Information processing device and information processing method

Abstract

An information processing device according to one embodiment of the present invention includes: a generation unit that generates a prompt including information specifying a provided article group; an extraction unit that inputs the prompt into a large-scale language model and uses the output of the large-scale language model as a basis to extract user characteristics indicating the characteristics of a user corresponding to the provided article group; a filtering unit that filters the designated provided article group in order to obtain a plurality of provided articles corresponding to the user characteristics from among the designated provided article group; and an output unit that outputs a list of the plurality of provided articles.

Description

Information processing device and information processing method

[0001] The present invention relates to the art of recommending offerings.

[0002] There are known technologies for suggesting products to customers based on customer information. For example, Patent Literature 1 discloses a technology for estimating the purpose of purchasing a target product from the customer's purchase history, such as purchase information, and extracting customer groups and product groups from a database according to the purpose of purchase. Patent Literature 2 discloses a technology for selecting candidates using a machine learning model based on customer information and determining products to be suggested. Patent Literature 3 discloses a technology for estimating the deep needs of each user based on the user's behavior when browsing a website and extracting product groups according to those needs.

[0003] International Publication No. 2024 / 054072 JP 2024-53861 A Japanese Patent No. 7482557 A

[0004] In the techniques described in Patent Documents 1 to 3, in order to obtain product proposals, it is necessary to input data in a predetermined data format into the system, which requires cumbersome preparation.

[0005] In contrast to this, the present invention provides a technology that allows desired information to be obtained more easily on a natural language basis.

[0006] One aspect of the present disclosure provides an information processing device having a generation unit that generates a prompt including information that identifies a group of offerings, an extraction unit that inputs the prompt into a large-scale language model and extracts user characteristics that indicate the characteristics of a user corresponding to the group of offerings based on the output of the large-scale language model, a filtering unit that filters the specified group of offerings to obtain multiple offerings from the specified group of offerings that correspond to the user characteristics, and an output unit that outputs a list of the multiple offerings.

[0007] Another aspect of the present disclosure provides an information processing method including the steps of generating a prompt including information identifying a group of offerings, inputting the prompt into a large-scale language model and extracting user characteristics indicating characteristics of a user corresponding to the group of offerings based on the output of the large-scale language model, filtering a plurality of offerings from the specified group of offerings according to the user characteristics, and outputting a list of the plurality of offerings.

[0008] According to the present invention, it is possible to more easily obtain information about an item on a natural language basis.

[0009] 1 is a diagram showing an overview of an information processing system 1 according to an embodiment. A diagram illustrating an example of the functional configuration of the information processing system 1. A diagram illustrating an example of the hardware configuration of an information processing device 10. A flowchart illustrating an overview of the operation of the information processing system 1. A flowchart illustrating an example of details of the user characteristic extraction process in step S2. A diagram illustrating an example of data recorded in the database 111. A diagram illustrating an example of a product list output from the large-scale language model 90. A diagram illustrating an example of a prompt and output. A diagram illustrating an example of a prompt and output. A flowchart illustrating an example of details of the product group selection process in step S3. A diagram illustrating an example of a prompt and output. A flowchart illustrating an example of a filtering process in step S4. A diagram illustrating a specific example of a process for filtering a candidate product group. A diagram illustrating another specific example of a process for filtering a candidate product group. A diagram illustrating an example of a prompt and output. A flowchart illustrating an example of details of the output process in step S5. A diagram illustrating an example of an output summary report.

[0010] 1. Configuration FIG. 1 is a diagram illustrating an overview of an information processing system 1 according to an embodiment. The information processing system 1 provides a service (hereinafter referred to as an "offer selection service") that selects or suggests an item to be offered based on user identification. An "offer" refers to the subject of information provided to a user, such as a product, service, or person. The offer selection service can be used, for example, as follows: Consider a store such as a convenience store. In this example, the offer is a product. The store has shelves on which products are displayed. It can be said that the store's sales depend on which products are displayed on the shelves from among the products currently in circulation. The selection of products to be displayed on the shelves is performed by the store manager. While the store manager can predict the best-selling product based on past experience or trends, predicting the second-best-selling product or products thereafter can be difficult. In particular, in stores that adopt a low-margin, high-volume sales business model, it is difficult to actually display potential products on the shelves and measure their sales due to the cost of verification. The information processing system 1 addresses this problem.

[0011] The information processing system 1 includes an information processing device 10, a user terminal 20, and a large-scale language model 90. The information processing device 10 functions as a server in the offering selection service. The user terminal 20 functions as a client in the offering selection service. The large-scale language model 90 is a machine learning model designed to perform natural language processing (NLP) tasks and is trained on a large amount of text data. As a result, the large-scale language model has the ability to understand and generate human language (i.e., natural language). In this example, the large-scale language model 90 is managed and operated by a business operator separate from the information processing device 10. The information processing device 10, the user terminal 20, and the large-scale language model 90 communicate via a network 9. The network 9 is a computer network such as the Internet. The user of the user terminal 20 is, for example, a store operator. When a user inputs information specifying a store into the user terminal 20, the information processing device 10 selects products to be displayed on shelves based on the user characteristics of the store's potential customers. The customers are users who use the store, i.e., end users. The information processing device 10 outputs a list of selected, i.e., recommended, products to the user terminal 20. The store manager can refer to this list to decide which products to display on the shelves. In this example, the information processing device 10 causes the large-scale language model 90 to recommend products.

[0012] 2 is a diagram illustrating an example of the functional configuration of the information processing system 1. The information processing system 1 includes a characteristic extraction unit 11, a selection unit 12, a filtering unit 13, an output unit 14, a storage unit 18, and a control unit 19. In this example, these elements are implemented in an information processing device 10.

[0013] The characteristic extraction unit 11 extracts user characteristics in response to a request from the user terminal 20. The user characteristics are information indicating the characteristics of a customer when selecting a group of offerings. The user characteristics are obtained, for example, from information regarding past actions taken by the customer (end user) regarding the group of offerings. The user characteristics can be subdivided. In this example, the characteristic extraction unit 11 extracts multiple types of subdivided user characteristics, including a first user characteristic and a second user characteristic.

[0014] In detail, the characteristic extraction unit 11 includes a database 111, a generation unit 112, an extraction unit 113, a generation unit 114, and an extraction unit 115. The database 111 is a database (characteristic database) that records information about past actions taken by multiple users with respect to a group of offerings. The generation unit 112 generates a prompt for extracting a first user characteristic. This prompt includes, for example, information identifying the group of offerings. The group of offerings refers to a portion of offerings that satisfy specific conditions and are selected from all the offerings handled by the information processing system 1. As an example, this group of offerings is a group of products available at a specific convenience store. The prompt is written in, for example, a natural language. The extraction unit 113 inputs this prompt to the large-scale language model 90. The large-scale language model 90 outputs information specified by the prompt. This output is written in a natural language. The extraction unit 113 extracts user characteristics based on the output of the large-scale language model 90. The generation unit 114 generates a prompt for extracting a second user characteristic. The prompt includes information specifying a group of offerings. The extraction unit 115 inputs the prompt to the large-scale language model 90. The large-scale language model 90 outputs the information specified by the prompt. This output is written in natural language. The extraction unit 115 extracts a second user characteristic based on the output of the large-scale language model 90.

[0015] The selection unit 12 selects an offering based on the user characteristics. Specifically, the selection unit 12 has a database 121, a generation unit 122, and an extraction unit 123. The database 121 is a database (product database) that records information about offerings. The generation unit 122 generates a prompt for selecting a candidate offering. This prompt includes at least a portion of the user characteristics. The extraction unit 123 inputs this prompt to the large-scale language model 90. The large-scale language model 90 outputs information indicated by the prompt. This output is written in natural language. The extraction unit 123 extracts candidate offerings based on the output of the large-scale language model 90.

[0016] The filtering unit 13 filters the candidate offerings output from the selection unit 12 using at least a significant portion of the user characteristics output from the characteristic extraction unit 11. In detail, the filtering unit 13 has a generation unit 131 and an extraction unit 132. The generation unit 131 generates a prompt for filtering the candidate offerings. The extraction unit 132 inputs this prompt to the large-scale language model 90. The large-scale language model 90 outputs information indicated by the prompt. This output is written in natural language. The extraction unit 132 extracts a list of offerings based on the output of the large-scale language model 90.

[0017] The output unit 14 processes the list of offerings output from the filtering unit 13 as required and outputs the result. In detail, the output unit 14 has a processing unit 141 and an output unit 142. The processing unit 141 processes the list of offerings. The output unit 142 outputs a final output obtained by the processing. The final output includes a list of recommended offerings. The memory unit 18 stores various data and programs. The control unit 19 performs various controls.

[0018] FIG. 3 is a diagram illustrating an example of the hardware configuration of the information processing device 10. The information processing device 10 is physically configured as a computer including a processor 101, a memory 102, a storage 103, a communication device 104, and a bus connecting these devices. Each of these devices operates using power supplied from a battery (not shown). In the following description, the term "device" can be interpreted as a circuit, device, unit, etc. The hardware configuration of the information processing device 10 may be configured to include one or more of the devices shown in FIG. 3, or may be configured without including some of the devices. Furthermore, the information processing device 10 may be configured by communicating with multiple devices each having a different housing.

[0019] Each function of the information processing device 10 is realized by loading specified software (programs) onto hardware such as the processor 101, memory 102, etc., so that the processor 101 performs calculations, controls communication via the communication device 104, and controls at least one of reading and writing data in the memory 102 and storage 103.

[0020] The processor 101 controls the entire computer by running, for example, an operating system. The processor 101 may be configured as a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. Furthermore, for example, a baseband signal processing unit, a call processing unit, etc. may be realized by the processor 101.

[0021] The processor 101 reads programs (program codes), software modules, data, etc. from at least one of the storage 103 and the communication device 104 into the memory 102 and executes various processes in accordance with these. The programs used are those that cause a computer to execute at least some of the operations described below. The functional blocks of the information processing device 10 may be implemented by a control program stored in the memory 102 and running on the processor 101. Various processes may be executed by one processor 101, or may be executed simultaneously or sequentially by two or more processors 101. The processor 101 may be implemented by one or more chips. The programs may be transmitted to the information processing device 10 via a telecommunications line.

[0022] The memory 102 is a computer-readable recording medium and may be configured by, for example, at least one of a read-only memory (ROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a random access memory (RAM), etc. The memory 102 may also be called a register, a cache, a main memory (primary storage device), etc. The memory 102 can store executable programs (program codes), software modules, etc. for implementing the method according to this embodiment.

[0023] Storage 103 is a computer-readable recording medium, and may be composed of at least one of, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy (registered trademark) disk, a magnetic strip, etc. Storage 103 may also be called an auxiliary storage device.

[0024] The communication device 104 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also called, for example, a network device, a network controller, a network card, or a communication module.

[0025] Each device, such as the processor 101 and the memory 102, is connected by a bus for communicating information. The bus may be configured using a single bus, or different buses may be used between each device.

[0026] The information processing device 10 may be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, the processor 101 may be implemented using at least one of these pieces of hardware.

[0027] In this example, the programs stored in the storage 103 include a program (hereinafter referred to as a "server program") for causing the computer to function as the information processing device 10 (i.e., a server) in the information processing system 1. When the processor 101 is executing the server program, the processor 101 is an example of the characteristic extraction unit 11, the selection unit 12, the filtering unit 13, and the control unit 19. At least one of the memory 102 and the storage 103 is an example of the storage unit 18. The processor 101 and the communication device 104 are an example of the output unit 14. The characteristic extraction unit 11, the selection unit 12, and the filtering unit 13 are so-called RAG (Retrieval-Augmented Generation).

[0028] 2. Operation FIG. 4 is a flowchart showing an outline of the operation of the information processing system 1. Here, the explanation will focus on the processing of the information processing device 10. The processing of FIG. 4 is initiated, for example, when the user terminal 20 transmits a request for a list of offerings recommended by the information processing device 10. Here, an example is used in which a store in a convenience store chain is used. In this example, the operator of the store is the user, and the products stocked at the store are the offerings.

[0029] In step S1, the characteristic extraction unit 11 of the information processing device 10 receives a request for providing a list from the user terminal 20. In this example, this request is text data written in natural language and includes information that identifies a target store. The target store is the store that is the target of the request among multiple stores. As a specific example, this request is a natural sentence such as "What products would be suitable for store Y of convenience store chain X?". Of these, the character string "store Y of convenience store chain X" is an example of information that identifies the target store.

[0030] In step S2, the characteristic extraction unit 11 extracts user characteristics related to the target store.

[0031] 5 is a flowchart illustrating the details of the user characteristic extraction process in step S2. In this flow, the characteristic extraction unit 11 causes the large-scale language model 90 to extract user characteristics. In this example, two types of subdivided user characteristics are extracted: taste characteristics (an example of a first user characteristic) and behavioral characteristics (an example of a second user characteristic). The taste characteristics are information indicating the taste preferences of a customer. The behavioral characteristics are information indicating the behavior of a customer when using a product, such as a use case.

[0032] In step S21, as preprocessing for extracting user characteristics related to the target store, the generation unit 112 of the characteristic extraction unit 11 extracts data related to the target store from the database 111. Hereinafter, the data extracted in step S21 will be referred to as "specific data."

[0033] FIG. 6 is a diagram illustrating an example of data recorded in the database 111. In this example, the database 111 includes a behavioral sequence, a questionnaire, customer data, and purchase data. The behavioral sequence indicates the transition of location information for each user. Specifically, the behavioral sequence includes a set of timestamps and coordinate data for each of multiple users. This data set includes multiple sets of data corresponding to multiple times. Each data set is associated with a user identifier. The questionnaire indicates the results of a survey conducted on a certain topic for multiple users. Specifically, the behavioral sequence includes a set of questions and their answers for each of multiple users. This data set includes multiple sets of data corresponding to multiple questions. Each data set is associated with a user identifier. The customer data indicates the attributes of customers who have used each store. Specifically, the customer data includes statistical information for each of multiple stores regarding the attributes of customers of that store. As a more detailed example, the customer data may include information such as the following about customers of Store Y of convenience store chain X: 56% male, 44% female; and 14% in their 20s, 17% in their 30s, 35% in their 40s, 27% in their 50s, and 7% over 60. The purchase data indicates purchase information for each user. Specifically, the purchase data includes a set of timestamp, product identification information, price, and store identification information for each of multiple users. This data set includes multiple sets of data corresponding to multiple purchase behaviors. Each data set is associated with a user identifier. All of these are examples of information regarding the structures multiple users have taken in the past with respect to a group of offerings. In particular, the purchase data is an example of information regarding the fact that a user selected a single offering. Note that FIG. 6 is merely an example; other data may be recorded in database 111, or some of this data may be omitted.

[0034] Here, the specific data is data that includes, in the case of a behavioral pattern, a data set including location information corresponding to the target store, in the case of a questionnaire, a data set including questions related to the target store, in the case of customer data, customer data of the target store, and in the case of purchase data, purchase data at the target store. Here, an example will be described in which purchase data is used as the specific data.

[0035] Specifically, the processing of step S21 is, for example, as follows. First, the generation unit 112 extracts purchase data for a predetermined period of time at the target store from the database 111. Next, the generation unit 112 generates a prompt for outputting the specific data. This prompt includes, for example, a natural sentence such as, "Please count the sales numbers from the attached purchase data and output a list of products ranked 1st to 20th in order of sales number." In one example, this prompt is a fixed form. Next, the generation unit 112 inputs the generated prompt to the large-scale language model 90. The large-scale language model 90 outputs information specified by the prompt. Here, the large-scale language model 90 outputs a list of products ranked 1st to 20th in order of sales number among the products included in the purchase data. In this example, this product list is the specific data. The extraction unit 113 obtains the output, i.e., the specific data. This completes the preprocessing.

[0036] 7 is a diagram illustrating an example of a product list output from the large-scale language model 90. In this example, the product list includes a ranking number and a product name. In this example, the number 0 is assigned to the top-selling product, and the number 19 is assigned to the 20th-selling product. This is a group of top-selling products that have actually sold well in the target store in the past.

[0037] Referring again to Figure 5, in this example, taste characteristics and behavioral characteristics are extracted separately. Here, an example in which taste characteristics and behavioral characteristics are extracted in order will be described, but the order in which these user characteristics are extracted is arbitrary. Alternatively, taste characteristics and behavioral characteristics may be extracted in parallel.

[0038] In step S22, the generation unit 112 generates a prompt for extracting taste characteristics from the product list. In step S23, the extraction unit 113 inputs the generated prompt into the large-scale language model 90. In step S24, the extraction unit 113 obtains the output from the large-scale language model 90, i.e., the taste characteristics.

[0039] 8 is a diagram illustrating an example of a prompt and an output. The output obtained here is adopted as the taste characteristics. The taste characteristics are described in natural language and, in this example, include consideration of the customer's (the "consumer" in the example shown).

[0040] Referring again to FIG. 5 , in step S25, the generation unit 114 generates a prompt for causing the large-scale language model 90 to estimate a behavioral characteristic, i.e., a use case. In step S26, the extraction unit 115 inputs this prompt and the specific data extracted in step S1 to the large-scale language model 90. The large-scale language model 90 outputs information indicated by this prompt. In this example, the output of the large-scale language model 90 includes natural sentences indicating the use case to be estimated. In step S27, the extraction unit 115 acquires this output, i.e., the behavioral characteristic.

[0041] 9 is a diagram illustrating an example of a prompt and output. In this example, the prompt includes natural sentences that allow the user to infer 10 or more usage scenarios or use cases. The output includes 12 natural sentences that indicate use cases, i.e., behavioral characteristics.

[0042] Referring again to Fig. 4, in step S3, the selection unit 12 selects a group of candidate products according to the user characteristics.

[0043] FIG. 10 is a flowchart illustrating the details of the product group selection process in step S3. In step S31, the generation unit 122 generates a prompt that causes the large-scale language model 90 to select a candidate product group. The candidate product group is an example of a specified offering group, and is a group of products selected from the population as matching at least a portion of the user characteristics (behavioral characteristics in this example) obtained in step S2. In this example, the population is a group of products that are available at the target store among all products handled by the information processing system 1. The database 121 records product information for all products handled by the information processing system 1. The product information includes product attributes, such as product name, type, price, and information indicating the store where the product is sold. The attribute "type" indicates the product classification, such as food, beverage, stationery, newspaper, and magazine. The product classification may be hierarchical. The attribute "store" includes information indicating the store where the product can be purchased. The store from which the product can be purchased may be identified in any way, such as by the store's brand name (e.g., "Convenience Store Chain X"), a specific store name (e.g., "Tokyo Station Store"), or a region (e.g., City B, Prefecture A).

[0044] In step S32, the extraction unit 123 inputs this prompt to the large-scale language model 90. The large-scale language model 90 outputs information indicated by this prompt. In this example, the output of the large-scale language model 90 includes a list of candidate products. In step S33, the extraction unit 123 acquires this output, i.e., the candidate products.

[0045] FIG. 11 is a diagram illustrating an example of a prompt and output. In this example, the prompt specifies that the output content should include, for each product, the product name, product overview, and selection reason. Furthermore, the prompt specifies a minimum number of candidate products to be included in the output candidate product group. The corresponding user characteristic, in this example, a use case, i.e., a behavioral characteristic, is substituted for the portion of the prompt labeled "[behavior#prompt]." In response, the output includes a list of candidate product groups.

[0046] Referring again to Fig. 4, in step S4, the filtering unit 13 filters the candidate product group to select a product group according to the user characteristics.

[0047] 12 is a flowchart illustrating the details of the filtering process in step S4. In step S41, the generation unit 131 of the filtering unit 13 generates a prompt to cause the large-scale language model 90 to filter the group of candidate products, i.e., to select products that meet the conditions from the group of candidate products. In step S42, the extraction unit 132 inputs this prompt to the large-scale language model 90. The large-scale language model 90 outputs information indicated by this prompt. In this example, the output of the large-scale language model 90 includes a list of the final filtered products. In step S43, the extraction unit 132 acquires this output.

[0048] 13 is a diagram showing a specific example of a process for filtering a group of candidate products. In this example, the large-scale language model 90 selects a specified number (m) of products from all (n) candidate products in a single selection (m is a natural number greater than or equal to 3). The prompt may include a natural sentence such as, "Please select m products from the list of products that you would like to recommend based on the user's characteristics, and then sort them in order of recommendation."

[0049] FIG. 14 shows another specific example of a process for filtering candidate products. In this example, the large-scale language model 90 selects m products from a group of n candidate products using a survival tournament (or knockout tournament). This process is performed, for example, as follows: First, all (n) candidate products are divided into product pairs consisting of two products. (n / 2) pairs are created (rounded up if n is an odd number). Next, for each pair, the large-scale language model 90 determines which one better matches the user characteristics. In other words, the large-scale language model 90 makes an alternative selection. Of the paired products, the one that better matches the user characteristics survives. When one round is completed, it is determined whether prediction is complete, i.e., whether a specified number of products have been selected. If prediction is not complete, the process proceeds to the next round, and new product pairs are created using the remaining products. In the next round, for each new product pair, the large-scale language model 90 determines which one better matches the user characteristics. This process is repeated until m products are determined. When the prediction is complete, the results are aggregated and output.

[0050] FIG. 15 is a diagram illustrating an example of a prompt and output. In this example, the prompt includes natural language for filtering a group of candidate products in one go. The part written as "[taste#prompt]" is substituted with the corresponding user characteristic, in this example, the taste characteristic. The output includes a list of filtered products. In addition to the product names, this list includes natural language for explaining why the products are recommended.

[0051] Referring again to Figure 4, in step S5, the output unit 14 processes the list of filtered products.

[0052] 16 is a flowchart illustrating the details of the output process in step S5. In step S51, the processing unit 141 of the output unit 14 generates a summary report as the final output. In this example, the summary report includes information on customer trends. In step S52, the output unit 142 outputs a product list and the summary report.

[0053] 17 is a diagram illustrating an example of a summary report that is output. The summary report includes a list of recommended products (top row of FIG. 17) and use cases (bottom row of FIG. 17). According to this example, a user (store manager) can obtain a list of recommended products to purchase at low cost.

[0054] 3. Modifications The present invention is not limited to the above-described embodiment, and various modifications are possible. Some modifications will be described below. Two or more of the following features may be applied in combination.

[0055] (1) Target Store The method by which the information processing system 1 identifies a target store is not limited to the method exemplified in the embodiment. The target store may be specified not by inputting natural language, but by the store's identification number or its location on a map.

[0056] (2) User Characteristics The user characteristics used in the information processing system 1 are not limited to the information exemplified in the embodiment. In the embodiment, an example has been described in which two types of user characteristics, taste characteristics and behavioral characteristics, are used. However, other types of user characteristics may be used instead of or in addition to either or both of these. As other types of user characteristics, any characteristics may be used, such as characteristics related to purchase amounts, characteristics related to time periods of purchasing behavior, and characteristics related to user attributes such as age or gender.

[0057] Furthermore, the user characteristics extracted in step S2 do not have to be directly related to the target store. For example, user characteristics related to other stores, such as stores of the same or similar brand, stores operated by the same or similar company, or stores in a certain area, may be extracted.

[0058] (3) Method for Extracting Specific Data The method for extracting specific data in step S21 is not limited to the method exemplified in the embodiment. For example, the specific data may be extracted using a database and a database function without using the large-scale language model 90.

[0059] (4) Large-Scale Language Model In the embodiment, an example has been described in which the same large-scale language model 90 is used four times (steps S23, S26, S32, and S42) for one target store. At least one of these times may use a different large-scale language model (independently trained or fine-tuned). Furthermore, some of these processes may be replaced with processes that do not use a large-scale language model.

[0060] In the embodiment, an example has been described in which the large-scale language model 90 is managed and operated by a business operator separate from the information processing device 10. However, the large-scale language model 90 may be managed and operated by the same business operator as the information processing device 10. In this case, the large-scale language model 90 may be implemented in the information processing device 10.

[0061] (5) Prompts The prompts presented in the embodiments are merely examples, and the prompts used in the information processing system 1 are not limited to these examples. The content of the prompts and the output format from the large-scale language model 90 can be set arbitrarily.

[0062] (6) Summary Report The final output of the information processing system 1 is not limited to the summary report exemplified in Fig. 17. The final output does not need to include other information as long as it includes at least a list of recommended products.

[0063] (7) Targets of Application The offerings handled by the information processing system 1 are not limited to products in stores. The information processing system 1 can also be applied to the following offerings, for example: (a) Staffing. In this case, the offering is human resource information. The user is a workplace or department (the person in charge of hiring). The user characteristics are information about the workplace or department. When a workplace or department is recruiting new personnel, the information processing system 1 provides human resource information of personnel recommended for the recruitment. (b) Personnel placement or personnel transfer. In this case, the offering is human resource information. The user is a human resources officer. The user characteristics are information about the department. When personnel are transferred between departments in a company, the information processing system 1 provides human resource information of personnel recommended to each department. (c) Software technology selection. In this case, the offering is software element information. The user is a software technology company. The user characteristics are information about software elements. The software element information is information about software elements. The software elements are software libraries or frameworks. When a software engineer develops software, the information processing system 1 provides software elements recommended for the functions. (d) Written expression. In this case, the offering is a written proposal. The user is the recipient of the document. For example, the writing is the copy of an advertisement and the user is the advertisement's target customer. The characteristic data is information about the target customer.

[0064] (8) Functional Configuration and Hardware Configuration The functional configuration and hardware configuration of the information processing system 1 described in the embodiment are merely examples. For example, some of the functional elements shown in FIG. 2 may be omitted. Furthermore, the information processing system 1 may have any hardware configuration as long as it can implement the required functions. For example, the information processing device 10 may be a so-called virtual server (or cloud).

[0065] (9) Others The various programs executed by the processor 101 may be provided by downloading via a network such as the Internet, or may be provided in a state recorded on a computer-readable non-transitory recording medium such as a DVD-ROM. Each processor may be, for example, a CPU, an MPU (Micro Processing Unit), or a GPU (Graphics Processing Unit).

[0066] The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and / or software. Furthermore, the method for realizing each functional block is not particularly limited. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are directly or indirectly connected (e.g., wired, wireless, etc.) and these multiple devices. The functional block may also be realized by combining software with the single device or multiple devices.

[0067] Functions include, but are not limited to, judgment, determination, judgment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0068] For example, the information processing device 10 according to an embodiment of the present disclosure may function as a computer that performs the processing of the present disclosure.

[0069] Each aspect / embodiment described in the present disclosure may be applied to at least one of systems using LTE (Long Term Evolution), LTE-Advanced (LTE-A), SUPER 3G, IMT-Advanced, 4G (4th generation mobile communication system), 5G (5th generation mobile communication system), FRA (Future Radio Access), NR (New Radio), W-CDMA (registered trademark), GSM (registered trademark), CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), IEEE 802.20, UWB (Ultra-Wideband), Bluetooth (registered trademark), or other suitable systems, and next-generation systems enhanced based on these. Furthermore, a combination of multiple systems (e.g., a combination of at least one of LTE and LTE-A with 5G) may also be applied.

[0070] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described in this disclosure may be changed unless it is consistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0071] Input and output information may be stored in a specific location (for example, memory) or may be managed using a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0072] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0073] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0074] Software, whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise, should be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc. Additionally, software, instructions, information, etc. may be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)), and / or wireless technologies (such as infrared, microwave), then such wired and / or wireless technologies are included within the definition of a transmission medium.

[0075] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof. Note that terms described in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings.

[0076] Furthermore, the information, parameters, etc. described in this disclosure may be expressed using absolute values, may be expressed using relative values ​​from a predetermined value, or may be expressed using other corresponding information.

[0077] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0078] As used in this disclosure, any reference to an element using a designation such as "first," "second," etc. does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must in some way precede the second element.

[0079] The "unit" in the configuration of each of the above devices may be replaced with "means," "circuit," "device," or the like.

[0080] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Furthermore, when the term "or" is used in this disclosure, it is not intended to be an exclusive or.

[0081] In this disclosure, where articles are added by translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0082] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different."

[0083] 1...information processing system, 9...network, 10...information processing device, 11...characteristic extraction unit, 12...selection unit, 13...filtering unit, 14...output unit, 18...storage unit, 19...control unit, 20...user terminal, 90...large-scale language model, 101...processor, 102...memory, 103...storage, 104...communication device, 111...database, 112...generation unit, 113...extraction unit, 114...generation unit, 115...extraction unit, 121...database, 122...generation unit, 123...extraction unit, 131...generation unit, 132...extraction unit, 141...processing unit, 142...output unit

Claims

1. An information processing device having: a generation unit that generates a prompt including information that identifies a group of offerings; an extraction unit that inputs the prompt into a large-scale language model and extracts user characteristics that indicate characteristics of a user corresponding to the group of offerings based on the output of the large-scale language model; a filtering unit that filters a specified group of offerings to obtain a plurality of offerings that correspond to the user characteristics from the specified group of offerings; and an output unit that outputs a list of the plurality of offerings.

2. The information processing device described in claim 1, wherein the extraction unit inputs, in addition to the prompt, information extracted from a database recording information regarding past actions taken by multiple users regarding the group of offerings into the large-scale language model, the prompt includes a string that uses the extracted information to infer information regarding the actions of the multiple users regarding the group of offerings, and the extraction unit extracts information regarding the actions of the multiple users regarding the group of offerings as the user characteristics.

3. The information processing device of claim 2, wherein the information regarding past behavior of the plurality of users regarding the group of offerings includes information regarding the fact that the plurality of users selected one offering from the group of offerings, and the user characteristics include information indicating a use case in which the plurality of users uses one offering selected from the group of offerings.

4. The information processing device according to claim 2, wherein the information relating to past actions taken by the plurality of users with respect to the group of offerings includes information relating to the fact that the plurality of users selected one offering from the group of offerings, and the user characteristics include information indicating the reasons why the plurality of users selected one offering from the group of offerings.

5. The information processing device described in claim 2, wherein the database records multiple types of information regarding past actions taken by multiple users regarding the group of offerings, and the extraction unit inputs at least two types of information from the multiple types of information into the large-scale language model.

6. The information processing device according to claim 1, wherein the user characteristics are described in natural language.

7. The information processing device according to claim 1, wherein the list includes, for each of the offerings included in the list, a natural sentence indicating the reason for selecting the offering.

8. The information processing device according to claim 1, wherein the filtering unit selects two offerings from the specified group of offerings, and selects the plurality of offerings by repeating a process of having the large-scale language model select the offering from the two selected offerings that is more suitable for the user characteristics.

9. The information processing device according to claim 1, wherein the plurality of offerings are m offerings (m is a natural number greater than or equal to 3), and the filtering unit selects the plurality of offerings by having the large-scale language model select n offerings that match the user characteristics from the specified group of offerings.

10. An information processing method comprising the steps of: generating a prompt including information identifying a group of offerings; inputting the prompt into a large-scale language model and extracting user characteristics indicating the characteristics of a user corresponding to the group of offerings based on the output of the large-scale language model; filtering a plurality of offerings from the specified group of offerings according to the user characteristics; and outputting a list of the plurality of offerings.

Images