Information processing system, information processing method, and program

Abstract

To provide an information processing system and the like that can grasp the tendency of numerical values and the cause of the tendency as well.SOLUTION: According to one aspect of the present invention, an information processing system comprising one or more processors is provided, wherein the processors acquire, in an acquisition step, a numerical value data group obtained by collecting numerical values related to an organization for a first organization and a second organization, execute, in a generation step, generation processing of, on the basis of the acquired numerical value data groups of the first organization and the second organization, causing a sentence generation AI to generate a summary sentence that includes the result of comparison of at least one of the tendency of numerical values indicated by the numerical value data group and the cause of the tendency between the first organization and the second organization; and perform, in a display step, display processing for causing display means to display the generated summary sentence.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing system, an information processing method, and a program. [Background technology]

[0002] Patent document 1 discloses a technology that classifies financial data into text data and numerical data according to the creation details set for each type of corporate financial document, classifies the numerical data into common data that can be used commonly across types of corporate financial documents and unique data for each type, determines an output format based on the creation details of the corporate financial document specified as the target of creation, extracts the relevant common data and unique data from the numerical data according to the determined output format, and identifies the relevant text data to output information on the specified corporate financial document. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-222385 A Summary of the Invention [Problem to be solved by the invention]

[0004] Even when a comment on a numerical trend is generated from a group of numerical data such as financial data, it is not possible to understand from the comment what factors caused the numerical trend.

[0005] In view of the above circumstances, the present invention provides an information processing system and the like that can grasp not only the tendency of numerical values ​​but also the factors behind the tendency. [Means for solving the problem]

[0006] According to one aspect of the present invention, there is provided an information processing system having one or more processors, in which the processor performs an acquisition step of acquiring a numerical data group collecting numerical values ​​related to a first organization and a second organization, a generation step of executing a generation process in which a sentence generation AI generates a summary sentence including a result of comparing at least one of the numerical trend indicated by the numerical data group and the factor of the trend for the first organization and the second organization based on the acquired numerical data group for each of the first organization and the second organization, and a display step of executing a display process in which the generated summary sentence is displayed on a display means.

[0007] According to this embodiment, in addition to the tendency of the numerical values, the factors behind the tendency can be understood. [Brief description of the drawings]

[0008] [Figure 1] 1 is a diagram showing an example of the overall configuration of a summary creation system 1. FIG. [Diagram 2] 2 is a diagram illustrating an example of a hardware configuration of a server device 10. FIG. [Diagram 3] FIG. 2 is a diagram illustrating an example of a hardware configuration of a user terminal 20. [Figure 4] FIG. 11 is an activity diagram showing an example of a summary creation process. [Diagram 5] FIG. 13 is a diagram showing an example of a displayed summary creation screen. [Figure 6] FIG. 13 is a diagram showing an example of a screen for instructing the creation of a financial summary. [Figure 7] FIG. 13 is a diagram showing an example of a displayed financial summary. [Figure 8] FIG. 13 is a diagram showing an example of a displayed journal entry summary creation screen. [Figure 9] FIG. 13 is a diagram showing an example of a journal entry summary creation instruction screen. [Figure 10] FIG. 13 is a diagram showing an example of a displayed journal entry summary. [Figure 11] FIG. 13 is a diagram showing another example of a displayed journal entry summary. [Figure 12]It is a figure showing an example of a screen for instructing the creation of a final summary. [Figure 13] It is a figure showing an example of the displayed final summary. [Figure 14] It is a figure showing an example of the display screen of the displayed final summary. [Figure 15] It is a figure showing an example of the displayed financial summary. [Figure 16] It is a figure showing an example of the displayed journal summary. [Figure 17] It is a figure showing another example of the display screen of the displayed final summary. [Figure 18] It is a figure showing an example of the displayed extraction location. [Figure 19] It is a figure showing another example of the display screen of the displayed final summary. [Figure 20] It is a figure showing an example of the displayed extraction location. [Figure 21] It is a figure showing another example of the displayed summary creation screen. [Figure 22] It is a figure showing another example of the screen for instructing the creation of a financial summary. [Diagram 23] It is a figure showing an example of the displayed financial comparison summary. [Figure 24] It is a figure showing an example of the displayed journal comparison summary. [Diagram 25] It is a figure showing an example of the displayed final comparison summary. [Figure 26] It is a figure showing an example of a screen for selecting a target organization. [Figure 27] It is a figure showing another example of a screen for selecting a target organization.

Mode for Carrying Out the Invention

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Various characteristic matters shown in the embodiments described below can be combined with each other.

[0010] Incidentally, a program for realizing the software appearing in one embodiment may be provided as a non-transitory computer-readable recording medium, or may be provided so as to be downloadable from an external server, or may be provided so that the program is launched on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] In addition, in various information processing according to an embodiment, an input and an output according to the input can be realized. Here, as long as an output is obtained as a result of the input, the form of information referenced in such information processing (hereinafter referred to as reference information) is not limited. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a judgment formula such as a regression formula constructed by a statistical method), or may be a trained model that has previously trained the correlation between the input and the output, or may be a large-scale language model that can output a desired result by inputting a prompt.

[0012] In one embodiment, the term "unit" may include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In one embodiment, various information is handled, and this information is represented by, for example, physical values ​​of signal values ​​representing voltage and current, high and low signal values ​​as a binary bit collection consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculation can be performed on the circuit in the broad sense.

[0013] Furthermore, a circuit in the broad sense is a circuit realized by at least appropriately combining a circuit, circuitry, a processor, and a memory. The processor may be a general-purpose processor or a dedicated circuit. In other words, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0014] <Embodiment 1> 1. System Configuration The system configuration according to the first embodiment will be described below. Fig. 1 is a diagram showing an example of the overall configuration of a summary creation system 1. Fig. 1 shows an overview of each device provided in the summary creation system 1 and users who use those devices. Each overview will be explained from time to time with reference to other figures. The summary creation system 1 is an information processing system that executes a summary creation process that analyzes numerical trends and the factors that caused the trends to appear from a group of numerical data, and creates a document (hereinafter referred to as a "summary") that summarizes the main points of the analysis results.

[0015] The summary creation system 1 includes a communication line 2, a server device 10, and a user terminal 20. The communication line 2 is not particularly limited, but may be configured, for example, by the Internet network. The communication line 2 may also include a local area network, a mobile communication network, and a VPN (Virtual Private Network), etc. The communication line 2 mediates data exchange between devices connected to the line. In the example of FIG. 1, the server device 10 is connected to the communication line 2 by wire, and the user terminal 20 is connected wirelessly. The connection of each device to the communication line 2 may be wired or wireless.

[0016] The server device 10 is an information processing device that executes summary creation processing. The server device 10 stores a journal entry database DB1 and a financial database DB2. The journal entry database DB1 stores journal entry data, i.e., data that is organized to be useful for creating ledgers and financial statements, including information on transactions and economic activities. The financial database DB2 stores financial data organized based on the journal entry data, i.e., data that shows information on the financial status and economic activities of a company or organization (including financial statements such as balance sheets and profit and loss statements).

[0017] The server device 10 also includes a summary creation AI module 110, a trend determination AI module 120, and a factor determination AI module 130 (hereinafter, when there is no need to distinguish between them, they are referred to as "AI modules 100"). The AI ​​modules 100 are modules that are adjusted (tuned) to realize predetermined functions using AI (Artificial Intelligence) technology.

[0018] The AI ​​module 100 has a natural language processing model with improved accuracy through machine learning using a large-scale data set called LLM (Large Language Models), and realizes a sentence generation function capable of generating natural sentences. The summary creation AI module 110 is adjusted to realize a summary creation function that creates a summary that summarizes the numerical trend and the factors of the trend when the numerical trend and the factors of the trend indicated by the numerical data group are input.

[0019] The trend determination AI module 120 is adjusted to realize not only a text generation function but also a trend analysis function that, when a group of numerical data is input, analyzes the trend of the numerical values ​​indicated by the group of numerical data. The factor determination AI module 130 is adjusted to realize not only a text generation function but also a factor analysis function that, when a group of numerical data and the trend of the numerical values ​​indicated by the group of numerical data are input, analyzes the parts of the group of numerical data that are the factors of the trend. These analysis functions are realized, for example, by performing learning based on a group of data related to a statistical analysis method (such as text explaining the statistical analysis method and sample data).

[0020] The user terminal 20 is a terminal used by a user, such as a smartphone, a tablet terminal, or a personal computer. The user terminal 20 executes display of images in the summary creation process and acceptance of operations by the user. The server device 10 executes a display process for displaying images on the user terminal 20 and an authentication process for authenticating the user of the user terminal 20.

[0021] The server device 10 performs, for example, processing such as generating and transmitting an HTML (Hyper Text Markup Language) file as display processing, and displays a web page showing a system screen on the user terminal 20. Note that an application program for using the summary creation system 1 may be installed in the user terminal 20, and the server device 10 may perform processing such as generating and transmitting display data in the application as display processing.

[0022] 2. Hardware Configuration The hardware configuration according to the first embodiment will be described below. 2 is a diagram showing an example of a hardware configuration of server device 10. Server device 10 includes a control unit 11, a storage unit 12, a communication unit 13, and a bus 14. Bus 14 electrically connects each unit included in server device 10.

[0023] (Control unit 11) The control unit 11 has at least one processor. The at least one processor may be configured, for example, by a central processing unit (CPU), a micro processing unit (MPU), a graphics processing unit (GPU), one or more integrated circuits, one or more discrete circuits, or a combination of these (not shown).

[0024] The control unit 11 is a computer that realizes various functions related to the summary creation system 1 by reading out a predetermined program stored in the storage unit 12. That is, information processing by software stored in the storage unit 12 is specifically realized by the control unit 11, which is an example of hardware, and can be executed as each functional unit included in the control unit 11. Note that the control unit 11 is not limited to being single, and may be implemented with multiple control units 11 for each function. Also, a combination of these may be used.

[0025] (Storage unit 12) The storage unit 12 stores various information defined by the above description. This can be implemented, for example, as a storage device such as a solid state drive (SSD) or a hard disk drive (HDD) that stores various programs and the like related to the summary creation system 1 executed by the control unit 11, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the program's calculations. The storage unit 12 stores various programs, variables, etc. related to the summary creation system 1 executed by the control unit 11.

[0026] (Communications Department 13) The communication unit 13 is configured by a communication module. The communication module may be a wireless communication module conforming to standards such as IEEE802.11a / b / g / n / ac / ax, LTE, 5G, and 6G, or may be a wired communication module conforming to standards such as IEEE802.3. The communication unit 13 is configured to be capable of transmitting various electrical signals from the server device 10 to external components. The communication unit 13 is also configured to be capable of receiving various electrical signals from the external components to the server device 10. More preferably, the communication unit 13 has a network communication function, and may be implemented so that various information can be communicated between the server device 10 and an external device via the communication line 2.

[0027] Fig. 3 is a diagram showing an example of a hardware configuration of the user terminal 20. The user terminal 20 includes a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, an output unit 25, and a bus 26. The bus 26 electrically connects the various units included in the user terminal 20. The control unit 21, the storage unit 22, and the communication unit 23 are similar hardware to the control unit 11, the storage unit 12, and the communication unit 13 shown in Fig. 2, although the specifications, model, etc. may be different.

[0028] (Input section 24) The input unit 24 has keys, buttons, a touch screen, a mouse, etc., and receives input from the user. The input unit 24 also has a microphone, and receives voice input from the user.

[0029] (Output section 25) The output unit 25 has a display, a speaker, and the like, and displays visual information generated in a manner that is visible to the user, such as screens, images, icons, text, and the like, on the display surface of the display, and outputs sounds including voice.

[0030] 3. Information Processing The summary creation process described above will be described below as an information process according to the embodiment. In the following description, the server device 10 and the user terminal 20 are described as the subjects of each summary creation process, but the information processes are executed by one or more processors included in the control unit of each device.

[0031] Fig. 4 is an activity diagram showing an example of summary creation processing. The summary creation processing shown in Fig. 4 is started when a user of the user terminal 20 performs an operation to display a summary creation screen, which is a display screen for a summary creation service provided by the server device 10. First, the server device 10 generates screen data showing the summary creation screen at A11 and transmits the generated screen data to the user terminal 20. The user terminal 20 displays the summary creation screen shown by the transmitted screen data at A12.

[0032] Fig. 5 is a diagram showing an example of a displayed summary creation screen. In the example of Fig. 5, the server device 10 displays a summary creation screen C1 on the display surface of the user terminal 20. The summary creation screen C1 displays a character string saying "Please specify the financial data for which a summary is to be created", a target selection field D11, and a decision button B11. The financial data for which a summary is to be created is specified from the financial data stored in the financial database DB2 shown in Fig. 1 (hereinafter referred to as "stored financial data").

[0033] The target selection field D11 is an operation image for selecting financial data to be the subject of summary creation (hereinafter referred to as "target financial data") from the stored financial data. In the target selection field D11, "target company," "financial statement," "account item," "target period," "period unit," and "indicator type" are entered as information for selecting the target financial data. In "target company," a company or organization whose financial data will be analyzed is selected. In "financial statement," a balance sheet or profit and loss statement, etc., containing the financial data to be analyzed is selected.

[0034] In "Account item", the specific account item to be analyzed is selected. In "Target period", the period during which the economic activity indicated by the financial data to be analyzed took place is selected. In "Period unit", the unit of time for analyzing the financial data (monthly or yearly, etc.) is selected. In "Indicator type", the type of indicator of the financial data to be analyzed (current year value or year-on-year comparison, etc.) is selected. Selection in the target selection field D11 may be made by entering a string from the keyboard, or by displaying a pull-down list (a list based on the stored financial data) and allowing the user to make a selection.

[0035] In the example of FIG. 5, among the "accounts receivable" included in the "balance sheet" of "Company A," the "year-on-year change" of the "monthly" accounts receivable for "2021-2022 fiscal year" is selected as the target financial data. When the decision button B11 is operated, the financial data selected in the target selection field D11 is determined as the target financial data. The user terminal 20 accepts, at A21, the input to the target selection field D11 and the operation of the decision button B11 as a selection operation of the target financial data.

[0036] When selecting the target financial data, the user terminal 20 may display a list of the financial data selected in the target selection field D11 so that the selected financial data can be confirmed. Instead of allowing the user to select the target financial data in the target selection field D11, the user terminal 20 may display a list of stored financial data and allow the user to select the target financial data from the list. Next, the user terminal 20 accepts an instruction to create a financial summary at A22. A financial summary is a document summarizing the results of an analysis of the trends in the numerical values ​​indicated by the target financial data. When the target financial data is selected, the user terminal 20 displays an instruction screen for creating a financial summary.

[0037] 6 is a diagram showing an example of a financial summary creation instruction screen. The user terminal 20 displays a financial summary creation instruction screen C2 including a character string "Please instruct the creation of a financial summary", an outline of the target financial data D21, an input field E21 for instructions to create a financial summary, a back button B21, and a creation start button B22. The outline D21 of the target financial data shows information such as the "target company" selected in the target selection field D11 shown in FIG. 5.

[0038] The back button B21 is an operation image for returning to the previous screen (summary creation screen C1 shown in FIG. 5). When returning to the previous screen with the back button, the operations available on that screen can be performed again. For example, when operating the back button B21 to return to the summary creation screen C1, the target financial data can be reselected. In the summary creation system 1, similarly in the screens described below, the operation on the previous screen can be redone by returning to the screen with the back button.

[0039] The financial summary creation instruction input field E21 is an input field for entering instructions to create a financial summary. Instructions to create a financial summary are given by sentences that indicate instructions to the AI, known as prompts. In the example of Figure 6, the following prompt has been entered in the creation instruction input field E21: "The target financial data shows the year-on-year increase / decrease in monthly accounts receivable for Company A. The accounting period for this company is from December to November of the following year. Explain the trends in this data. Please do not explain the factors or reasons why such trends are observed."

[0040] The server device 10 may create a prompt based on information included in the summary D21 of the target financial data. In this case, the server device 10 creates a prompt by, for example, storing a template for a prompt in which unique information is embedded, and embedding the information included in the summary D21 as unique information. In this case, the creation instruction input field E21 does not need to be displayed.

[0041] When the creation start button B22 is operated, the user terminal 20 accepts the input prompt at A22 as an instruction to create a financial summary, and transmits financial creation instruction data indicating the selected target financial data and the text of the creation instruction to the server device 10. When the financial creation instruction data is transmitted, the server device 10 acquires the target financial data indicated by the financial creation instruction data at A23, and inputs the acquired target financial data to the trend determination AI module 120. In the following figures, the summary creation AI module 110 is referred to as the "first AI", the trend determination AI module 120 as the "second AI", and the factor determination AI module 130 as the "third AI".

[0042] Next, in A24, the server device 10 instructs the trend determination AI module 120 (second AI) to output the trend of the numerical values ​​indicated by the target financial data based on the creation instruction indicated by the financial creation instruction data. The trend determination AI module 120 generates and outputs a sentence indicating the numerical value trend based on the input target financial data.

[0043] In A25, the server device 10 acquires the text output from the trend determination AI module 120 as a financial summary of the target financial data. In this manner, the server device 10 generates a financial summary using the trend determination AI module 120. The server device 10 transmits the generated financial summary to the user terminal 20. In A26, the user terminal 20 displays the transmitted financial summary.

[0044] 7 is a diagram showing an example of a displayed financial summary. The user terminal 20 displays a financial summary display screen C3 including a character string "Financial summary has been created," an overview D21 of the target financial data, a financial summary F31, a back button B31, and a journal summary creation button B32. The back button B31 is an operation image for returning to the previous screen (the financial summary creation instruction screen C2 shown in FIG. 6).

[0045] Financial summary F31 shows, in text form, trend 1: "1. Monthly fluctuations: The increase or decrease in accounts receivable varies greatly from month to month. For example, there is a large increase from December in January, an even larger increase in February, and a fairly small increase in March." trend 2: "2. Magnitude: There are months in which the increase or decrease in accounts receivable exceeds 500 million yen (e.g. August, September, October, November)." trend 2: "3. Total compared to the same month of the previous year: The cumulative increase or decrease in accounts receivable from December to November of the following year is... This is the net fluctuation for the year, and shows that accounts receivable increased by this much over the course of 12 months." trend 3:

[0046] Financial summary F31 also shows trend 4, "4. High peak: The year-on-year changes in accounts receivable were the largest in February, August, October, and November, all exceeding 300 million yen," and trend 5, "5. Relative low: Conversely, the year-on-year changes in accounts receivable were relatively small in December, March, June, and July, falling well below 100 million yen." In this way, a financial summary is created from the target financial data.

[0047] After the financial summary is created, a process of creating a journal entry summary is performed. A journal entry summary is a document summarizing the results of an analysis of the journal entries that caused the trend in the numerical values ​​indicated by the financial summary. To proceed to creating a journal entry summary, the user operates the journal entry summary creation button B32. Then, the user terminal 20 first displays a journal entry summary creation screen for creating a journal entry summary.

[0048] 8 is a diagram showing an example of a displayed journal entry summary creation screen. The user terminal 20 displays a journal entry summary creation screen C4 including a character string "Please specify the journal entry data for which a summary is to be created," a target selection field D41, journal entry data D42, a back button B41, and a decision button B42. The journal entry data for which a summary is to be created is specified from the journal entry data stored in the journal entry database DB1 shown in FIG. 1 (hereinafter referred to as "stored journal entry data").

[0049] The target selection field D41 is an operation image for selecting the journal data for which a summary is to be created (hereinafter referred to as "target journal data") from the stored journal data. The target selection field D41 displays "target company," "client," "account item," and "target period" as information for selecting the target journal data. "Target company," "account item," and "target period" display the information selected when creating the financial summary. "Client" selects a company that does business with the target company.

[0050] The "business partner" may be selected by inputting a character string from the keyboard, or by displaying a pull-down list (a list based on stored journal data) and making a selection. In the example of FIG. 8, "Company D" is selected as the "business partner." When a "business partner" is selected, the user terminal 20 displays the journal data between the selected business partner and the target company. In the example of FIG. 8, journal data D42 of transactions between companies A and D is displayed.

[0051] The displayed journal data D42 may be used to create the journal summary, or the range to be used to create the journal summary may be selected. When the decision button B42 is operated, the journal data selected in the target selection field D41 is determined as the target journal data. The user terminal 20 accepts, in A31, the input to the target selection field D41 and the operation of the decision button B42 as the selection operation of the target journal data.

[0052] The method of selecting the target journal data is not limited to this, and for example, a list of stored journal data may be displayed and the target journal data may be selected from the list. Also, multiple business partners may be selected at the same time. When the target journal data is selected as described above, the user terminal 20 displays a screen instructing the creation of a journal summary. The creation of the journal summary is performed by the factor determination AI module 130.

[0053] 9 is a diagram showing an example of a journal summary creation instruction screen. The user terminal 20 displays a journal summary creation instruction screen C5 including a character string "Please instruct the creation of a journal summary.", an outline D51 of the input data to be input to the factor determination AI module 130, a journal summary creation instruction input field E51, a back button B51, and a creation start button B52. The back button B51 is an operation image for returning to the previous screen (the journal summary creation screen C4 shown in FIG. 8).

[0054] In the example of Fig. 9, the input data overview D51 shows the financial summary F31 of company A shown in Fig. 7 and the journal entry data D42 of companies A and D shown in Fig. 8. The journal entry summary creation instruction input field E51 is an input field for inputting instructions for creating a journal entry summary. In the creation instruction input field E51, a prompt representing an instruction for creating a journal entry summary is entered, as in the case of the financial summary.

[0055] In the example of Fig. 9, the prompt "Data 1 is a summary summarizing the trends of accounts receivable for Company A compared to the same month of the previous year for each month. Data 2 is a table summarizing journal entry data regarding transactions between Company A and Company D. Using these two data, explain the factors that caused the trends written in the financial summary from the journal entry data. However, do not explain trends that are not written in the financial summary or factors that cannot be determined from the journal entry data" is input in the creation instruction input field E51. Note that in this case, as in the example described in Fig. 6, the server device 10 may also create a prompt based on information included in the input data summary D51.

[0056] When the creation start button B52 is operated, the user terminal 20 accepts the input prompt as an instruction to create a journal summary in A32, and transmits journal entry creation instruction data indicating the created financial summary, the selected target journal entry data, and the creation instruction text to the server device 10. When the journal entry creation instruction data is transmitted, the server device 10 inputs the financial summary and the target journal entry data indicated by the journal entry creation instruction data to the factor determination AI module 130 (third AI) in A33.

[0057] Next, in A34, the server device 10 instructs the factor determination AI module 130 (third AI) to output the factors of the trend of the numerical values ​​indicated by the target journal data based on the creation instruction indicated by the journal creation instruction data. The factor determination AI module 130 generates and outputs a sentence indicating the factors of the numerical value trend based on the input financial summary and the target journal data.

[0058] The server device 10 acquires the text output from the factor determination AI module 130 as a journal entry summary of the target journal entry data in A35. The server device 10 generates a journal entry summary using the factor determination AI module 130 in this manner. The server device 10 transmits the generated journal entry summary to the user terminal 20. The user terminal 20 displays the transmitted journal entry summary in A36.

[0059] Fig. 10 is a diagram showing an example of a displayed journal entry summary. The user terminal 20 displays a journal entry summary display screen C6 including a character string "Journal entry summary has been created," an overview of input data D51 shown in Fig. 9, a journal entry summary F61, a back button B61, and a final summary creation button B62. The back button B61 is an operation image for returning to the previous screen (the journal entry summary creation instruction screen C5 shown in Fig. 9).

[0060] Journal entry summary F61 indicates factor 1 in the form of "1. Monthly fluctuations: The increase or decrease in accounts receivable varies greatly from month to month. This difference is due to various transactions, such as the difference in bad debt losses, the ToC department, and the ToB department. For example, in January, Company No. 1's accounts receivable increased from the ToC department compared to December, and in June it decreased significantly due to bad debt losses." and factor 2 in the form of "2. Magnitude: The large fluctuations in accounts receivable, especially in August, September, October, and November, are largely due to transactions with the ToC department. The items for these months have high absolute values."

[0061] In addition, journal entry summary F61 states in text that factor 3 is, "3. Total compared to the same month of the previous year: The cumulative change in accounts receivable from December to November can be explained by related transactions in the journal entry data," factor 4 is, "4. High peak: The large year-on-year fluctuations in accounts receivable in February, August, October, and November are likely due to high values ​​in these months, especially for transactions related to the toC division," and factor 5 is, "5. Relative low: The relatively small fluctuations in December, March, June, and July are seen in certain transactions such as those related to unearned revenue in the toB division."

[0062] Factors 1 to 5 for "Monthly Fluctuation," "Magnitude," "Total Compared to Same Month of Last Year," "High Peak," and "Relative Low" represent trends 1 to 5 included in financial summary F31 shown in Fig. 7, that is, factors that caused trends 1 to 5 in the same numerical values ​​for "Monthly Fluctuation," "Magnitude," "Total Compared to Same Month of Last Year," "High Peak," and "Relative Low." Fig. 10 shows a journal entry summary based on journal entry data with company D, but a journal entry summary based on journal entry data with company E will be explained with reference to Fig. 11.

[0063] 11 is a diagram showing another example of a displayed journal entry summary. Journal entry summary F61 for Company E indicates, in the form of text, factor 1, "1. Monthly fluctuation: This can be seen in the fluctuation in the debit amount for each month, with a large increase from December to January and an even larger increase in February. The variation in these figures across different rows for different companies or departments reflects the large changes from month to month." and factor 2, "2. Magnitude: High figures for certain months, such as August to November, indicate periods of high activity. We can see that the debit amounts for these months are over 500 million."

[0064] In addition, journal entry summary F61 indicates factor 3 in the form of, "3. Total compared to the same month last year: The cumulative change can be inferred from the total debit amounts, which matches the total year-on-year change shown in the financial summary," and factor 4 in the form of, "4. High peak: From the journal entry data for February, August, October, and November, we can see that these are the months with the largest year-on-year changes in accounts receivable. These can be observed from the corresponding debit amounts."

[0065] Furthermore, the journal entry summary F61 indicates factor 5 in the form of a sentence: "5. Relatively low value: Similarly, in the journal entry data, the values ​​for December, March, June, and July are low. The values ​​for these months are well below 100 million, which is consistent with the observation of the relatively low level in the financial summary." In this manner, the journal entry summary is created from the financial summary and the target journal entry data. In this embodiment, it is assumed that the journal entry summary is created based on the journal entry data for companies F, G, and H in addition to companies D and E.

[0066] After the journal summary is created, a process of creating a final summary is performed. The final summary is a document summarizing the trends of the values ​​indicated by the financial summary and the factors behind the trends of the values ​​indicated by the journal summary. If the user wishes to proceed to creating a final summary, the user operates the Create Final Summary button B62. Then, the user terminal 20 first displays a final summary creation screen for creating a final summary. The creation of the final summary is performed by the summary creation AI module 110.

[0067] 12 is a diagram showing an example of a final summary creation instruction screen. The user terminal 20 displays a final summary creation instruction screen C7 including a character string "Please instruct the creation of a final summary," an outline D71 of input data to be input to the summary creation AI module 110, a final summary creation instruction input field E71, a back button B71, and a creation start button B72. The back button B71 is an operation image for returning to the previous screen (the journal entry summary display screen C6 shown in FIG. 10 or FIG. 11).

[0068] In the example of Fig. 12, the input data overview D71 shows the financial summary F31 of company A shown in Fig. 7 and the journal entry summary F61 shown in Fig. 10 and Fig. 11. The final summary creation instruction input field E71 is an input field for inputting instructions for creating a final summary. A prompt representing an instruction for creating a final summary is entered in the creation instruction input field E71, as in the case of the financial summary and the journal entry summary.

[0069] In the example of FIG. 12, a prompt is entered in the creation instruction input field E71, stating, "Data 1 is a financial summary summarizing the trend of the increase / decrease in Company A's "accounts receivable" from December 2021 to November 2022 compared to the same month of the previous year. Data 2 is a journal entry summary summarizing the trend of Company A's journal entry data. Journal entry data with the debit account item being accounts receivable was extracted for trading partner companies D, E, F, G, and H, and a journal entry summary for each was created. Using data 1 and 2, explain the factors of the trends described in Company A's accounts receivable summary data based on the journal entry summary data. If the factors cannot be determined from the journal entry summary data, state that you do not know. ". In this case, as in the example described in FIG. 6, the server device 10 may create a prompt based on the information included in the input data summary D51.

[0070] When the creation start button B72 is operated, in A41, the user terminal 20 accepts the creation instruction entered in the creation instruction input field E71 as an instruction to create a final summary, and transmits final creation instruction data indicating the created financial summary, the created journal entry summary, and the text of the creation instruction to the server device 10. When the final creation instruction data is transmitted, in A42, the server device 10 inputs the financial summary and journal entry summary indicated by the final creation instruction data to the summary creation AI module 110 (first AI).

[0071] Next, in A43, the server device 10 instructs the summary creation AI module 110 (first AI) to output a final summary that combines the financial summary and the journal entry summary based on the creation instruction indicated by the final instruction data. The summary creation AI module 110 generates and outputs the final summary based on the input financial summary and journal entry summary.

[0072] In A44, the server device 10 acquires the text output from the summary creation AI module 110 as a final summary of the target financial data and the target journal entry data. In this manner, the server device 10 generates a final summary using the summary creation AI module 110. The server device 10 transmits the generated final summary to the user terminal 20. In A45, the user terminal 20 displays the transmitted final summary.

[0073] Fig. 13 is a diagram showing an example of a displayed final summary. The final summary display screen C8 shown in Fig. 13 displays the character string "Final summary has been created", the outline of the input data D71 shown in Fig. 12, the final summary F81, a back button B81, and a save button B82. The back button B81 is an operation image for returning to the previous screen (the final summary creation instruction screen C7 shown in Fig. 12).

[0074] Final summary F81 presents summary 1, "1. Monthly fluctuations: The accounting entries of several companies show large variations between successive months. The toC and toB divisions contribute to these fluctuations. Not all companies have data for each month, and there may be some companies that do not contribute much to the overall fluctuations."; and summary 2, "2. Magnitude: The high figures seen from August to November, for example, are reflected in the accounting entries of some companies, but not all companies. In addition, there are some companies that show relatively low values ​​without explaining their magnitude."

[0075] In addition, Final Summary F81 states in the text Summary 3 that "3. Year-on-year change total: The cumulative increase or decrease can be tracked through various subaccounts and business partners, but not all companies' accounting data can fully explain this figure. Transactions other than those provided in the accounting data must contribute to the total fluctuation." and Factor 4 that "4. High peaks: The high year-on-year changes in February, August, October, and November are consistent with the accounting data of some companies. Other customers or factors not included in the data provided may be influencing these peaks."

[0076] In addition, Final Summary F81 indicates factor 5 in the following text: "5. Relative Lows: The low values ​​for December, March, June, and July can be traced back to specific accounts and subaccounts in some company data. However, not all company data provide information for these months, so other factors may have contributed to these relative lows."

[0077] The final summary F81 can be modified by a user operation. The user checks the displayed final summary F81, modifies it if necessary, and then operates the save button B82. When the save button B82 is operated, the server device 10 stores the final summary F81. At that time, the server device 10 may store the target financial data, financial summary, target journal data, journal summary, and each prompt used when generating the final summary F81 in association with each other. In this way, the results of the analysis of the numerical data group and the data that became the material for the analysis are stored.

[0078] As described above, the server device 10 executes an acquisition step of acquiring stored journal data as the first numerical data group and acquiring stored financial data as the second numerical data group. The second numerical data group is a data group including numerical values ​​derived from the first numerical data group. For example, the stored financial data includes numerical values ​​of sales and accounts receivable, etc., derived from the stored journal data. A23 and A33 shown in FIG. 4, i.e., A23 in which the server device 10 acquires the target financial data as an example of the second numerical data group and A33 in which the server device 10 acquires the target journal data as the first numerical data group, are examples of an acquisition step.

[0079] The server device 10 executes a generating step of making the first artificial intelligence module generate text indicating the trend of the numbers indicated by the second numerical data group and the factors of the trend indicated by the first numerical data group, based on the first numerical data group and the second numerical data group acquired in the acquiring step. The summary creating AI module 110 is an example of a first artificial intelligence module, and the final summary F81 is an example of text indicating the trend of the numbers and the factors of the trend. A42 and A43, in which the server device 10 makes the summary creating AI module 110 generate the final summary, are examples of the generating step.

[0080] The first artificial intelligence module is an artificial intelligence module that performs learning based on a data group (e.g., LLM) including natural language, and outputs a sentence according to input information. The server device 10 executes a display step of performing a display process for displaying the sentence generated in the generation step on a display means. A44 in which the server device 10 displays the final summary F81 on the user terminal 20 is an example of the display step. According to this embodiment, the displayed final summary not only makes it possible to grasp the trend of the numerical values ​​indicated by the second numerical data group, but also makes it possible to grasp the factors behind the trend.

[0081] The server device 10 also executes a first determination step of making a trend determination on the trend of the numerical values ​​indicated by the second numerical data group, based on the second numerical data group acquired in the acquisition step. In the embodiment, the server device 10 inputs the second numerical data group to a second artificial intelligence module in A25 shown in FIG. 4, and the output by the second artificial intelligence module is the result of the trend determination. The trend determination AI module 120 is an example of a second artificial intelligence module. A23 to A25 shown in FIG. 4 are steps of acquiring a financial summary as a result of the trend determination, and are an example of the first determination step.

[0082] The second artificial intelligence module is an artificial intelligence module that can perform learning based on a data group including natural language and output a sentence showing the result of analyzing the trend of an inputted numerical value, like the trend determination AI module 120. According to this embodiment, the trend determination result can be made easier for people to understand than when the trend determination result is not in a sentence.

[0083] The server device 10 also executes a second determination step of performing factor determination on the factors of the determined numerical value trend based on the acquired first numerical data group. In the embodiment, the server device 10 inputs the trend determination result and the first numerical data group to the third artificial intelligence module in A35 shown in FIG. 4, and the output by the third artificial intelligence module is the result of the factor determination. The factor determination AI module 130 is an example of the third artificial intelligence module. A33 to A35 shown in FIG. 4 are steps of obtaining a journal entry summary as a result of the factor determination, and are an example of the second determination step.

[0084] The third artificial intelligence module is an artificial intelligence module that can perform learning based on a data group including natural language and output a sentence that analyzes the factors of the trend of the input numerical value, like the factor determination AI module 130. According to this embodiment, the result of the factor determination can be made easier for people to understand than when the result of the factor determination is not in the form of a sentence.

[0085] Furthermore, the server device 10 generates a final summary by inputting the trend judgment result and the factor judgment result to the first artificial intelligence module. In the embodiment, the server device 10 inputs the financial summary, which is the trend judgment result, and the journal entry summary, which is the factor judgment result, to the summary creation AI module 110, which is an example of the first artificial intelligence module, at A42. According to this aspect, it is possible to eliminate the need for judgment functions such as trend judgment and factor judgment in the first artificial intelligence module.

[0086] <Modification: Display of financial summary and journal entry summary> In the embodiment, the server device 10 only displays the final summary on the display screen C8 of the final summary shown in Fig. 13, but may display, for example, materials for judging whether the text of the final summary is appropriate or not. Such judgment materials include, for example, the financial summary and the journal summary used to generate the final summary.

[0087] In this case, the server device 10 executes a display step of performing a process for displaying either or both of the trend judgment result and the factor judgment result corresponding to the sentence (final summary) generated in the generation step as a display process. The server device 10 displays a display screen of the final summary that accepts an operation for executing such a display process.

[0088] Fig. 14 is a diagram showing an example of a display screen of the displayed final summary. In the display screen C9 of the final summary shown in Fig. 14, a financial summary display button B91 and a journal entry summary display button B92 are additionally displayed in addition to what is displayed on the display screen C8 shown in Fig. 13. When the financial summary display button B91 is operated, the server device 10 reads and displays the financial summary shown in data 1 of the input data summary D71, that is, the financial summary F31 used when generating the final summary F81.

[0089] Fig. 15 is a diagram showing an example of a displayed financial summary. On the final summary display screen C9 shown in Fig. 15, the financial summary F31 shown in Fig. 7 is displayed instead of the final summary F81, and a final summary display button B93 is displayed instead of the financial summary display button B91. When the final summary display button B93 is operated, the server device 10 displays the final summary F81 alongside the financial summary F31, allowing the two summaries to be compared.

[0090] When the journal entry summary display button B92 is operated on the display screen C9, the server device 10 reads out and displays the journal entry summary represented in data 2 of the input data summary D71, i.e., the journal entry summary F61 used when generating the final summary F81.

[0091] Fig. 16 is a diagram showing an example of a displayed journal summary. In the final summary display screen C9 shown in Fig. 16, the journal summary F61 shown in Fig. 10 and Fig. 11 is displayed instead of the final summary F81, and the final summary display button B93 is displayed instead of the journal summary display button B92. Note that since the journal summary F61 is generated only for transactions with five companies, if it does not fit on the screen, the journal summary F61 for each transaction can be displayed by scrolling.

[0092] When the display final summary button B93 is operated, the server device 10 displays the final summary F81 alongside the journal summary F61, allowing the two summaries to be compared. According to the above embodiment, the user can visually compare the final summary F81 with the financial summary F31 or the journal summary F61, making it easier to find grounds for examining the text of the final summary.

[0093] The server device 10 may also perform, as a display process, a process for displaying either or both of the result of the trend judgment and the result of the factor judgment corresponding to a portion of the displayed final summary designated by the user. The operation method by the user and the information displayed in this case will be described with reference to Figs. 17 and 18.

[0094] Fig. 17 is a diagram showing another example of the display screen of the displayed final summary. In the display screen C9 of the final summary shown in Fig. 17, a selection image G91 for selecting the portion of the description related to "3. Year-on-year comparison total" is displayed in the display screen C9 shown in Fig. 14. The portion selected by the selection image G91 is an example of "a portion of the text of the final summary designated by the user."

[0095] When the financial summary display button B91 is operated in this state, the server device 10 extracts a portion related to the selected portion from the financial summary shown in data 1 of the input data summary D71, i.e., the financial summary F31 used in generating the final summary F81. The server device 10 also extracts a portion corresponding to the selected portion from the journal entry summary shown in data 2 of the input data summary D71, i.e., the journal entry summary F61 used in generating the final summary F81. The server device 10 causes the user terminal 20 to display the extracted portions of the financial summary F31 and the journal entry summary F61 thus extracted.

[0096] Fig. 18 is a diagram showing an example of a displayed extracted portion. In the display screen C9 of the final summary shown in Fig. 18, a character string "A summary corresponding to the final summary has been displayed", a selected portion H91 selected in the final summary F81, an extracted portion H92 described for "3. Total compared to the same month of the previous year" in the financial summary F31 shown in Fig. 7, and an extracted portion H93 described for "3. Total compared to the same month of the previous year" in the journal entry summary F61 of each company and other companies shown in Figs. 10 and 11 are displayed. The extracted portion H93 in the journal entry summary F61 is scrolled to display the extracted portion H93 in the transaction with each company.

[0097] The server device 10 may display the selected portion H91 and the extracted portions H92, H93 of the final summary together as in the example of Fig. 18, but is not limited to this and may display them in order by switching pages. Regardless of the display mode, the final summary F81, financial summary F31, and journal entry summary F61 can be compared for the portion selected by the user. According to this mode, compared to the case where the entire judgment result is always displayed, text that is narrowed down to the portion to be compared is displayed, making it easier to compare the text of the final summary with the judgment result.

[0098] <Modification: Display of financial data and journal data> In the above example, the server device 10 displays the financial summary or the journal summary to examine the text of the final summary, but the present invention is not limited to this and may display the numerical data on which the summaries are based. In this case, the server device 10 performs a process for displaying the numerical data corresponding to the portion of the text of the displayed final summary designated by the user as a display process. The operation method by the user and the information displayed in this case will be described with reference to Figs. 19 and 20.

[0099] Fig. 19 is a diagram showing another example of a display screen of the displayed final summary. In the display screen C10 of the final summary shown in Fig. 19, the input data outline D71 shown in Fig. 12, the final summary F81 shown in Fig. 13, a back button B81, a save button B82, a display summary button B101, and a display data button B102 are displayed. In the display screen C10, a selection image G101 for selecting the entry for "3. Year-on-year comparison total" is displayed.

[0100] When the summary display button B101 is operated in this state, the server device 10 displays the parts related to the selected part from the financial summary F31 and journal entry summary F61 used in generating the final summary F81, as in the example of Fig. 18. Furthermore, when the data display button B102 is operated, the server device 10 extracts the parts corresponding to the selected part from the financial data and journal entry data used in generating the final summary F81. The server device 10 displays the extracted parts of the financial data and the extracted parts of the journal entry data thus extracted on the user terminal 20.

[0101] Fig. 20 is a diagram showing an example of a displayed extracted portion. In the display screen C10 of the final summary shown in Fig. 20, a character string "The data corresponding to the final summary has been displayed", a selected portion H101 of the final summary F81, a financial portion D101 corresponding to the portion H101 of the financial data, a journal entry portion D102 corresponding to the portion H101 of the journal entry data, a back button B103, a save button B104, a financial summary display button B105, and a journal entry summary display button B106 are displayed.

[0102] Since the journal entry locations D102 are extracted only for transactions with five companies, the journal entry locations D102 for each transaction can be displayed by scrolling. When the financial summary display button B105 is operated, the server device 10 displays the financial summary generated based on the financial location D101 (extracted location H92 of the financial summary shown in FIG. 18). When the journal entry summary display button B106 is operated, the server device 10 displays the journal entry summary generated based on the journal entry location D102 (extracted location H93 of the journal entry summary shown in FIG. 18).

[0103] 20, when examining the text of the final summary, the data on which the text is based can be viewed, making it possible to find more specific grounds than when examining the financial summary and journal entry summary. Also, since it is possible to switch between displaying the financial summary and journal entry summary corresponding to a selected portion of the final summary and displaying the financial data and journal entry data corresponding to the same portion, the text of the final summary can be examined from more diverse perspectives than when this switching is not possible.

[0104] <Modification: Reassessment of factors> The stored journal data includes multiple types of journal data, such as journal data showing transactions with company D and journal data showing transactions with company E. In this way, the first numerical data group may include multiple types of numerical data groups. Here, when the text of the final summary generated in the generation step does not satisfy a predetermined condition, the server device 10 may perform factor determination using a numerical data group of a different type from the first numerical data group used to generate the text as the first numerical data group.

[0105] For example, when the degree of certainty of the sentences indicated by the final summary is less than a threshold, the server device 10 determines that the sentences in the final summary do not satisfy the predetermined condition. The degree of certainty of a sentence is expressed, for example, by the ratio of the number of sentences that are not definitive, such as "may," "should," and "possible," to the number of sentences included in the final summary, and the smaller this ratio, the higher the degree of certainty. Note that the predetermined condition is not limited to this. For example, the server device 10 may determine that the predetermined condition is not satisfied when the amount of sentences in the final summary is less than a threshold.

[0106] Also, whether or not the final summary satisfies a predetermined condition may be determined by a human. In this case, the server device 10 displays, for example, a re-determination button on the display screen C8 of the final summary shown in Fig. 13, which indicates that the text of the final summary does not satisfy the predetermined condition and therefore the factor determination is to be performed again. When the re-determination button is operated, the server device 10 determines that the text of the final summary has been determined by a human to not satisfy the predetermined condition, and performs the factor determination again.

[0107] If the final summary F81 does not satisfy the predetermined conditions with the journal data of companies D, E, F, G, and H, the server device 10 further acquires journal data of company I, performs factor determination again, and regenerates the final summary, for example. Also, if the final summary does not satisfy the predetermined conditions even after acquiring the journal data of company I, the server device 10 further acquires journal data of company J, performs factor determination again, and regenerates the final summary. In this way, the server device 10 may repeatedly perform factor determination and regenerate the final summary until the final summary satisfies the predetermined conditions.

[0108] In the above example, the server device 10 acquires additional journal data while leaving the acquired journal data unchanged. However, this is not limited to the above, and for example, the server device 10 may discard a portion of the acquired journal data and acquire new journal data. The server device 10 may also discard a portion of the acquired journal data and perform factor determination again without acquiring new journal data. The journal data to be acquired again may also be determined by a person. According to such an embodiment, the accuracy of factor determination can be improved compared to the case where factor re-determination is not performed.

[0109] <Variation: Feedback on corrections> As described above, the final summary may be revised by the user. In that case, the server device 10 may execute a revision acquisition step of acquiring revisions to the text of the displayed final summary, and may execute a learning step of making the first artificial intelligence module learn the revisions acquired in the revision acquisition step.

[0110] The first artificial intelligence module is, for example, the summary creation AI module 110. As described above, the summary creation AI module 110 is adjusted to realize a summary creation function that, when a financial summary and a journal summary are input, creates a final summary text that summarizes the trends of the numerical values ​​of the target financial data and the factors of the trends. The server device 10 adds a pair of the input financial summary and journal summary and the corrected final summary to the teacher data, and causes the summary creation AI module 110 to learn. By feeding back the correction results to the AI ​​in this way, the generated text can be closer to a text created by a person than when this feedback is not provided.

[0111] Similarly, in the learning step, the server device 10 may cause the second artificial intelligence module to learn the correction content acquired in the correction acquisition step. The second artificial intelligence module is, for example, the trend determination AI module 120. As described above, the trend determination AI module 120 is adjusted to realize a trend analysis function that analyzes the trend of the numerical values ​​indicated by the target financial data when the target financial data is input.

[0112] The server device 10 adds a pair of the input target financial data and the corrected final summary to the teacher data and causes the trend determination AI module 120 to perform learning. If the corrected portion of the final summary is also included in the financial summary used to generate the final summary, by feeding back this correction, the financial summary output by the trend determination AI module 120, i.e., the generated trend determination result, can be made closer to the result of a human determination than if this feedback is not provided.

[0113] In addition, in the learning step, the server device 10 may cause the third artificial intelligence module to learn the correction content acquired in the correction acquisition step. The third artificial intelligence module is, for example, the factor determination AI module 130. As described above, the factor determination AI module 130 is adjusted to realize a factor analysis function that analyzes the factors of the trend of the numerical values ​​indicated by the target journal entry data when the target journal entry data is input.

[0114] The server device 10 adds a pair of the input target journal data and the corrected final summary to the teacher data and causes the factor determination AI module 130 to learn. If the corrected portion of the final summary is also included in the journal summary used to generate the final summary, by feeding back this correction to the AI, the journal summary output by the factor determination AI module 130, i.e., the generated result of the factor determination, can be made closer to the result of a human determination than when this feedback is not provided.

[0115] It should be noted that the corrections fed back to the AI ​​module 100 are not limited to corrections to the final summary. Corrections to the financial summary may be made available, and the corrections to the financial summary may be fed back to the trend determination AI module 120. Also, corrections to the journal entry summary may be made available, and the corrections to the journal entry summary may be fed back to the factor determination AI module 130. In either case, the results of trend determination or factor determination can be made closer to the results of determination by a human, compared to when feedback is not provided.

[0116] <Variation: Configuration of artificial intelligence module> The artificial intelligence modules (summary creation AI module 110, trend determination AI module 120, and factor determination AI module 130) may be internal or external components of the server device 10, or may be internal or external components of the summary creation system 1. Furthermore, the function realized by one artificial intelligence module may be distributed and realized by two or more artificial intelligence modules, or the function realized by two or more artificial intelligence modules may be integrated and realized by one artificial intelligence module.

[0117] For example, the summary creation AI module 110 may be adjusted to realize a trend analysis function and a factor analysis function in addition to the text generation function and summary creation function described above. In this case, the server device 10 can generate text output by the summary creation AI module 110 as a final summary showing the trend of the values ​​indicated by the target financial data and the factors of the trend indicated by the target journal entry data simply by inputting the target financial data and the target journal entry data acquired in the acquisition step to the summary creation AI module 110.

[0118] The server device 10 may generate a final summary using only the summary creation AI module 110 and the trend determination AI module 120 (in this case, the summary creation AI module 110 performs factor determination), or may generate a final summary using only the summary creation AI module 110 and the factor determination AI module 130 (in this case, the summary creation AI module 110 performs trend determination). In this way, the range of functions realized by the AI ​​module 100 may be determined as necessary.

[0119] <Modification: Trend Judgment / Factor Judgment> In the embodiment, the server device 10 performed trend determination and factor determination using an AI module 100 (trend determination AI module 120 and factor determination AI module 130) that was adjusted to realize a sentence generation function, but this is not limited to the above, and trend determination and factor determination may also be performed using an AI module 100 that does not realize a sentence generation function.

[0120] Furthermore, the server device 10 may use an algorithm for determining a trend in a predetermined procedure based on the target financial data, or an algorithm for determining factors in a predetermined procedure based on the trend determination results and the target journal entry data, thereby performing trend determination and factor determination without using the AI ​​module 100. In either case, if the summary creation AI module 110 has performed learning using the trend determination results and factor determination results for each case, it can generate a final summary based on those determination results.

[0121] <Modification: First Numeric Data Group and Second Numeric Data Group> The first and second numerical data groups are not limited to financial data and accounting data. For example, the server device 10 may acquire, in A33, detailed expenditure data such as receipt data in a household accounting application, mail order purchase data, and credit card usage details data as the first numerical data group, and acquire, in A23, monthly expenditure data in which the numerical values ​​indicated by the first numerical data group are summarized by item (fixed cost, variable cost, etc.) as the second numerical data group.

[0122] In this case, the second numerical data group, which is monthly expenditure data, is a data group including numerical values ​​derived from the first numerical data group, which is detailed expenditure data. The period for which the expenditure data is compiled is not limited to one month, but may be any predetermined period such as one day, one week, one quarter, or one half year. The server device 10 then executes the summary creation process shown in Fig. 4, performs trend judgment on the tendency of the numerical values ​​indicated by the monthly expenditure data based on the acquired monthly expenditure data, and generates an expenditure trend summary instead of a financial summary in A25.

[0123] The server device 10 also performs a factor determination on the factors of the determined numerical trend based on the acquired detailed expenditure data, and generates an expenditure detail summary instead of a journal entry summary in A35. The server device 10 then causes the first artificial intelligence module to generate a final summary, which is a text indicating the numerical trend indicated by the monthly expenditure data and the factors of the trend indicated by the detailed expenditure data in A42 and A43, and performs a display process for displaying the generated text on a display means in A44. In this case, too, the displayed final summary allows the user to grasp not only the numerical trend indicated by the second numerical data group, but also the factors of the trend.

[0124] Alternatively, for example, meteorological data such as daily precipitation and temperature may be the first numerical data group, and meteorological information (e.g., monthly precipitation and monthly average temperature) that is a compilation of meteorological data for a predetermined period may be the second numerical data group. Furthermore, traffic data such as daily traffic volume at each observation point and the number of accidents in each region may be the first numerical data group, and traffic information that summarizes the number of accidents in a predetermined period by region (e.g., monthly number of accidents by region) may be the second numerical data group. In short, any data group may be used as long as the second numerical data group is a data group that includes numerical values ​​derived from the first numerical data group.

[0125] <Variation: Comparison between organizations> In the above embodiment, a sentence showing the tendency and factors of the numerical values ​​of one organization, the target company, is generated, but the present invention is not limited to this. A sentence comparing the tendency and factors of the numerical values, or both, of two or more organizations may be generated. In this case, the server device 10 executes an acquisition step of acquiring a numerical data group that collects numerical values ​​related to the organizations for the first organization and the second organization.

[0126] The organization referred to here is, for example, a company, but also includes public institutions such as local governments, associations, non-profit organizations (NPOs), and incorporated foundations. In short, any organization that generates a large amount of numerical data through its organizational activities can be included. For example, the first organization is the company to which the user belongs, and the second organization is another company, etc.

[0127] In a generation step, the server device 10 executes a generation process in which the sentence generation AI generates a summary sentence based on the acquired numerical data groups of the first organization and the second organization. More specifically, the server device 10 causes the sentence generation AI to generate a summary sentence (hereinafter referred to as a "comparison summary sentence") including a result of comparing at least one of the numerical trend indicated by the numerical data groups of the first organization and the second organization and the factor of the trend, for the first organization and the second organization.

[0128] The comparative summary text here may include any of a trend comparison summary that shows a comparison of numerical trends such as a financial summary, a factor comparison summary that shows a comparison of factors of numerical trends such as a journal summary, or a final comparison summary that shows a comparison of those trends and factors together. In addition, when the numerical data group includes a first numerical data group and a second numerical data group as in the embodiment, the generation process is a process of having the AI ​​module 100 (an example of a text generation AI) generate a summary text showing the result of a comparison of at least one of the numerical trend indicated by the second numerical data group and the factor of the trend indicated by the first numerical data group based on the first numerical data group and the second numerical data group included in the acquired numerical data group.

[0129] Then, the server device 10 executes a display step of performing a display process for displaying the generated comparison summary text on the display means. The screen transition until the comparison summary text showing such a comparison result is displayed will be described with reference to FIG. 21 etc.

[0130] Fig. 21 is a diagram showing another example of a displayed summary creation screen. The summary creation screen C11 shown in Fig. 21 displays a character string "Please specify the target organization and financial data for creating a comparative summary", a selection field D111 for a first target organization, a selection field D112 for a second target organization, a selection field D113 for a target data range, and a decision button B111. "Company A" is input as the first target organization in the selection field D111, and "Company B" is input as the second target organization in the selection field D112.

[0131] In the selection field D113, "financial statement" and "account item" etc. are entered as information for selecting target financial data. In the example of Fig. 21, only one selection field is displayed to select a common range for the first target organization and the second target organization, but selection fields may be displayed separately for selecting different ranges for the first target organization and the second target organization. When the decision button B111 is operated, the server device 10 displays a screen instructing the creation of a financial summary.

[0132] Fig. 22 is a diagram showing another example of a screen for instructing the creation of a financial summary. In the screen C12 for instructing the creation of a financial comparison summary shown in Fig. 22, a character string "Please instruct the creation of a financial comparison summary", a display field D121 of the organizations to be compared, an overview D122 of the target financial data, an input field E121 for instructions to create a financial comparison summary, and a creation start button B122 are displayed. The display field D121 displays two organizations to be compared (Company A and Company B in the example of Fig. 22). The overview D122 displays information such as "Financial statements" and "Account items" that indicate the scope of the target financial data.

[0133] The creation instruction input field E121 contains the following prompt, instructing the user to create a summary document comparing the numerical trends shown in the selected range of financial data for the two selected organizations: "The financial data to be compared show the year-on-year increase / decrease in monthly accounts receivable for Company A and Company B. The accounting periods for these companies are from December to November of the following year. Compare and explain the trends in this data. Please do not explain the factors or reasons why such trends are seen."

[0134] When the creation start button B122 is operated and financial creation instruction data is transmitted, the server device 10 acquires the target financial data of the two selected organizations and inputs both acquired target financial data to the trend determination AI module 120. Next, based on the creation instructions indicated by the financial creation instruction data, the server device 10 instructs the trend determination AI module 120 to output the results of comparing the trends of the numerical values ​​indicated by both target financial data. The trend determination AI module 120 generates and outputs a comparison summary text indicating the results of comparing the trends of the numerical values ​​of the two selected organizations based on the input target financial data of both organizations.

[0135] Fig. 23 is a diagram showing an example of a displayed financial comparison summary. The financial comparison summary display screen C13 shown in Fig. 23 displays a character string "A financial comparison summary has been created", a display field D121 for comparison target organizations, an overview D122 of the target financial data, a financial comparison summary F131, and a journal entry comparison summary creation button B132.

[0136] Financial comparison summary F131 shows trend comparison 1, "Monthly fluctuation: Company A's increase from January to February is much higher than Company B's. On the other hand, Company B's increase in March is larger," trend comparison 2, "Magnitude: Company A has months in April where the increase or decrease in accounts receivable exceeds 500 million yen (e.g. August, September, October, November), but Company B does not do so in February," and also shows, in text, trend comparison 3 for "Year-on-year total," trend comparison 4 for "high peak," and trend comparison 5 for "relative low." In this way, a financial comparison summary is created from the financial data of both comparison target organizations.

[0137] The server device 10 also generates a journal entry comparison summary from the financial comparison summary and the journal entry data of the selected comparison target organizations. The journal entry data may be specified, for example, by using a journal entry summary creation screen C4 shown in FIG. 8. The server device 10 may, for example, automatically specify the range of journal entry data specified for the first target organization for the second target organization, or may specify the range of journal entry data separately for the first target organization and the second target organization. The server device 10 creates a journal entry comparison summary based on the journal entry data specified in this way and the created financial comparison summary.

[0138] Fig. 24 is a diagram showing an example of a displayed journal entry comparison summary. The journal entry comparison summary display screen C16 shown in Fig. 24 displays the character string "Journal entry comparison summary has been created", an input data summary D161, a journal entry comparison summary F161, and a Create final comparison summary button B162.

[0139] The journal entry comparison summary F161 shows, in text form, factor comparison 1, "Monthly fluctuations: A's accounts receivable fluctuations are mainly due to... On the other hand, B's accounts receivable fluctuations are mainly due to...," and factor comparison 2, "Magnitude: A's accounts receivable fluctuations are large from August to November, mainly due to... On the other hand, B's accounts receivable fluctuations are large from September and October, mainly due to....". The following factors are shown in text form: factor comparison 3 for "Comparative total compared to the same month of the previous year," factor comparison 4 for "High peak," and factor comparison 5 for "Relative low price." These factor comparisons 1 to 5 represent the factors that caused trend comparisons 1 to 5, which are included in the financial comparison summary F131 shown in Figure 23. In this way, the journal entry comparison summary is created from the financial comparison summary of the comparison target organizations and the journal entry data of both targets.

[0140] The server device 10 also creates a final comparison summary based on the financial comparison summary and journal entry comparison summary created as described above. The instruction to create the final comparison summary may be given, for example, by using a creation instruction screen similar to the final summary creation instruction screen C7 shown in FIG. 12. In this case, the created financial comparison summary and journal entry comparison summary are displayed in the input data summary D71, and a prompt including an instruction such as "Using data 1 and 2, explain the comparison of trends described in the accounts receivable summary data of companies A and B and the comparison of the factors thereof based on the journal entry summary data of companies A and B." is input in the creation instruction input field E71. The server device 10 creates the final comparison summary based on the financial comparison summary and journal entry comparison summary thus specified.

[0141] Fig. 25 is a diagram showing an example of a displayed final comparison summary. A character string "Final comparison summary has been created", an input data summary D171, a final comparison summary F181, and a save button B182 are displayed on the final comparison summary display screen C18 shown in Fig. 25. In the example of Fig. 25, the summary D171 shows the financial comparison summary F131 shown in Fig. 23 and the journal entry comparison summary F161 shown in Fig. 24.

[0142] Final comparison summary F181 shows comparison summary 1, "1. Monthly fluctuation: The large variation seen in Company A's accounting summary is contributed by the ... department. On the other hand, the variation seen in Company B's accounting summary is contributed by the ... department," and comparison summary 2, "2. Magnitude: The high figures seen in Company A's August to November are due to the ... factor. On the other hand, the high figures seen in Company B's September and October are due to the ... factor." In addition, comparison summary 3 of "Comparison with the same month of the previous year" comparison summary 4 of "High peak" comparison summary 5 of "Relative low value" are shown in text.

[0143] These comparison summaries 1-5 are a compilation of trend comparisons 1-5 shown in FIG. 23 and factor comparisons 1-5 shown in FIG. 24. In this way, a final comparison summary is created from the financial comparison summary and journal entry comparison summary of the compared organizations. According to this embodiment, it is possible to grasp the trend of the values ​​and their factors related to an organization (first organization) based on a comparison with another organization (second organization). Furthermore, by creating a summary text using both the first and second numerical data groups, it is possible to display a summary text that shows more detailed trends or factors than when only one of the numerical data groups is used.

[0144] <Modification: Comparison with Virtual Organization> In the above example, this can be implemented when the comparison subject for company A is a single organization, company B, and company A is in a position to refer to the financial data and journal entry data of company B. An example of a position to refer to is company A being the parent company of company B. Also, if only a trend comparison summary is to be created, an organization that publishes financial data, such as a listed company, an NPO, or a public organization, can be the second organization on its own.

[0145] However, naturally, there are cases where financial data and journal data cannot be referenced by other organizations. In such cases, a set of numerical data that can be used as a reference for comparison may be used while preventing the first organization from knowing that the data are financial data and journal data of a specific organization. Specifically, the set of numerical data for the second organization may be a set of data obtained by aggregating sets of numerical data for multiple organizations. For example, when the first organization is Company A, a virtual organization (hereinafter referred to as a "virtual organization") that includes N organizations (N is a natural number) is used as the second organization.

[0146] The financial data of the virtual organization is data obtained by aggregating the financial data of N companies included in the virtual organization, for example, data obtained by averaging the values ​​of the same item of the financial data of each company. Note that the aggregated data is not limited to the average value, and the median, mode, minimum value, maximum value, etc. may also be used. Furthermore, if trends of change are to be compared, a total value may be used. According to such an embodiment, it is possible to make it more difficult to identify the organization to be compared, compared to a case where the comparison target is not a virtual organization.

[0147] In addition, all of the multiple organizations included in the virtual organization may be organizations that satisfy a predetermined condition. The predetermined condition here includes a condition that at least one of attributes or situations is similar to that of the first organization. In this case, a screen for selecting a target organization will be described with reference to FIG.

[0148] Fig. 26 is a diagram showing an example of a selection screen for a target organization. In addition to the selection fields shown in Fig. 21, the summary creation screen C11a shown in Fig. 26 displays input fields D114 and D115 for the conditions of the virtual organization and a display field D116 for the number of organizations included in the virtual organization. In the input field D114, a condition called "common attribute" of the organizations included in the virtual organization is input. In the input field D115, a condition called "common situation" of the organizations included in the virtual organization is input by the user.

[0149] An "attribute" is information that indicates the characteristics or properties inherent to an organization, and is stable without significant change. On the other hand, a "situation" is information that indicates the environment, conditions, or state at a certain point in time or a specific period of time, and is more likely to change than an attribute. In the example of FIG. 26, common attributes such as "industry type," "business type," and "scale" are input in the input field D114, and common situations such as "financial situation" and "customer situation" are input in the input field D115. These are, for example, AND conditions, but may be OR conditions or may be mixed. In addition, the common attributes may include other areas of activity or corporate form, and the common situations may include other business situations or market situations.

[0150] In addition, the common situation may be found after the financial comparison summary or the journal entry comparison summary is created. For example, when a financial comparison summary or journal entry comparison summary is created based on a second organization (hereinafter referred to as a "first virtual organization") including multiple organizations whose "financial situation" and "customer situation" are common to the first organization, it may be shown that some of the multiple organizations have a sales change trend (one of the business situations, for example, "high sales in summer") that is also common to the first organization. In that case, the server device 10 may cause the text generation AI to generate a comparison summary text (which may be a financial comparison summary or a journal entry comparison summary) by treating the multiple organizations whose common situation is further shown as a "second virtual organization". According to this embodiment, it is possible to create a comparison summary text with a virtual organization (second virtual organization) that has a higher similarity to the first organization than the first virtual organization.

[0151] When the above common attributes and common situations are input, the display field D116 displays the number of organizations (hereinafter referred to as the "number of organizations"; in the example of FIG. 26, "15") that share the input attributes and situations with the first target organization "Company A". If the number of organizations is too small, there is a possibility that an organization included in the virtual organization will be identified. Therefore, if the number of organizations is less than a predetermined threshold Th1, the user increases or decreases the common attributes or common situations to make the number of organizations equal to or greater than the threshold Th1, and then operates the decision button B111 to decide the comparison target organization. The threshold Th1 is, for example, the number of organizations that can reduce the possibility that an organization included in the virtual organization will be identified.

[0152] The threshold value Th1 may be stored in advance by the server device 10, or may be a value input by the user. The pre-stored threshold value Th1 may or may not be presented to the user. Furthermore, the input by the user may be a multiple-choice type or a free input. In the case of a multiple-choice type, the options may be expressed by words such as "take a rough look" or "take a detailed look", and values ​​may be associated with these words. According to such an embodiment, the threshold value can be determined intuitively. Furthermore, in the case of a free input, a value smaller than a predetermined value may not be input. According to such an embodiment, the possibility of identifying an organization included in the virtual organization can be reduced.

[0153] As described above, by comparing with a virtual organization that includes organizations that have attributes or situations in common with the first organization, commonalities are more likely to emerge in the trends and factors of the numerical data groups than when other organizations are included in the virtual organization, thereby improving the usefulness of the comparison results.

[0154] In addition, the server device 10 may not display the summary text when the number of organizations included in the virtual organization is less than the threshold Th1. In this case, the server device 10 may disable the decision button B111 when the number of organizations displayed in the display field D116 is less than the threshold Th1, or may disable any operation thereafter until the summary comparison text is displayed. In addition, the server device 10 may create the summary comparison text but not display it. According to such an embodiment, the number of organizations in the virtual organization compared by the displayed summary comparison text is always equal to or greater than the threshold Th1, so that it is possible to make it more difficult to identify the organizations included in the virtual organization to be compared, compared to the case where the summary comparison text is displayed even if the number of organizations is less than the threshold Th1.

[0155] Furthermore, the conditions for organizations to be included in the virtual organization are not limited to the above conditions. For example, the server device 10 may store summary information indicating an overview of each organization, and the summary information may include a keyword (such as a phrase related to the business of the first organization) input by the user as a condition. According to this embodiment, even if the attributes or circumstances of the first organization are different, an organization that has some commonality with the first organization can be added to the virtual organization.

[0156] In addition, in the above example, the organizations to be included in the virtual organization are determined by inputting the condition of an organization similar to the first target organization, but the method of determining the virtual organization is not limited to this. Fig. 27 is a diagram showing another example of the target organization selection screen. In addition to the selection fields shown in Fig. 21, the summary creation screen C11b shown in Fig. 27 displays an automatic virtual organization generation button B115, a similarity condition display field D117, and a number of organizations display field D118.

[0157] When the automatic generation button B115 is operated, the server device 10 executes a specification step of specifying a condition (hereinafter referred to as a "similar condition") under which the number of organizations having at least one of attributes or situations similar to the first organization is equal to or greater than a threshold Th1. The server device 10, for example, stores a list of attributes and situations used as similar conditions, randomly selects one attribute or situation from the list, and extracts organizations whose selected attribute or situation is similar to the first target organization.

[0158] If the number of extracted organizations is less than the threshold Th1, the server device 10 starts over from selecting an attribute or situation, and if the number of extracted organizations is equal to or greater than the threshold Th1, it identifies the selected attribute or situation as a similar condition. If the number of extracted organizations is not equal to or greater than the threshold Th1 with one attribute or situation, the server device 10 randomly selects two from the attributes or situations and extracts organizations in the same manner. The server device 10 repeats this process until the number of extracted organizations is equal to or greater than the threshold Th1, and identifies similar conditions.

[0159] The server device 10 displays the identified similarity conditions in a display field D117, and displays the number of organizations extracted based on the identified similarity conditions in a display field D118. The server device 10 performs the above extraction and identification each time the automatic generation button B115 is operated. If the user sees the displayed similarity conditions and the number of organizations and judges them to be appropriate as a virtual organization, the user operates the decision button B111 to decide on the virtual organization. According to this embodiment, by operating the automatic generation button B115, it is possible to decide on a virtual organization in which the number of organizations is always equal to or greater than the threshold value Th1, so that trends or factors can be compared without wasting extra time.

[0160] The server device 10 may display a plurality of buttons associated with different threshold values ​​Th1 as the automatic generation buttons for the virtual organization, and may perform the above extraction and identification using the threshold value Th1 associated with the operated button. For example, by setting the threshold value Th1 to 10 for the "small button", 20 for the "medium button", and 30 for the "large button", the size of the virtual organization can be selected depending on the button operated.

[0161] The server device 10 may also execute an item presenting step of presenting items whose commonality is equal to or greater than a threshold Th11 among the items whose numerical values ​​are included in the numerical data sets of the multiple organizations included in the virtual organization. The commonality is an index showing the degree of commonality among the multiple organizations included in the virtual organization. For example, when the proportion of organizations whose numerical values ​​for the item "sales" are included in a common range among the multiple organizations included in the virtual organization is equal to or greater than a threshold Th12, the server device 10 determines that the commonality of the item "sales" is equal to or greater than a threshold Th11.

[0162] In addition, when the number of organizations in the virtual organization that have a common range of values ​​for the item "Personnel Cost" is equal to or greater than the threshold Th13, the server device 10 determines that the commonality of the item "Personnel Cost" is equal to or greater than the threshold Th11. The items for which the commonality is expressed are not limited to the above, and may be any journal entry item, account item, business partner, or journal entry summary. Since any character string can be entered into the journal entry summary, it is possible to perform highly flexible search and calculation of commonality that takes into account semantic agreement in addition to exact matches. In this way, by presenting items whose commonality is equal to or greater than the threshold Th11, it is possible to grasp the characteristics of the multiple organizations included in the virtual organization.

[0163] <Variation: Public Information> The information included in the numerical data set is not limited to information provided by the first organization or the second organization for reference. For example, the server device 10 may execute a generating step of executing a generating process by adding or substituting statistical information provided by a public institution to the numerical data set. In this case, the statistical information is information indicating statistics about a plurality of organizations including at least one of the first organization and the second organization.

[0164] Specifically, for example, government statistical information (provided by each ministry and agency), economic census (provided by Statistics Bureau of the Ministry of Internal Affairs and Communications), corporate enterprise statistics information (provided by the Policy Research Institute), or Basic Survey on Small and Medium Enterprises (provided by the Small and Medium Enterprise Agency) may be used as statistical information. These statistical information can be freely viewed, so they can also be used as a set of numerical data for the second organization. In this case, the second organization is a virtual organization including multiple organizations, similar to the virtual organization described above. Note that statistical information provided by not only public institutions but also private institutions for a fee (or free of charge) may be used as the set of numerical data. According to this embodiment, even if there is a lack of information on the second organization that can be provided by the summary creation system 1, it is possible to compare the first organization and the second organization.

[0165] <Modification: Method of acquiring first numerical data group> In the above example, the range of the first numerical data group to be acquired is specified by the user, but this is not limited thereto. For example, the server device 10 may execute an acquisition step of acquiring a numerical data group related to the numerical trend indicated by the second numerical data group of the first organization as the first numerical data group of the first organization. For example, the server device 10 acquires a numerical data group highly related to a word or phrase included in the trend summary (which may be a trend comparison summary; the same applies below) as the first numerical data group.

[0166] Specifically, for example, when the trend summary includes a term indicating a time (such as a month or a season), the server device 10 acquires a group of numerical data for the time indicated by the term as the first numerical data group. Also, when the trend summary includes a term indicating a specific item (such as an account entry or an account item), the server device 10 acquires a group of numerical data for the item indicated by the term as the first numerical data group. Also, when the trend summary includes a term indicating a specific supplier or business partner, etc., the server device 10 acquires a group of numerical data indicating transactions with the supplier or business partner, etc. indicated by the term as the first numerical data group.

[0167] According to this aspect, the first numerical data group is narrowed down to a numerical data group associated with the determined numerical trend, thereby reducing the processing load for analyzing the factors behind the trend compared to when all numerical data groups are acquired as the first numerical data group.

[0168] The server device 10 may automatically acquire the first numerical data group based on the trend summary without asking the user about the range of the first numerical data group to be acquired. The server device 10 may also execute a numerical value presentation step of presenting the first numerical data group acquired as a numerical data group related to the trend to the user. When an item included in the presented first numerical data group is corrected by the user, the server device 10 executes an acquisition step of acquiring the numerical data group of the corrected item as the first numerical data group of the first organization. According to this aspect, the accuracy of the first numerical data group can be improved compared to a case where the user cannot make corrections.

[0169] Whether or not to present the range of the first numerical data group to the user may be set by the user's operation. In addition, the server device 10 may present the first numerical data group to the user when the number of the first numerical data group acquired as the numerical data group related to the trend is equal to or greater than the threshold Th21. If the number of the first numerical data group is too large, the effect of reducing the processing load of analyzing the above-mentioned trend factors is reduced, so the first numerical data group is presented to the user so that unnecessary items can be removed from the range of the first numerical data group. Therefore, the threshold Th21 may be determined based on the number of the first numerical data group that can be expected to have the effect of reducing the processing load.

[0170] The server device 10 may acquire the first numerical data group by narrowing down the numerical data group to one having a higher relevance to the numerical trend so that the number of the first numerical data group is less than the threshold value Th21. The degree of relevance to the numerical trend is determined, for example, by the strength of the connection between the word included in the trend summary and the item of the numerical data, or the number of times the word appears.

[0171] Conversely, the server device 10 may present the first numerical data group to the user when the number of the first numerical data group acquired as the numerical data group related to the trend is less than the threshold value Th22. If the number of the first numerical data group is too small, the factors of the trend described above cannot be analyzed appropriately, so the first numerical data group is presented to the user so that the user can add items to be included in the first numerical data group. Therefore, the threshold value Th22 may be determined based on the number of the first numerical data group that can be expected to have the effect of appropriately analyzing the factors of the trend.

[0172] The server device 10 may acquire as the first numerical data group a numerical data group having a relatively high relevance to the trend so that the number of the first numerical data groups is equal to or greater than the threshold Th22. In addition, even if the number of the first numerical data groups is equal to or greater than the threshold Th22, if the first numerical data groups have a low relevance to the words and phrases included in the trend summary, the server device 10 may present the first numerical data groups to the user to allow the user to determine whether the first numerical data groups are appropriate.

[0173] <Variations: highlighted> In the above example, the first numerical data group related to the words included in the trend summary is acquired, but the server device 10 may execute a process for displaying a sentence highlighting words related to the acquired first numerical data group among words explaining the numerical trend as a display process. Methods for highlighting words include highlighting, bolding, underlining, etc. According to this embodiment, the user can know what related numerical data was used in the factor analysis.

[0174] <Examples of variations: composition variations> The configuration (overall configuration, hardware configuration, functional configuration, etc.) shown in FIG. 1 etc. is an example, and other configurations may be used as long as there is no inconvenience in implementation. For example, the server device 10 may be distributed across two or more devices, and may be provided in the form of SaaS (Software as a Service) or a cloud computing system. Furthermore, the information processing executed by the server device 10 may be executed collectively by the user terminal 20. In short, as long as the necessary information processing is executed in the entire summary creation system 1, the devices that execute the information processing may have any configuration.

[0175] The output destination of information or data (hereinafter referred to as "information, etc.") may be another device, a display, a memory unit (including a built-in memory unit and an external memory unit), etc. Acquisition of information, etc. includes acquiring information, etc. generated by the device itself, in addition to acquiring information, etc. transmitted from another device. Tables, etc. (tables, databases, etc.) that associate parameters are not limited to the tables, etc. shown in the figures, and the number of parameters may be reduced or increased. Furthermore, information, etc. corresponding to parameters may be obtained by a formula, a conditional formula, etc., without using a table, etc.

[0176] The above-described aspects of the embodiment are information processing devices such as the server device 10 and the user terminal 20, and information processing systems equipped with one or more processors such as the summary creation system 1, but may also be information processing methods. The information processing methods include the same steps as those executed by the information processing system. The above-described aspects of the embodiment may also be programs. The programs cause a computer to execute the same steps as those executed by the information processing system.

[0177] <Additional Notes> Furthermore, it may be provided in the following aspects:

[0178] (1) An information processing system having one or more processors, wherein the processor performs an acquisition step of acquiring a set of numerical data for a first organization and a second organization, the set of numerical data being a collection of numerical values ​​relating to the first organization and the second organization, a generation step of executing a generation process in which a text generation AI generates a summary text including a result of comparing at least one of the numerical trends indicated by the set of numerical data and the factors of the trend for the first organization and the second organization based on the acquired set of numerical data for each of the first organization and the second organization, and a display step of executing a display process in which the generated summary text is displayed on a display means.

[0179] According to this embodiment, it is possible to grasp the trend of the numerical values ​​and the factors behind them based on a comparison with other organizations.

[0180] (2) In the information processing system described in (1) above, the numerical data group includes a first numerical data group and a second numerical data group, the second numerical data group includes numerical values ​​derived from the first numerical data group, and the generation process is a process of having the sentence generation AI generate, as the summary sentence, a sentence indicating the result of the comparison of at least one of the numerical trend indicated by the second numerical data group and the factor of the trend indicated by the first numerical data group based on the first numerical data group and the second numerical data group included in the acquired numerical data group.

[0181] According to this embodiment, it is possible to display a summary text showing more detailed trends or factors.

[0182] (3) In the information processing system described in (1) or (2) above, the group of numerical data of the second organization is a data group obtained by aggregating the groups of numerical data of a plurality of organizations.

[0183] According to this embodiment, it is possible to make it difficult to identify the tissue to be compared.

[0184] (4) In the information processing system described in (3) above, all of the multiple organizations are organizations that satisfy predetermined conditions, and the conditions include a condition that at least one of attributes or circumstances is similar to that of the first organization.

[0185] According to this embodiment, the usefulness of the comparison result can be improved.

[0186] (5) In the information processing system described above in (4), when the number of organizations included in the plurality of organizations is less than a threshold, the summary text is not displayed.

[0187] According to this embodiment, it is possible to make it more difficult to identify the tissue to be compared.

[0188] (6) In the information processing system described above in (5), in the identification step, the processor identifies the condition under which the number of similar organizations is equal to or greater than the threshold.

[0189] In this manner, trends or factors can be compared without wasting additional time.

[0190] (7) In the information processing system described in any one of (3) to (6) above, in the item presentation step, the processor presents items among the numerical items included in the numerical data group of the multiple organizations, the items having a commonality indicating the degree of commonality among the multiple organizations that is equal to or greater than a threshold value.

[0191] According to this embodiment, it is possible to grasp the characteristics of a plurality of tissues.

[0192] (8) In the information processing system described in any one of (1) to (7) above, in the generation step, the processor executes the generation process by adding or substituting statistical information provided by a public or private institution to the group of numerical data, and the statistical information is information indicating statistics about a plurality of organizations including at least one of the first organization or the second organization.

[0193] According to this embodiment, a comparison can be made even if information about the second tissue is insufficient.

[0194] (9) In the information processing system described in (2) above, in the acquisition step, the processor acquires a group of numerical data related to the numerical trend indicated by the second group of numerical data of the first organization as the first group of numerical data of the first organization.

[0195] According to this embodiment, the processing load for analyzing the factors of the trend can be reduced.

[0196] (10) In the information processing system described in (9) above, in the numerical presentation step, the processor presents to a user the first numerical data group acquired as a numerical data group related to the trend, and in the acquisition step, if an item included in the presented first numerical data group is corrected, acquires the numerical data group of the corrected item as the first numerical data group of the first organization.

[0197] According to this embodiment, the accuracy of the first numerical data group can be improved.

[0198] (11) In the information processing system described in (9) or (10) above, in the display step, the processor executes a process for displaying a sentence that highlights words that are related to the acquired first numerical data group among words that explain the numerical trend as the display process.

[0199] According to this embodiment, it is possible to know what the numerical data related to is used in the factor analysis.

[0200] (12) An information processing method comprising the steps of the information processing system according to any one of (1) to (11) above.

[0201] According to this embodiment, it is possible to grasp the trend of the numerical values ​​and the factors behind them based on a comparison with other organizations.

[0202] (13) A program for causing a computer to execute each step of the information processing system according to any one of (1) to (11) above.

[0203] According to this embodiment, it is possible to grasp the trend of the numerical values ​​and the factors behind them based on a comparison with other organizations. Of course, this is not the case. Furthermore, the above-described embodiments and modifications may be combined in any desired manner.

[0204] Finally, although various embodiments of the present invention have been described, these are presented as examples and are not intended to limit the scope of the invention. The new embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. The embodiments and their modifications are within the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0205] 1: Summary creation system 2: Communication lines 10: Server device 11: Control section 20: User terminal 21: Control section 100: AI module 110: Summary creation AI module 120: Trend Judgment AI Module 130: Factor Judgment AI Module DB1: Journal entry database DB2: Financial Database

Claims

1. An information processing system including one or more processors, The processor, In the acquisition step, a set of numerical data is acquired for the first organization and the second organization, the numerical data being a collection of numerical values ​​related to the organizations; In the generation step, a generation process is executed to cause a sentence generation AI to generate a summary sentence including a result of comparing at least one of the tendency of the numerical values ​​indicated by the numerical data group and the factor of the tendency for the first organization and the second organization based on the acquired numerical data group of each of the first organization and the second organization; In the display step, a display process is performed to display the generated summary text on a display means. Information processing system.

2. 2. The information processing system according to claim 1, The numerical data group includes a first numerical data group and a second numerical data group, the second set of numerical data includes numerical values ​​derived from the first set of numerical data; The generation process is a process of causing the sentence generation AI to generate a sentence as the summary sentence, the sentence indicating the result of the comparison of at least one of the tendency of the numerical values ​​indicated by the second numerical data group and the factor of the tendency indicated by the first numerical data group, based on the first numerical data group and the second numerical data group included in the acquired numerical data group. Information processing system.

3. 2. The information processing system according to claim 1, The numerical data group of the second organization is a data group obtained by aggregating the numerical data groups of a plurality of organizations. Information processing system.

4. 4. The information processing system according to claim 3, The plurality of organizations are all organizations that satisfy a predetermined condition, The conditions include a condition that at least one of attributes or circumstances is similar to that of the first organization. Information processing system.

5. 5. The information processing system according to claim 4, When the number of organizations included in the plurality of organizations is less than a threshold, the summary text is not displayed. Information processing system.

6. 6. The information processing system according to claim 5, The processor, In the identification step, the condition in which the number of similar tissues is equal to or greater than the threshold is identified. Information processing system.

7. 4. The information processing system according to claim 3, The processor, In the item presenting step, items whose commonality indicating a degree of commonality among the plurality of organizations is equal to or greater than a threshold value are presented among the items whose numerical values ​​are included in the numerical data set of the plurality of organizations. Information processing system.

8. 2. The information processing system according to claim 1, The processor, In the generating step, the generating process is performed using statistical information provided by a public institution or a private institution in addition to or in place of the set of numerical data; The statistical information is information indicating statistics about a plurality of organizations including at least one of the first organization and the second organization. Information processing system.

9. 3. The information processing system according to claim 2, The processor, In the acquiring step, a set of numerical data related to a tendency of numerical values ​​indicated by the second set of numerical data of the first organization is acquired as the first set of numerical data of the first organization. Information processing system.

10. 10. The information processing system according to claim 9, The processor, In the numerical value presentation step, the first numerical value data group acquired as the numerical value data group related to the trend is presented to a user; In the acquiring step, when an item included in the presented first numerical data group is corrected, the corrected numerical data group of the item is acquired as the first numerical data group of the first organization. Information processing system.

11. 10. The information processing system according to claim 9, The processor, In the display step, a process for displaying a sentence in which a phrase related to the acquired first numerical data group is highlighted among the phrases explaining the tendency of the numerical values ​​is executed as the display process. Information processing system.

12. 1. An information processing method, comprising: The information processing system according to any one of claims 1 to 11, Information processing methods.

13. A program, A computer is caused to execute each step of the information processing system according to any one of claims 1 to 11. program.

Images