Information processing systems, information processing methods, and programs

Abstract

This reduces the effort required for users to enter their own information. [Solution] The information processing system includes a receiving means for receiving identification information from users of the service, a first acquisition means for acquiring user information, which is information about the user, based on the input identification information, a second acquisition means for inputting the user information into a pre-trained AI module and obtaining the category to which the user belongs by outputting it to the AI ​​module, and a utilization means for using the acquired category as the user's basic information.

Description

Technical Field

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

Background Art

[0002] As documents disclosing the background art of this technical field, there are Non-Patent Document 1 and Non-Patent Document 2. Non-Patent Document 1 describes that "based on the company name entered by the user, detailed company data of that company is automatically obtained". On the other hand, Non-Patent Document 2 describes that "when the selling company enters its own information on "M&A Cloud", just by entering the company URL, a function of automatically generating "business keywords", "business content", "strengths", and "title of project information" used when the buying company searches for the selling company" is described.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Non-Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] Non-patent document 1, mentioned above, describes how to obtain detailed company data based on the entered company name. On the other hand, non-patent document 2 describes how to automatically generate "business keywords," etc., simply by entering the company URL. However, neither document discloses a mechanism for obtaining user information based on identification information entered by the user, and then inputting that user information into an AI module to obtain the category to which the user belongs.

[0005] This invention has been made in view of these circumstances, and aims to reduce the effort required for users to input their own information. [Means for solving the problem]

[0006] To solve the above problems, for example, the configuration described in the claims is adopted. The present invention includes multiple means for solving the above-mentioned problems, but one example is an information processing system comprising: a receiving means for receiving identification information from a user using the service; a first acquisition means for acquiring user information, which is information about the user, based on the input identification information; a second acquisition means for inputting the user information into a pre-trained AI module and obtaining the category to which the user belongs, by having the AI ​​module output it; and a utilization means for using the acquired category as basic information of the user. [Effects of the Invention]

[0007] According to the present invention, the effort required for users to input their own information can be reduced. Other issues, configurations, and effects not mentioned above will be clarified by the following description of the embodiments. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 shows an example of the procedure in the embodiment. [Figure 2]Figure 2 shows an example of the configuration of the information processing system 200. [Figure 3] Figure 3 shows an example of the configuration of the business support server 201. [Figure 4] Figure 4 shows an example of the configuration of LLM server 202. [Figure 5] Figure 5 shows an example of the industry estimation process 500. [Figure 6] Figure 6 shows an example of the industry list expansion process 600. [Figure 7] Figure 7 shows an example of the industry estimation process 700. [Figure 8] Figure 8 shows an example of a web page that can be accessed from the top page. [Figure 9] Figure 9 shows a specific example of the industry list expansion process 600. [Figure 10] Figure 10 shows an example of the relationship between the target user and the industry of their trading partner users. [Figure 11] Figure 11 shows an example of the relationship between the target user and the industry of their trading partner users. [Figure 12] Figure 12 shows an example of the relationship between the target user and the industry of their trading partner users. [Figure 13] Figure 13 shows an example of the relationship between a target user and the names of the products they bought and sold. [Modes for carrying out the invention]

[0009] 1. Examples Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0010] (1) Overview First, an overview of the embodiments of the present invention will be described. Traditionally, service user statistics have been obtained based on "accounting information" and "user information" entered into software as a service (SaaS), such as accounting software, to recommend other services or analyze one's own services.

[0011] However, users who use the above services have no problem as long as they can use the services. In the case of the user's own information input, it is very rare for all information (accountant information, number of employees, deployed services, future situation strategies, etc.) to be input. Also, since the person in charge does not know appropriate accounting information, there may be cases where such information cannot be input. Under such circumstances, it is impossible to recommend services that match the user's needs.

[0012] Also, even when the user himself / herself registers, it is not always correct, and it is assumed that appropriate statistical information cannot be obtained and effective in-house analysis cannot be performed.

[0013] This embodiment is made in view of the above circumstances, and greatly reduces the labor for the user to input his / her own information, and enables more accurate registration of user information.

[0014] In order to solve the above problems, in this embodiment, based on the information on the user's web page and the information on the web page where other user information is described, etc., the user is allowed to select an industry type from an industry type list. If there seems to be nothing related to the industry type list, an industry type to be added to the industry type list is generated.

[0015] The procedure of this embodiment will be described with reference to FIG. 1. The procedure of this embodiment is as follows. A. Prepare an industry type list. At this stage, the industry type list does not have to be complete. B. The user inputs business operator information (at least one of business operator name, corporate number, location, telephone number, representative name, capital, number of employees, industry type, business form, business content, products / services, history, base, major customers, shareholders, corporate philosophy, future vision, invoice number, etc.). C. Search the web page based on the business operator information, and obtain information on the user's corporate website and other web pages where user information is described (Step 101). E. Provide the LLM with a list of industries and a webpage containing user sites and other user information, and have it solve the following two tasks (Step 102). (a) A task to select several appropriate industries from the list of industries (Step 103) (i) A task to generate industries that are not included in the industry list but are deemed appropriate (Step 104) (o) Output the industry obtained from the selection task as the user's industry (step 105). (k) Output the industry type obtained in the generation task as the user's industry type (step 106). The industries obtained in the generation task are passed to the industry list expansion algorithm (step 107), and those that pass this process are added to the industry list (step 108).

[0016] In step 102, the LLM is given prompts such as, "Select an appropriate industry from the industry list. If there is no appropriate industry in the industry list, generate a new industry name."

[0017] Furthermore, in step 104, there may be cases where "this is the one I would choose, but it's not a perfect fit." Therefore, in addition to generating if a selection cannot be made, it may also be possible to generate while making a selection.

[0018] The following modifications may be applied to the above procedure. These modifications may also be combined with each other. A. Provide LLM with several concrete examples (few-shot learning / in-context learning). If selection / generation fails in step 102, change the user site obtained in step 101. In step 102, selection and generation are performed separately. For example, if no LLMs are selected, the generation process is run separately. E. Narrow down the list of industries provided in Step 102 using some kind of information. For example, if the broad category of the target user's industry is known, only the industries included in that broad category should be provided. In this case, if each industry in the industry list (e.g., the Japan Standard Industrial Classification) has a description or examples of business activities, you may also check the consistency between these and the user's information. The few-shot examples provided should be tailored to the specific user. If the industry list has a hierarchical structure, perform the above process for each level of the hierarchy.

[0019] (2) Composition Next, the configuration of this embodiment will be described. Figure 2 shows an example of the configuration of the information processing system 200 according to this embodiment. The information processing system 200 shown in the figure comprises a business support server 201, an LLM server 202, and a number of user terminals 203. These devices are connected via a wired or wireless network and are able to send and receive information from each other.

[0020] Business support server 201 is a server designed to support users' business operations. The LLM server 202 is a server that receives prompts from the business support server 201 and outputs responses to the received prompts. The multiple user terminals 203 are terminal devices used by individual users. Each user receives business support services provided by the business support server 201.

[0021] Each of these devices comprises a processor that runs an operating system, applications, programs, etc.; main memory such as RAM (Random Access Memory); auxiliary storage such as IC cards, hard disk drives, SSDs (Solid State Drives), flash memory, etc.; a communication control unit such as a network card, wireless communication module, or mobile communication module; input devices such as a touch panel, keyboard, mouse, voice input, or motion detection input via a camera; and output devices such as a monitor or display. The output device may also be a device or terminal that transmits information for output to an external monitor, display, printer, or other device.

[0022] The main memory stores various programs and applications (modules), and the processor executes these programs and applications to realize each functional element of the overall system. These modules may be implemented in hardware, such as through integration. Furthermore, each module may be an independent program or application, or it may be implemented as a subprogram or function within a single integrated program or application.

[0023] In this specification, each module is described as the entity (subject) that performs the processing; however, in reality, the processor that processes various programs and applications (modules) executes the processing.

[0024] The auxiliary storage device stores various databases (DBs). A "database" is a functional element (storage unit) that stores a collection of data so that it can be manipulated by the processor or an external computer (e.g., extraction, addition, deletion, overwriting, etc.). The implementation method of a database is not limited; for example, it may be a database management system, spreadsheet software, or text files such as XML or JSON. The following sections provide a detailed explanation of the business support server 201 and the LLM server 202.

[0025] A. Business support server 201 The business support server 201 is composed of, for example, one or more servers located on the cloud. Figure 3 shows an example of the configuration of the business support server 201. The main memory 301 of the server shown in the figure stores programs and applications such as the reception module 310, the first acquisition module 311, the second acquisition module 312, the utilization module 313, the update module 314, the first estimation module 315, and the second estimation module 316. The processor 303 executes these programs and applications to realize each functional element of the business support server 201. Each module will be described below.

[0026] The reception module 310 accepts identification information from users of the service. Specifically, this identification information includes the company name, corporate number, address, telephone number, representative name, capital, number of employees, industry, business type, business content, products / services, history, locations, major business partners, shareholders, corporate philosophy, future vision, invoice number, user's website URL, and URLs of other websites containing user information. The reception module 310 registers the entered identification information as part of the business information described later.

[0027] The first acquisition module 311 acquires user information, which is information about the user, based on the identification information received by the reception module 310. Specifically, user information refers to web pages on which the user's information is posted or other web pages that contain user information. More specifically, it refers to at least a part of the user's company website (for example, the top page) or web pages found using a search engine.

[0028] When user information is acquired by the first acquisition module 311, the second acquisition module 312 inputs the user information into a pre-trained AI module and obtains the category to which the user belongs by having the AI ​​module output it.

[0029] The AI ​​module referred to here is, for example, a deep learning model (in other words, a generative AI such as an LLM) that has been pre-trained on a large dataset. This AI module outputs a response to an input prompt (in other words, a command). Because this AI module is pre-trained on a deep learning model using a large dataset, it can be used even without training data for category estimation. This AI module corresponds to the generative AI module 410 described later. Furthermore, the category to which a user belongs, as referred to here, specifically refers to the industry name. However, the industry name is merely an example, and the user may belong to a different category.

[0030] More specifically, the second acquisition module 312 inputs user information and a category list into the AI ​​module when user information is acquired. If the category to which the user belongs is included in the category list, the AI ​​module selects and acquires it. On the other hand, if the category to which the user belongs is not included in the category list, the AI ​​module generates and acquires it. Specifically, the category list referred to here is the industry list.

[0031] The second acquisition module 312 may also provide concrete examples to the AI ​​module (few-shot learning / in-context learning). Furthermore, the second acquisition module 312 may separately instruct the AI ​​module to select and generate categories. In other words, the prompts for category selection and generation may be separated.

[0032] The utilization module 313 uses the categories obtained by the second acquisition module 312 as the user's basic information. For example, the utilization module 313 uses the obtained categories for the following purposes: (a) Analysis, comparison (i) Preparation of various application / notification forms and declaration forms (c) Recommendations for subsidies, grants, and other forms of fundraising. (e) Estimates, simulations (e) Recommendations for proper input of accounting information

[0033] Furthermore, the user module 313 may use the acquired category as the user's basic information only if the user has given their approval for that category. This prevents the use of inaccurate categories.

[0034] The update module 314 updates the category list mentioned above. Specifically, for a predetermined number of users among the multiple users of the service, the update module 314 adds the category to the category list when a category to which a user belongs has been generated and retrieved by the second retrieval module 312.

[0035] Furthermore, when a category is generated for each of the multiple users using the service and retrieved by the second retrieval module 312, the update module 314 clusters the retrieved categories and adds the representative category of each cluster to the category list.

[0036] The first estimation module 315 estimates the category to which the user belongs based on information of other users who are the user's business partners, if user information is not obtained by the first acquisition module 311. In this case, the first estimation module 315 inputs information of other users who are the user's business partners into the category estimation model M1 and outputs the category to which the user belongs.

[0037] The category estimation model M1 referred to here is a pre-trained machine learning model. This category estimation model M1 takes information about other users who are business partners of the user as input and outputs the category to which the user belongs.

[0038] The second estimation module 316 estimates the category to which a user belongs based on the user's transaction history if user information is not obtained by the first acquisition module 311. In this process, the second estimation module 316 identifies one or more other users who are the user's trading partners from the user's transaction history. The module then identifies information on one or more items that each of the identified other users has bought or sold. The module then inputs the information on one or more items that each of the other users has bought or sold into the category estimation model M2 and outputs the category to which the other users belong.

[0039] The category estimation model M2 referred to here is a pre-trained machine learning model. This category estimation model M2 takes information about one or more products that a user has bought and sold with a trading partner as input and outputs the category to which the trading partner belongs.

[0040] The second estimation module 316 estimates a category for each of the other users, then inputs the estimated categories into the category estimation model M1 described above to estimate the category to which the user belongs.

[0041] As a variation, the second estimation module 316 may directly estimate the category from the user's trading history. In this case, the module inputs the product information registered in the user's trading history into the category estimation model M3 and outputs the category to which the user belongs. The category estimation model M3 referred to here is a pre-trained machine learning model. This category estimation model M3 takes the product information registered in the user's trading history as input and outputs the category to which the user belongs.

[0042] Next, the auxiliary storage device 302 will be described. The auxiliary storage device 302 stores business information 320, generated industry information 321, sales history 322, etc. Each piece of information will be described below.

[0043] Business information 320 is information about users of the service (in other words, businesses, companies, sole proprietors, etc.). Specifically, this information includes business name, corporate number, address, telephone number, representative name, capital, number of employees, industry, business type, business content, products / services, history, locations, major business partners, shareholders, corporate philosophy, future vision, invoice number, etc. Each piece of business information 320 is stored in association with a business ID.

[0044] The generated industry information 321 is a set of industry names generated by the AI ​​module described above and acquired by the second acquisition module 312. Each industry name is stored in association with the business ID of the user to whom the industry name was estimated and the URL of the web page used to estimate the industry name.

[0045] Transaction history 322 is the transaction history of a user of the service. Specifically, this information consists of information such as invoices, delivery slips, and receipts. Each transaction history 322 is stored in association with a business ID.

[0046] I LLM Server 202 The LLM server 202 consists of one or more servers located, for example, on the cloud. Figure 4 shows an example of the configuration of the LLM server 202. The main memory 401 of the server shown in the figure stores programs and applications such as the generation AI module 410. The processor 403 executes these programs and applications to realize each functional element of the LLM server 202.

[0047] Among the functional elements to be implemented, the generative AI module 410 is, for example, a deep learning model (in other words, a generative AI such as an LLM) that has been pre-trained on a large dataset. This module outputs a response to an input prompt (in other words, an instruction). Because this module is a deep learning model that has been pre-trained on a large dataset, it can be used even without training data for category estimation.

[0048] (3) Operation Next, we will explain the various processes performed by the business support server 201.

[0049] A. Industry Estimation Processing 500 Figure 5 is a flowchart showing an example of the industry estimation process 500. The process shown in the figure is for estimating the industry of the target user.

[0050] In step 501 of this process, the first acquisition module 311 refers to the business information of the target user and determines whether the user's URL is registered (step 501). If the result of this determination is that the user's URL is registered (YES in step 501), the module uses the user's URL to obtain the top page of the user's site (step 502). On the other hand, if the result of this determination is that the user's URL is not registered (NO in step 501), the module enters the business information into a search engine and obtains the top page of the user's site (step 503). The business information entered into the search engine here includes, for example, the business name, corporate number, address, telephone number, representative name, capital, number of employees, industry, business type, business content, products / services, history, locations, major customers, shareholders, corporate philosophy, future vision, invoice number, etc.

[0051] Next, the second acquisition module 312 generates a prompt to be input to the generation AI module 410 (step 504). The prompt generated here includes the following information: (a) Information posted on the acquired web page, or the URL of the web page (i) Industry list (c) A statement that instructs the system to select the target user's industry from a list of industries based on the information posted on the target user's web page. (e) A statement that commands the system to generate a new industry name if there is no suitable industry in the industry list.

[0052] Next, the second acquisition module 312 inputs the generated prompt to the generation AI module 410 (step 505). The generation AI module 410 outputs a response to the input prompt.

[0053] The second acquisition module 312 refers to the outputted response and determines whether an industry has been selected or generated (step 506). If the result of this determination is that an industry has been selected or generated, the process proceeds to step 507 (YES in step 506). Next, the module determines whether the industry was selected from the industry list (step 507). If the result of this determination is that the industry was selected from the industry list (YES in step 507), the module registers the outputted industry as part of the target user's business information (step 508). On the other hand, if the result of this determination is that the industry was generated by the generation AI module 410 (NO in step 507), the module registers the outputted industry as generated industry information (step 509) and also registers it as part of the target user's business information (step 508).

[0054] If, as a result of the determination in step 506 above, no industry is selected or generated (NO in step 506), the first acquisition module 311 acquires another web page of the target user (step 510). The other web page acquired here is a web page that can be accessed from the top page acquired in step 502 or 503 (in other words, a web page linked to the top page). The method of selecting another web page is, for example, one of the following:

[0055] (a) Select randomly. (i) Select pages with a large amount of information (specifically, a large number of words). (c) Select a page that has the specified content (for example, the "about us" page). (e) Allow the generation AI module 410 to make the selection.

[0056] Figure 8 shows an example of a web page that can be accessed from the top page. Web pages 802 and 803 shown in the figure are examples of web pages that can be accessed from the top page 801.

[0057] If another web page is retrieved in step 510, step 504 is executed again based on the retrieved web page. Steps 504-506 and 510 are repeated until an industry is selected or generated. The above is an explanation of the industry estimation process 500.

[0058] In the industry estimation process 500 described above, the user's industry is selected by the generation AI module 410. If the user's industry is not registered in the industry list, the generation AI module 410 will create it. Therefore, the effort required for the user to register their own industry is reduced.

[0059] The above process 500 may be modified as follows. (a) If the top page was obtained using a search engine in step 503, then in step 510, a web page with a lower search ranking (for example, the second-ranked web page) may be obtained. Web pages 804 and 805 shown in Figure 8 are examples of such web pages. This method is effective when a third-party website contains more information than the user's website.

[0060] Alternatively, the generation AI module 410 can be used to generate search keywords, and these keywords can then be entered into a search engine to retrieve other web pages.

[0061] (i) After obtaining the top page using a search engine in step 503, the user may be asked to confirm whether the obtained top page is appropriate. If the result of this confirmation is that it is not appropriate, the user may be asked to enter the correct user URL.

[0062] I. Industry list expansion process 600 Figure 6 is a flowchart showing an example of the industry list expansion process 600. The process shown in the figure adds the industry names generated by the generation AI module 410 to the industry list. This process is executed periodically.

[0063] In step 601 of this process, the update module 314 obtains multiple industry names registered as generated industry information. Next, the module converts each of the obtained multiple industry names into a vector representation using a well-known embedding method (e.g., Word2vec or BERT) (step 602). Next, the module applies a well-known clustering method (e.g., K-means) to the generated vector representation to cluster the multiple industry names and obtain the representative industry name for each cluster (step 603). Finally, the module adds the obtained representative industry names to the industry list (step 604). The above is an explanation of the industry list expansion process 600.

[0064] Figure 9 shows a specific example of the industry list expansion process 600. In the example shown in the figure, embedding + clustering 902 is applied to the generated set of industry names 901, generating multiple clusters 903. Then, a representative industry name is selected from each cluster and added to the industry list 904.

[0065] According to the industry list expansion process 600 described above, new industry names can be automatically added to the industry list. In addition, by clustering the generated industry names, it is possible to narrow down the list to include only the representative industry names of each cluster.

[0066] The above process 600 may be modified as follows. (a) Instead of steps 602 and 603, multiple industry names may be entered into the generating AI module 410 to perform clustering and identify the representative industry name for each cluster. In this case, the representative industry name may be selected from the industry names entered into the generating AI module 410, or a cluster name that reflects the characteristics of the entered industry names may be generated. In addition, as reference information to improve the accuracy of clustering, at least one of the following may be entered into the generating AI module 410: information or URL posted on the web page used to identify each industry name, and business information or sales history (especially journal entries) of the user whose industry name was estimated.

[0067] (i) Of the representative industry names obtained in step 603, only those that meet at least one of the following conditions may be added to the industry list. a. The representative industry name or a similar industry name is generated frequently. b. For the user whose representative industry name was estimated, no other industry name was estimated in the same way (in other words, the representative industry name is different from an industry name already registered in the industry list and is therefore new). c. The similarity between users whose representative industry names were estimated is high.

[0068] Condition a above, more precisely, means that the frequency with which a representative industry name or a similar industry name is generated is above a threshold. In other words, the number of industry names that make up the cluster to which the representative industry name belongs is above a threshold. Condition b above, more precisely, means that the number or percentage of users whose industry names are estimated to be duplicates of the representative industry name is below a threshold. More precisely, condition c above means that when comparing a certain attribute (e.g., sales) between users whose representative industry name has been estimated, there are a predetermined number of users for whom the difference in that attribute value is less than or equal to a predetermined value.

[0069] (c) The representative industry names obtained in step 603 may be confirmed by the service provider, and only those approved may be added to the industry list. Alternatively, the obtained representative industry names may be entered into the generation AI module 410 to determine their validity, and only those deemed valid may be added to the industry list. In this case, the current industry list may be entered into the generation AI module 410 as reference information.

[0070] (c) Industry estimation processing 700 The industry estimation process 500 described above uses the user's web information to estimate their industry. However, some users may not have access to web information (especially small and medium-sized enterprises and sole proprietors). Therefore, we will now explain the industry estimation process 700 that is executed when web information is unavailable.

[0071] Figure 7 is a flowchart showing an example of the industry estimation process 700. In the process shown in the figure, information on the target user's trading partners and sales history are used instead of the target user's web information. This process is executed, for example, in step 501 of the industry estimation process 500 if it is determined that the user's URL is not registered. Alternatively, it is executed in the industry estimation process 500 if the generation AI module 410 fails to select or generate an industry more than a predetermined number of times.

[0072] In step 701 of the industry estimation process 700, the first estimation module 315 refers to the business information of the target user and determines whether or not a trading partner user is registered (step 702). If the result of this determination is that a trading partner user is not registered (NO in step 702), the process proceeds to step 708. On the other hand, if the result of this determination is that a trading partner user is registered (YES in step 702), the first estimation module 315 then refers to the business information of each trading partner user (step 703) and determines whether or not the industry of each trading partner user is registered (step 704). If the result of this determination is that the industry is not registered for any trading partner user (NO in step 704), the industry estimation process 500 described above is executed for that trading partner user (step 705) to estimate the industry of that company.

[0073] Figure 10 shows an example of the relationship between the target user and the industry of their trading partner users. Figure 11 shows another example of the relationship between the industry of the target user and the trading partner user. In the example shown in Figure 11, the industry estimation process 500 is executed to estimate the industry of each trading partner user. Note that the "users" shown in Figures 10 and 11 include, for example, businesses, companies, and sole proprietors.

[0074] Once the industry types of multiple trading partner users are identified, the first estimation module 315 inputs the identified industry types into the category estimation model M1 and outputs the industry type of the target user (step 706). The category estimation model M1 referred to here is a pre-trained machine learning model. This category estimation model M1 takes the industry types of trading partner users as input and outputs the industry type of the target user.

[0075] The first estimation module 315 registers the outputted industry as part of the target user's business information (step 707). Then this process ends.

[0076] In step 708, the second estimation module 316 determines whether or not the target user's sales history is registered. If the result of this determination is that no sales history is registered (NO in step 708), the process ends. On the other hand, if the result of this determination is that a sales history is registered (YES in step 708), the second estimation module 316 identifies one or more names of products that were bought and sold for each of the target user and the users who bought and sold products (in other words, trading partners) (step 709). The module then inputs the one or more names of products bought and sold for each trading partner into the category estimation model M2 and outputs the industry of the trading partner user (step 710).

[0077] The category estimation model M2 referred to here is a pre-trained machine learning model. This category estimation model M2 takes the names of one or more products that the target user has bought and sold with a trading partner as input and outputs the industry of the trading partner.

[0078] Figure 12 shows an example of the relationship between the target user and the industry of the trading partner user. In the example shown in the figure, the industry of the trading partner user is estimated from the names of one or more products that were bought and sold. Note that the "users" shown in Figure 12 include, for example, businesses, companies, and sole proprietors.

[0079] The second estimation module 316 estimates the industry of each trading partner user, inputs the estimated industry into the category estimation model M1 described above, and outputs the industry of the target user (step 706). The module then registers the output industry as part of the target user's business information (step 707). The above is an explanation of the industry estimation process 700.

[0080] According to the industry estimation process 700 described above, it is possible to estimate the industry of a user even if the user's web information is unavailable.

[0081] The above process 700 may be modified as follows. (a) In step 706, the industry of the target user may be estimated using a rule-based method instead of the category estimation model M1. Alternatively, the industry of the trading partner user may be input into the generation AI module 410 to estimate the industry of the target user.

[0082] (i) The target user's industry may be directly estimated from the target user's trading history. In this case, the second estimation module 316 inputs one or more product names registered in the target user's trading history into the category estimation model M3 to estimate the target user's industry. The category estimation model M3 referred to here is a pre-trained machine learning model. This category estimation model M3 takes one or more product names registered in the target user's trading history as input and outputs the target user's industry.

[0083] Figure 13 shows an example of the relationship between a target user and the names of the products they bought and sold. Note that the "users" shown in Figure 13 include, for example, businesses, companies, and sole proprietors.

[0084] 2. Variations The above embodiment may be modified as follows. The following modifications may be combined with each other.

[0085] (1) Organization of the industry list In the above embodiment, if an industry added to the industry list is not used for a predetermined period, it may be deleted or replaced with another industry. This helps to streamline the list of industries that are not used frequently.

[0086] In that case, the update module 314 will either remove the category to which the user belongs from the category list or replace it with another category that has a similarity of a predetermined value or higher if the category to which the user belongs is not output by the AI ​​module for other users for a predetermined period of time. Specifically, "category," "AI module," and "category list" here correspond to "industry," "generating AI module 410," and "industry list," respectively.

[0087] The update module 314 monitors the frequency of use of industry names added to the industry list, and if an industry name is not selected by the generation AI module 410 for a predetermined period, it removes it from the industry list. Alternatively, the module replaces the unselected industry name with another similar industry name in the industry list. The other similar industry name is generated by the generation AI module 410, for example.

[0088] (2) Category update In the above embodiment, if the user's industry is not updated for a predetermined period, the industry estimation process 500 described above may be executed to update the industry. This ensures the accuracy of the industry information. Furthermore, by notifying the user of the update, the user can be prompted to check the updated industry information.

[0089] In that case, if the categories used as the user's basic information are not updated for a predetermined period, the first acquisition module 311 acquires new user information, which is information about the user, based on the identification information entered by the user.

[0090] The second acquisition module 312 inputs the newly acquired user information into the AI ​​module and acquires the category to which the user belongs. The module then updates the old industry name with the new industry name.

[0091] The business support server 201 is further equipped with a notification module (not shown). This notification module outputs an alert to the user when a newly acquired category differs from a previously acquired category.

[0092] In this context, "category," "identification information," "user information," and "AI module" specifically correspond to "industry," "information such as business name, corporate number, and company URL," "web page containing user information," and "generated AI module 410," respectively.

[0093] (3) Category In the above embodiment, the user's industry is estimated. This industry is merely one example of a category that can be estimated in the present invention. In another embodiment, other categories may be estimated, such as business type, number of employees, location, and type of target customer (e.g., B2B, B2C, hospitals, students).

[0094] (4) Components of the system Each device constituting the above-described information processing system 200 may be, for example, a mobile device such as a smartphone, tablet, mobile phone, or personal digital assistant (PDA), or a wearable device such as glasses, a wristwatch, or clothing. Each device may also be a stationary or portable computer, or a server located on the cloud or a network. Functionally, each device may be a VR (Virtual Reality) terminal, an AR (Augmented Reality) terminal, or an MR (Mixed Reality) terminal. Alternatively, a combination of multiple such terminals may be used. For example, a combination of one smartphone and one wearable device can logically function as a single terminal. Other types of information processing terminals may also be used.

[0095] (5) Others It should be noted that the present invention is not limited to the embodiments described above, and various modifications are included. For example, the embodiments described above are described in detail to make the present invention easier to understand, and are not necessarily limited to those having all the configurations described. Furthermore, it is possible to replace parts of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add configurations from other embodiments to the configuration of one embodiment. In addition, it is possible to add, delete, or replace parts of the configuration of each embodiment with other configurations.

[0096] Furthermore, each of the above configurations, functions, processing units, and processing means may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. Alternatively, each of the above configurations and functions may be implemented in software by having the processor interpret and execute programs that implement each function. Information such as programs, tables, and files that implement each function can be stored in memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0097] Furthermore, the control lines and information lines shown are those deemed necessary for explanatory purposes, and not all control lines and information lines are necessarily shown in the actual product. In reality, it is safe to assume that almost all components are interconnected. Furthermore, the above-described embodiments disclose at least the configuration described in the claims. [Explanation of symbols]

[0098] 200... Information processing system, 201... Business support server, 202... LLM server, 203... User terminal

Claims

1. A reception method for receiving identification information from users of the service, A first acquisition means for acquiring user information, which is information about the user, based on the input identification information, When the user information is acquired, a second acquisition means inputs the user information into a pre-trained AI module and obtains the category to which the user belongs by having the AI ​​module output it. A means of using the acquired category as the user's basic information, An information processing system equipped with the following features.

2. The second acquisition means, when the user information is acquired, inputs the user information and the category list into the AI ​​module, and if the category to which the user belongs is included in the category list, it has the AI ​​module select and acquire it; if the category to which the user belongs is not included in the category list, it has the AI ​​module generate and acquire it. The information processing system according to claim 1.

3. The system further includes an update means that adds a category to the category list when a predetermined number of users among multiple users of the service have a category to which that user belongs has been generated and acquired by the second acquisition means. The information processing system according to claim 2.

4. When a category is generated for each of the multiple users using the service and acquired by the second acquisition means, the update means further includes clustering the acquired multiple categories and adding the representative category of each cluster to the category list. The information processing system according to claim 2.

5. If the user information is not obtained by the first acquisition means, the system further includes an estimation means for estimating the category to which the user belongs based on information of other users who are business partners of the user or the user's sales history. The information processing system according to claim 1.

6. The update means further includes, when a category to which the user belongs is generated and acquired by the second acquisition means, an update means to add the category to the category list. The update means, if the category to which the user belongs is not output by the AI ​​module for other users for a predetermined period of time, either deletes it from the category list or replaces it with another category having a similarity of a predetermined value or higher. The information processing system according to claim 2.

7. The aforementioned means of use utilizes the acquired category as the user's basic information when the user's approval is obtained for that category. The information processing system according to claim 1.

8. If the category used as the user's basic information is not updated for a predetermined period, the first acquisition means acquires new user information, which is information about the user, based on the input identification information. The second acquisition means inputs the newly acquired user information into the AI ​​module to newly acquire the category to which the user belongs. The system further includes a notification means that outputs an alert to the user if the newly acquired category differs from the previously acquired category. The information processing system according to claim 1.

9. A method of information processing performed by a computer, The steps include receiving identification information from users of the service, The steps include obtaining user information, which is information about the user, based on the input identification information, When the user information is obtained, the user information is input into a pre-trained AI module, and the category to which the user belongs is output and obtained by the AI ​​module. The steps include using the acquired categories as the user's basic information, Information processing methods including

10. On the computer, The steps include receiving identification information from users of the service, The steps include obtaining user information, which is information about the user, based on the input identification information, When the user information is obtained, the user information is input into a pre-trained AI module, and the category to which the user belongs is output and obtained by the AI ​​module. The steps include using the acquired categories as the user's basic information, A program to execute.

Images