows the results of the journal entries made by the accounting information processing system in Example 1. [Figure 7] This figure shows the accounting period of Company A. [Figure 8] This chart shows the amount of expenses recorded for each accounting period. [Modes for carrying out the invention]
[0013] The following describes an accounting information processing system according to an embodiment of the present invention. Figure 1 is a diagram showing the configuration of this system.
[0014] This system includes a voucher text information acquisition unit 11 that applies character recognition technology to a voucher image to obtain voucher text information. The voucher text information includes information regarding the billing period and information regarding the billing date. Based on the information regarding the billing period and the information regarding the billing date, a prepayment expense determination unit 12 determines whether the claim regarding the voucher image is a claim regarding prepayment expenses, and an output unit 13 outputs the determination result of the prepayment expense determination unit 12.
[0015] In addition, this system includes an accounting period reception unit 14 that receives information regarding the accounting period of the organization to which the user belongs. The voucher text information includes information regarding the claim amount. Based on the information regarding the accounting period received by the accounting period reception unit 14 and the information regarding the claim amount, a claim amount calculation unit 15 calculates the amount of expenses recorded for each accounting period. The output unit 13 outputs the amount of expenses recorded for each accounting period calculated by the claim amount calculation unit 15.
[0016] In this specification, a voucher refers to a document or file that serves as evidence of the transaction details. Specifically, it includes, but is not limited to, quotations, order forms (including purchase orders), delivery notes, inspection reports, invoices, etc. An example of an invoice is shown in FIG. 4.
[0017] In this specification, text information refers to information composed of characters, numbers, symbols, etc., and includes text data and text files.
[0018] Figure 2 is a diagram showing the configuration of a computer 20 used for information processing in this system. The storage unit 21 is a device for storing information, and stores computer programs, data, etc. The storage unit 21 is composed of a hard disk, SSD, semiconductor memory, etc. The arithmetic unit 22 is a device that performs arithmetic operations and executes programs. The arithmetic unit 22 functions as the prepayment expense determination unit 12, the expense calculation amount calculation unit 15, etc. It also functions as a control unit that controls devices such as the storage unit 21, the input unit 23, and the output unit 24. The input unit 23 is a device for a user to input information into the computer 20, and is composed of a mouse, keyboard, camera, microphone, etc. The output unit 24 is a device for outputting the information read from the storage unit 21 in a form recognizable by the user, and is composed of a display (display unit), speaker, etc.
[0019] Figure 3 is a diagram showing a flowchart of an example of information processing by this system. First, apply AI-OCR to the evidentiary image such as the invoice, and sort and extract the text information recorded in the invoice image, etc. (ST1). AI-OCR (Optical Character Recognition / Reader. Optical character recognition) not only extracts the text information recorded in the image, but also extracts the text information corresponding to predetermined items such as the claim amount, claim date, claim period (claim target period), and invoice issuer name. Specifically, when an invoice image is input, the invoice image is input to a learning model that has been machine-learned so as to output the text information corresponding to predetermined items such as the claim amount, claim date, claim period, and invoice issuer name, and the text information corresponding to the predetermined items is extracted. Figure 5 is an example of the text information of the invoice image extracted by AI-OCR.
[0020] Next, based on the relationship between the claim date and the claim period, determine that the expenses related to the claim period before the previous month of the claim date are expenses, and determine that the expenses related to the claim period after the month to which the claim date belongs are prepayment expenses (ST2). Then, based on the determination result, the account titles for each billing period are derived, and these account titles for each billing period are displayed on the screen and output to the user (ST5). Figure 6 is an example of the output screen.
[0021] Furthermore, the user enters the accounting period of the organization to which they belong into this system (ST3). Figure 7 shows the accounting period of Company A. Company A has a December fiscal year end and its accounting period spans from the first quarter to the fourth quarter.
[0022] Then, the invoice amount is allocated to each accounting period, the expense to be recorded for each accounting period is calculated (ST4), and displayed on the screen (ST5). Figure 8, item 81, is an example of a display screen showing the expense amounts for each accounting period. [Examples]
[0023] The operation of the accounting information processing system according to Embodiment 1 of the present invention will be described below.
[0024] Figure 4 shows a PDF invoice image 41 issued by ABC Systems Co., Ltd. to Company A and sent as an email attachment. At Company A, the invoice image 41 is uploaded to the accounting information processing system and input into an LLM (Large Language Model). A language model is a learning model that has been trained with machine learning data to output a response in accordance with an output instruction in natural language. A language model that has been trained with a large amount of training data is called an LLM (Large Language Model). In this specification, an LLM is a language model that has been trained with 1 GB or more of training data. There are LLMs that can process image data and those that cannot, but in this embodiment, an LLM that can process image data is used.
[0025] When a standard template output instruction, "Output the invoice amount, consumption tax amount, invoice date, payment due date, invoice period, invoice issuer, and account title recorded in the attached invoice image," is input to the LLM (which functions here as a document text information acquisition unit) capable of processing image data, the text information corresponding to these items recorded in the invoice image 41 in Figure 4 is output. For example, from the text "A System License Fee (August 2023 - July 2025)" recorded in section 42 of the invoice image 41, the LLM determines that the invoice period is "August 1, 2024 - July 31, 2027." The text information corresponding to the predetermined items extracted from the invoice image 41 in this way is shown in Figure 5.
[0026] The LLM (Prepaid Expense Determination Unit) then compares the invoice date and the invoice period to determine whether the account is an expense or a prepaid expense. In Figure 5, since the "Invoice Date" is set to "April 2, 2024," expenses related to the invoice period from April 2024 onwards (April 1, 2024 to July 31, 2026) are determined to be prepaid expenses. Expenses related to the invoice period prior to the month in which the invoice date falls (August 1, 2024 to July 31, 2025) are determined to be expenses, specifically communication expenses, rather than prepaid expenses.
[0027] Next, the draft journal entry data, which is the output of LLM, is displayed on the screen and shown to the user. Figure 6 shows the display screen 61 of the journal entry data output by LLM.
[0028] The accounting period storage unit 14 stores information about the accounting period of the organization to which the user belongs. Figure 7 shows the accounting period of Company A. Company A's fiscal year runs from January 1st to December 31st, with December being its closing month. The accounting period consists of four periods: the first quarter (January 1st to March 31st), the second quarter (April 1st to June 30th), the third quarter (July 1st to September 30th), and the fourth quarter (October 1st to December 31st). The closing period and the length of the unit accounting period vary depending on the organization. Note that "1Q" in Figure 7 is an abbreviation of "1st quarter" and means "first quarter."
[0029] The expense calculation unit of this system calculates the expense amount for each accounting period based on the accounting period information stored in the accounting period storage unit.
[0030] Figure 8 shows the display screen for the expense calculation amount for each accounting period calculated by the expense calculation unit. For each period from 2023 / 4Q to 2025 / 2Q, 1,500,000 yen, which is the invoice amount for the three-month billing period, is recorded as an expense for that accounting period. For 2023 / 3Q, since the billing period is two months, August and September 2023, 1,000,000 yen, which is the invoice amount for two months, is recorded as an expense for 2023 / 3Q. For 2025 / 3Q, since the billing period is one month, July 2025, 500,000 yen, which is the invoice amount for one month, is recorded as an expense for 2025 / 3Q. In this way, the invoice amount for each accounting period is calculated from the billing period, and that amount is recorded as an expense for that accounting period. The display method may be changed, for example, by color-coding the backgrounds of the prepaid expense period and the expense period, so that the periods of prepaid expenses and expenses can be understood intuitively, or the words "prepaid expenses" and "expenses" may be displayed on the screen.
[0031] According to this embodiment, since expenses or prepaid expenses are calculated automatically, the burden on users of determining and calculating prepaid expenses, which was previously complicated, is reduced, and accounting operations can be performed accurately.
[0032] Furthermore, since the amount of expenses to be recorded for each accounting period is automatically calculated based on the fiscal year end and accounting period of the organization to which the user belongs, the burden of calculating the amount of expenses to be recorded for each accounting period, which was previously complicated, is reduced, and accounting operations can be performed accurately.
[0033] The embodiments of the present invention are not limited to the embodiments and examples described above, and the above embodiments and examples can be modified in various ways within the scope of the claims. For example, in Embodiment 1, a language model capable of processing image data was used to extract invoice text information such as the invoice amount, invoice date, and invoice issuer from an invoice image. However, invoice text information may also be extracted from an invoice image using a learning model other than a language model that has been trained to output invoice text information when an invoice image is input. [Explanation of Symbols]
[0034] 11. Document Text Information Storage Unit 12 Prepaid Expense Determination Unit 13 Output section 14 Accounting Period Storage Unit 20 Computers 21 Memory section 22. Calculation Unit (Control Unit) 23 Input section 24 Output section 25 Communications Department 41 Invoice 42. System License Fee (August 2023 - July 2025) 61 Journal Entry Preview Screen 81 Expense calculation amount display screen
Claims
1. It is equipped with a document text information storage unit that stores the document text information written on the document, The aforementioned document text information includes information regarding the billing period and information regarding the billing date, A prepaid expense determination unit determines whether the claim relating to the aforementioned document is a claim relating to prepaid expenses, based on the information relating to the aforementioned billing period and the information relating to the aforementioned billing date. An accounting information processing system comprising: an output unit that outputs the determination result of the prepaid expense determination unit.
2. A machine learning model that takes an image of the document as input and outputs the text information of the document recorded in that image, The system includes a document text information output unit that outputs document text information described in the document by inputting the document image of the document into the learning model, The aforementioned document text information storage unit stores the document text information output by the aforementioned document text information output unit. The accounting information processing system according to claim 1.
3. It includes an accounting period storage unit that stores information about the accounting period of the organization to which the user belongs, The aforementioned document text information includes information regarding the invoice amount, The system includes an expense calculation unit that calculates the expense amount for each accounting period based on the accounting period information and billing amount information stored in the accounting period storage unit, The accounting information processing system according to claim 1, wherein the output unit outputs the expense amounts for each accounting period calculated by the expense calculation unit.
4. Executed by a computer, A document text information storage step that stores document text information including information about the billing period and billing date stated on the document, A prepaid expense determination step, which determines whether the claim relating to the aforementioned document is a claim relating to prepaid expenses, based on the information relating to the aforementioned billing period and the information relating to the aforementioned billing date, An output step which outputs the determination result of the prepaid expense determination unit, An accounting information processing method having the following characteristics.
5. On the computer, A document text information storage step that stores document text information including information about the billing period and billing date stated on the document, A prepaid expense determination step, which determines whether the claim relating to the aforementioned document is a claim relating to prepaid expenses, based on the information relating to the aforementioned billing period and the information relating to the aforementioned billing date, An output step which outputs the determination result of the prepaid expense determination unit, An accounting information processing program to execute this process.