Information processing device, information processing method, and information processing program

Abstract

Update the information contained in the response with information from the official website, etc. [Solution] The system comprises: a generation unit that inputs a first prompt to a first trained language model to generate a response to a user's request and generates a response sentence; an acquisition unit that inputs a second prompt, which includes search conditions for searching for first information contained in the response sentence from information publicly available on the Internet, to a second trained language model and obtains search results; and an output control unit that updates the response sentence based on the search results and outputs it to the user.

Description

Technical Field

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and an information processing program.

Background Art

[0002] Correction apparatuses that automatically detect input errors in facility names and addresses in map data and correct data errors are known in the art.

Prior Art Literature

Patent Literature

[0003]

Patent Literature 1

Summary of the Invention

Problem to be Solved by the Invention

[0004] By the way, when searching for a facility that meets user needs using a large language model (LLM), the address information of the facility presented by the LLM may be incorrect. In the prior art, in such a case, it is conceivable to use the web to check whether the address of the facility is correct or not, but since the web is full of information of uncertain authenticity, there may occur cases where the check is performed based on incorrect information.

[0005] The present disclosure has been made in view of the above circumstances, and an object thereof is to provide an information processing apparatus, an information processing method, and an information processing program capable of updating a response sentence with correct information.

Means for Solving the Problem

[0006] To solve the above-mentioned problems and achieve the objective, the information processing device according to this disclosure is characterized by comprising: a generation unit that inputs a first prompt to a first trained language model to generate a response to a user's request and generates a response sentence; an acquisition unit that inputs a second prompt including search conditions for searching for first information contained in the response sentence from information publicly available on the Internet to a second trained language model and obtains search results; and an output control unit that updates the response sentence based on the search results and outputs it to the user.

[0007] Furthermore, the information processing method relating to this disclosure is an information processing method executed by an information processing device, and is characterized by comprising: a generation step of inputting a first prompt to a first trained language model to generate a response to a user's request to generate a response sentence; an acquisition step of inputting a second prompt including search conditions for searching for first information contained in the response sentence from information publicly available on the Internet to a second trained language model to obtain search results; and an output control step of updating the response sentence based on the search results and outputting it to the user.

[0008] Furthermore, the information processing program relating to this disclosure is characterized by causing a computer to execute a generation procedure for generating a response sentence by inputting a first prompt to a first trained language model to generate a response to a user's request; an acquisition procedure for obtaining search results by inputting a second prompt, which includes search conditions for searching for first information contained in the response sentence from information publicly available on the Internet, to a second trained language model; and an output control procedure for updating the response sentence based on the search results and outputting it to the user. [Brief explanation of the drawing]

[0009] [Figure 1] Figure 1 shows an example of how information processing equipment is used. [Figure 2] Figure 2 shows an example of an application of an information processing device. [Figure 3] Figure 3 is a block diagram showing an example configuration of an information processing device according to Embodiment 1. [Figure 4] Figure 4 shows an example of the process for generating a response statement based on a request according to Embodiment 1. [Figure 5] Figure 5 shows an example of information retrieval processing from the Web using the second trained language model according to Embodiment 1. [Figure 6] Figure 6 shows an example of the process of obtaining search results from the second trained language model according to Embodiment 1 and updating the response sentence. [Figure 7] Figure 7 is a flowchart showing an example of the information processing flow of the information processing device according to Embodiment 1. [Figure 8] Figure 8 shows an example of a process for modifying and updating the response statement based on the search results according to Embodiment 2. [Figure 9] Figure 9 shows an example of a process that adds and updates information to the response statement based on the search results according to Embodiment 2. [Figure 10] Figure 10 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing device in Embodiments 1, 2, and 3. [Modes for carrying out the invention]

[0010] Embodiments of the information processing apparatus, information processing method, and information processing program relating to this disclosure will be described in detail below with reference to the drawings. However, this disclosure is not limited to the embodiments described below. Furthermore, each embodiment can be combined as appropriate within a non-consistent scope.

[0011] [Embodiment 1] Figure 1 shows an example of how the information processing device 1 is used. As shown in Figure 1, the information processing device 1 is a computer or similar device installed in a vehicle that responds to requests from the vehicle's user (hereinafter sometimes simply referred to as "user"), who is either the driver or a passenger. For example, the vehicle's user activates the information processing device 1 installed in the vehicle and makes a request to the information processing device 1, such as searching for a store. Upon receiving this request, the information processing device 1 responds to the user's request.

[0012] Note that the information processing apparatus 1 may be a dedicated navigation device built in or mounted on a vehicle. For example, the information processing apparatus 1 may be a single apparatus having a navigation function, a display function, and an interactive function.

[0013] Furthermore, a user can install a predetermined application on a portable terminal device (e.g., a smartphone, a tablet terminal, a notebook PC, a desktop PC, a PDA, etc.) that the user uses on a daily basis, and substitute this as the information processing apparatus 1. For example, when a portable terminal device is utilized as the information processing apparatus 1, it is installed on a vehicle dashboard or the like during driving.

[0014] (Application Example) Incidentally, the information processing apparatus 1 is not limited to an in-vehicle device. The information processing apparatus 1 can be used not only for in-vehicle devices but also for various devices capable of interacting with a user. Application examples of the information processing apparatus 1 will be illustrated here.

[0015] FIG. 2 is a diagram illustrating an example of a usage status of the information processing apparatus 1. For example, as shown in (1) of FIG. 2, the information processing apparatus 1 can be used as a smart speaker. In this case, a user activates a smart speaker having functions equivalent to those of the information processing apparatus 1, and makes a request such as a store search to the smart speaker. The smart speaker that has received the request responds to the user's request.

[0016] Furthermore, for example, as shown in (2) of FIG. 2, the information processing apparatus 1 can be used as an interactive application or the like for a smartphone. In this case, a user activates an interactive application or the like including functions equivalent to those of the information processing apparatus 1, and makes a request such as a store search to the smartphone. The interactive application or the like that has received the request responds to the user's request. Note that the information processing apparatus 1 is not limited to smartphones, and can be similarly applied to mobile phones, tablets, and the like.

[0017] Further, for example, as shown in (3) of FIG. 2, the present invention can be used for a personal computer as the information processing apparatus 1. In this case, a user activates interactive software or the like that includes functions equivalent to those of the information processing apparatus 1, and issues a request such as a store search to the personal computer. The interactive software or the like that has received the request responds to the user's current request. Note that the information processing apparatus 1 is not limited to personal computers, and can be similarly applied to various computers such as game machines having an Internet communication function.

[0018] Incidentally, in conventional information processing apparatuses, even when a facility matching a user's needs is searched using a trained language model in response to a user's request, the facility information presented by the trained language model may differ from current information due to closure, relocation, or other reasons. For this reason, conventional information processing apparatuses check whether facility information is correct using the Web. However, since the Web is full of information of uncertain authenticity, a situation occurs in which whether the facility information presented by the trained language model is correct is determined based on incorrect information on the Web.

[0019] Therefore, in the present disclosure, an information processing apparatus 1 that can generate a prompt including search conditions for searching an official website, a reliable platform having a certain number or more of users, and the like from a facility name included in a response sentence, input the prompt to a trained language model, and update the response sentence based on a search result will be described.

[0020] Specifically, the information processing apparatus 1 of the present disclosure acquires a response sentence corresponding to a user's request generated by a first trained language model 61. The information processing apparatus 1 generates a prompt including search conditions for searching an official website or the like that provides information such as the address of a facility included in the response sentence, inputs the prompt to a second trained language model 62, and acquires a search result. Then, the information processing apparatus 1 updates information such as the address of the facility included in the response sentence based on information such as the address of the facility included in the search result.

[0021] For example, a user might request information processing device 1 to "tell me the opening hours of the nearby dry cleaners." The information processing device 1 then inputs this request into the first trained language model 61 and retrieves the response statement generated by the first trained language model 61: "The following dry cleaners are located nearby, and their opening hours are as follows: A. Dry Cleaning: 24 hours (staffed from 10:00 to 18:00) B. Dry Cleaning: 11:00 to 20:00."

[0022] The information processing device 1 then generates a prompt that includes search criteria such as the official websites of the facility names "A Cleaning" and "Cleaning B" included in the response sentence, and a platform where users can post facility ratings and reviews, and inputs it into the second trained language model 62. The second trained language model 62 then outputs the search results: "According to the official website, A Cleaning's business hours are 24 hours (staffed service from 10:00 to 20:00). According to the official website, Cleaning B's business hours are from 11:00 to 20:00." The information processing device 1 retrieves these search results, updates the business hours of "A Cleaning" and "Cleaning B" included in the response sentence based on the respective search results, and outputs them to the user.

[0023] In this way, the information processing device 1 can update the facility information included in the response statement with the latest information included in the search results to provide accurate information to the user.

[0024] (Configuration of Information Processing Device 1) Next, an example of the functional configuration of the information processing device 1 according to Embodiment 1 will be described. Figure 3 is a block diagram showing an example of the configuration of the information processing device 1 according to Embodiment 1.

[0025] As shown in Figure 3, the information processing device 1 consists of a communication unit 2, an input unit 3, an output unit 4, a control unit 5, a storage unit 6, and the like.

[0026] The communication unit 2, under the control of the control unit 5, connects to the server via the network and sends and receives various types of information. For example, if a trained language model or the like, contained in the storage unit 6 of the information processing device 1, is stored in an external device such as a server, the communication unit 2 sends and receives data via the network.

[0027] The input unit 3 includes a microphone for capturing sound and generates audio information based on the captured sound. The input unit 3 also includes a display for receiving operation input and generates input information based on the entered text and other information. For example, under the control of the control unit 5, the input unit 3 generates audio information when it detects a voice request from the user, such as a store search. The input unit 3 then outputs the generated audio information to the control unit 5.

[0028] The output unit 4 outputs information under the control of the control unit 5. For example, the output unit 4 includes an audio output device and outputs the information input from the control unit 5 as audio. The output unit 4 also includes a display device and, in addition to audio output, may display characters, symbols, etc., on an in-vehicle display of the information processing device 1 or on various displays connected to the information processing device 1.

[0029] The storage unit 6 stores various programs executed by the control unit 5 (information processing programs according to this embodiment), as well as data necessary for the control unit 5 to perform processing. As shown in Figure 3, the storage unit 6 includes a first trained language model 61, a second trained language model 62, and the like.

[0030] The first trained language model 61 is a language model trained to generate and output response sentences according to input prompts. For example, if the first trained language model 61 receives a prompt requesting the generation of a response sentence to a user's request, such as a store search, it will output a response sentence according to the instructions in the prompt.

[0031] The second trained language model 62 is a language model trained to search for information and output search results according to the input prompt. For example, when the second trained language model 62 receives a prompt that includes the name of a facility, search criteria such as an official website or platform, and the information to be retrieved, it will search for the official website and platform of the relevant facility from the internet or other sources and output the retrieved information.

[0032] Note that the first trained language model 61 and the second trained language model 62 may be the same. For example, large language models (LLMs) can be used for the first trained language model 61 and the second trained language model 62. Also, the first trained language model 61 and the second trained language model 62 may be a group of language models that includes multiple language models. Note that the configuration and type of the models are not particularly limited, as long as the first trained language model 61 and the second trained language model 62 are machine learning models that have been trained to handle the processing described later.

[0033] The control unit 5 is implemented by a controller such as a CPU (Central Processing Unit) or MPU (Micro Processing Unit) executing various programs stored in the memory unit 6, thereby controlling the operation of the entire information processing device 1. Note that the control unit 5 may be composed of integrated circuits other than a CPU or MPU, such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array).

[0034] As shown in Figure 3, this control unit 5 includes a generation unit 51, an acquisition unit 52, an output control unit 53, and the like.

[0035] When the generation unit 51 receives a request from the user, such as a store search, via the input unit 3, it requests the first trained language model 61 to generate a response sentence corresponding to the user's request.

[0036] Here, we will explain the generation of response sentences using Figure 4. Figure 4 is a diagram showing the generation of a response sentence using the first trained language model 61 after receiving a user request.

[0037] As shown in Figure 4(1), the generation unit 51 receives a user request such as a store address search according to Embodiment 1 and generates a prompt requesting the first trained language model 61 to generate a response sentence. In the example in Figure 4(1), when the generation unit 51 receives the request "Please tell me the address of Western Restaurant A", it generates the prompt "Please tell me the address of Western Restaurant A." and inputs it to the first trained language model 61. Then, as shown in Figure 4(2), the generation unit 51 outputs response sentence A "The address of Western Restaurant A is as follows. Address: F2-18-17, Kodaira City, Tokyo" based on the output result output by the first trained language model 61.

[0038] Returning to Figure 3, the acquisition unit 52 generates a prompt that includes the information contained in the response sentence and the conditions for searching the official website or platform, and inputs it into the second trained language model 62 to obtain the search results output from the second trained language model 62. The information contained in the response sentence includes the facility name or address, business hours, etc., and is included in the first information of this disclosure. This prompt is included in the first prompt of this disclosure.

[0039] Furthermore, if the acquisition unit 52 cannot obtain information from an official website or platform due to the absence of an official website or other reasons, it outputs the response sentence output by the first trained language model 61 to the output unit 4 and responds to the user.

[0040] Here, using Figures 5 and 6, we will explain the process of generating a prompt to search for facility information contained in the response sentence and obtaining search results from the second trained language model 62. Figure 5 is a diagram showing the flow of processing for searching for information based on the facility name contained in the response sentence, and Figure 6 is a diagram showing the process of obtaining the search results (process (1) in Figure 6).

[0041] For example, as shown in Figure 5, the acquisition unit 52 generates a prompt to search the official homepage and platform for the "address" of the facility name "Western Restaurant A" included in the response sentence A shown in Figure 4, and inputs it into the second trained language model 62. This prompt is included in the second prompt of this disclosure.

[0042] Specifically, the acquisition unit 52 includes the following prompt in Figure 5B: "Please search for the official website of Western-style restaurant A based on the following conditions, and obtain the address if at least two of the conditions are met. If none are found, please obtain the address from the information publicly available on the internet for Western-style restaurant A on a platform that meets at least two of the following conditions. In that case, please also let us know if there is a possibility of temporary closure or permanent closure based on recent posts about the establishment."

[0043] Furthermore, the acquisition unit 52 includes the following conditions in the prompt as an example of conditions for searching the official website, as shown in Figure 5B. 1. The URL or homepage title includes the facility name or part of it. 2. Contact information for the facility is provided. 3. Information has been posted or updated on the official website within the last six months. 4. There is a link to order online.

[0044] Furthermore, when searching for an official website, you may also include criteria such as "having a privacy policy and terms of use," "having links to social media," "official operating information," "clear and consistent brand, logo, and store information," "a reliable SSL certificate," and "including copyright information."

[0045] Furthermore, the acquisition unit 52 may include conditions for searching for a platform, as shown in Figure 5B. As an example of conditions for searching for a platform, the acquisition unit 52 includes the following conditions in the prompt. 1. The number of registered users must be 40 million or more. The website lists more than 24,500 stores. 3. There are over 1 million user-submitted reviews and comments.

[0046] As a result, the acquisition unit 52 generates the prompt "Official homepage of Western-style restaurant A..." shown in Figure 5B and inputs it into the second trained language model 62.

[0047] Then, the second trained language model 62, upon receiving the above prompt, performs a search according to the prompt, as shown in Figure 6(1), and outputs the search result in Figure 6C: "We could not find the official homepage for Western-style restaurant A, but we will provide you with information from Tabelog. Address of Western-style restaurant A: F2-18-17, Kodaira City, Tokyo Recent posts: Currently, the operating status of the store is unknown, and it is possible that it is closed or has closed down." As a result, the acquisition unit 52 acquires the outputted search result as shown.

[0048] Returning to Figure 3, the output control unit 53 updates the information contained in the response statement based on the information contained in the search results acquired by the acquisition unit 52, and responds to the user through the output unit 4.

[0049] Here, using Figure 6, we will explain the process by which the output control unit 53 updates the response statement based on the search results acquired by the acquisition unit 52 and outputs it to the user. Figure 6 is a diagram that specifically shows the process of updating and outputting the response statement (process (2) in Figure 6).

[0050] Specifically, after the search result C shown in Figure 6 (1) is obtained, the output control unit 53 first updates the address in the response statement A in Figure 6 (similar to A in Figure 4), as shown in Figure 6 (2). Next, since the search result C in Figure 6 includes "Currently, the store's operating status is unknown, and it may be closed or permanently closed," the output control unit 53 adds "Currently, the store's operating status is unknown, and it may be closed or permanently closed" to the response statement A in Figure 6. Through these processes, the output control unit 53 updates the response statement to D in Figure 6. Then, the output control unit 53 responds to the user with the response statement D in Figure 6 through the output unit 4.

[0051] Furthermore, the output control unit 53 may select an official website or platform from the information source of the search results obtained by the acquisition unit 52 from the internet, and update the response text based on the selected information.

[0052] Specifically, the acquisition unit 52 searches the internet using the facility name included in the response message and obtains the search results. Then, the output control unit 53 selects information in the search results whose URLs contain the facility name or part of it. The output control unit 53 acquires this selected information as the official homepage and updates the response message.

[0053] Furthermore, if the official website is not found, the output control unit 53 selects information in the search results that includes the name or part of the pre-specified platform "Tabelog (trademark registered)" in the URL of the information provider. The output control unit 53 retrieves this selected information as information from Tabelog and updates the response message.

[0054] (Processing flow of Information Processing Device 1) Next, an example of the processing procedure by the information processing device 1 according to Embodiment 1 will be described using Figure 7. Figure 7 is a flowchart showing an example of the information processing flow of the information processing device 1 according to Embodiment 1. Note that each step in the flowchart shown in Figure 7 can be executed in a different order, and additional or omitted processes may be included.

[0055] First, the information processing device 1 receives a user request (S1). Next, the information processing device 1 generates a response sentence using the first trained language model 61 (S2). The information processing device 1 obtains the generated response sentence (S3). The information processing device 1 creates a prompt that includes search conditions for searching the internet for facility information (name, address, etc.) contained in the response sentence, and if the prompt is input to the second trained language model 62 and search results are obtained (S4; YES), the information processing device 1 updates the facility information contained in the response sentence based on the facility information contained in the search results (S5). Then, the information processing device 1 responds to the user with the updated response sentence (S6) and terminates the process.

[0056] Furthermore, the information processing device 1 creates a prompt that includes search conditions for searching the internet for facility information (name, address, etc.) contained in the response sentence. If the prompt is input to the second trained language model 62 and no search results are obtained (S4; NO), the device responds to the user with the same response sentence without updating it (S6), and terminates the process.

[0057] (effect) The above-described embodiment 1 provides the following effects. The information processing device 1 creates a prompt that includes search conditions so that accurate information can be obtained from the internet based on the facility name and other information included in the response statement. The information processing device 1 updates the response statement based on the search results and responds to the user. As a result, the information processing device 1 can provide the user with the latest information, improving user convenience.

[0058] Furthermore, the information processing device 1 can also search for store information, including information on store closures and temporary suspensions, thus preventing users from wasting their time visiting stores due to closures or other reasons.

[0059] [Embodiment 2] Embodiment 2 describes how the information processing device 1 handles cases where the search results from the Internet are multiple different pieces of information.

[0060] Note that a detailed explanation of the configuration and processing of the information processing device 1 according to Embodiment 2 will be omitted, as it is similar to the information processing device 1 according to Embodiment 1. The following will describe the processing content that differs from that of Embodiment 1. In this Embodiment 2, the output control unit 53, which differs from that of Embodiment 1, will be described in detail.

[0061] The output control unit 53 updates the response statement based on a predetermined number of pieces of information if the search results acquired by the acquisition unit 52 contain multiple different pieces of information. For example, the output control unit 53 updates the response statement based on 10 or more pieces of information if the search results acquired by the acquisition unit 52 contain multiple different pieces of information.

[0062] Here, using Figure 8, we will explain the processing that the output control unit 53 handles when the search results contain multiple different pieces of information. Figure 8 is a diagram that specifically shows the process of updating the response statement when the search results contain multiple different pieces of information.

[0063] The output control unit 53 retrieves "(1) H2-10-16, Kodaira City, Tokyo," which is included in the search results shown in Figure 8E, "The search results for the address of Western-style restaurant A yielded the following two addresses..." and has a retrieve count of 10 or more. Then, the output control unit 53 updates the address in the response statement shown in Figure 8F, "The address of Western-style restaurant A is as follows: Address: F2-18-17, Kodaira City, Tokyo," which was obtained from the first trained language model 61, to "H2-10-16, Kodaira City, Tokyo."

[0064] Furthermore, if the search results acquired by the acquisition unit 52 contain multiple different pieces of information, the output control unit 53 may update the response statement based on the most frequent piece of information.

[0065] Specifically, the output control unit 53 retrieves the address with the highest number of occurrences, "(1) H2-10-16, Kodaira City, Tokyo," which is included in the search results shown in Figure 8E, "The search results for the address of Western-style restaurant A yielded the following two addresses..." Then, the output control unit 53 updates the address in the response statement shown in Figure 8F, "The address of Western-style restaurant A is as follows: Address: F2-18-17, Kodaira City, Tokyo," to "H2-10-16, Kodaira City, Tokyo."

[0066] Furthermore, the output control unit 53 may indicate that there are multiple pieces of information if they are included. For example, if the search results acquired by the acquisition unit 52 include multiple different pieces of information, the output control unit 53 updates the information to the most frequently included piece of information and indicates that there are multiple pieces of information, such as "Multiple pieces of information" or "Other information available."

[0067] Specifically, the output control unit 53 updates the response statement "The address of Western-style restaurant A is as follows: Address: H2-10-16, Kodaira City, Tokyo" with the most frequently acquired information, adding "Multiple pieces of information available," and updates the response statement shown in Figure 8F to the response statement shown in Figure 8G.

[0068] Furthermore, the output control unit 53 may update the response statement by adding the information from the search results if the information specified is not included in the response statement. For example, the output control unit 53 may find that there is missing information in the response statement generated by the first trained language model 61, and that missing information is included in the search results. In this case, the output control unit 53 updates the response statement by adding the missing information based on the information from the search results.

[0069] Here, using Figure 9, we will explain a specific example of how the output control unit 53 processes the case where the information specified in the response statement is not included, but the information specified in the search results is included. Figure 9 is a diagram that specifically shows the process of updating the response statement when the information specified in the response statement is not included, but the information specified in the search results is included.

[0070] For example, the output control unit 53 retrieves the search result shown in Figure 9H, "The search results for the address, business hours, and regular holidays of Western-style restaurant A are as follows: G1-15-12, Kodaira City, Tokyo; 11:30~21:30 (LO 21:00); Irregular holidays."

[0071] Then, the output control unit 53 adds the information included in the search results to the information that is currently "Not available" in the response statement shown in Figure 9I, "Information about Western-style restaurant A is as follows: Address: G1-15-12, Kodaira City, Tokyo Business hours: Not available Closed days: Not available". The output control unit 53 updates the response statement shown in Figure 9J, "Information about Western-style restaurant A is as follows: Address: G1-15-12, Kodaira City, Tokyo Business hours: 11:30~21:30 (LO 21:00) Closed days: Irregular".

[0072] According to the embodiment 2 described above, when there is multiple pieces of information, the information processing device 1 updates the response statement based on the most abundant piece of information, or on more than the planned number of pieces of information to be acquired. This makes it possible to acquire information based on more accurate information, and improves the accuracy of the information provided to the user.

[0073] Furthermore, if the information processing device 1 obtains multiple pieces of information, it outputs a message indicating that multiple pieces of information exist, allowing the user to know that multiple pieces of information are available and take appropriate action, such as contacting the store.

[0074] Furthermore, the information processing device 1 can meet user needs by supplementing missing information in the response statement with information included in the search results.

[0075] [Embodiment 3] Now, while embodiments of this disclosure have been described, this disclosure may be implemented in various other forms besides those described above.

[0076] (Numerical values, etc.) The numerical values ​​and thresholds (for example, arbitrary numerical values) used in the above embodiment are merely examples and can be changed as desired.

[0077] [Hardware configuration] Furthermore, the information processing device 1 according to Embodiments 1 to 3 described above is implemented by a computer 1000 having the configuration shown in Figure 10. The following explanation will use the information processing device 1 as an example. Figure 10 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing device 1 of Embodiments 1 to 3. The computer 1000 has a CPU 1100, RAM 1200, ROM 1300, HDD 1400, communication interface (I / F) 1500, input / output interface (I / F) 1600, and media interface (I / F) 1700.

[0078] The CPU 1100 operates based on programs stored in the ROM 1300 or HDD 1400, controlling various components. The ROM 1300 stores boot programs executed by the CPU 1100 when the computer 1000 starts up, as well as programs that depend on the computer 1000's hardware.

[0079] The HDD1400 stores programs executed by the CPU1100, as well as data used by such programs. The communication interface1500 receives data from other devices via a predetermined communication network and sends it to the CPU1100, and transmits data generated by the CPU1100 to other devices via the predetermined communication network.

[0080] The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, via the input / output interface 1600. The CPU 1100 acquires data from input devices via the input / output interface 1600. The CPU 1100 also outputs the generated data to output devices via the input / output interface 1600.

[0081] The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads the program from the recording medium 1800 onto the RAM 1200 via the media interface 1700 and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or PD (Phase Change Rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory.

[0082] For example, when computer 1000 functions as an information processing device 1 according to Embodiments 1 to 6, the CPU 1100 of computer 1000 realizes the functions included in the control unit 5 of the information processing device 1 by executing programs loaded on RAM 1200. The CPU 1100 of computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be obtained from other devices via a predetermined communication network.

[0083] 〔others〕 Furthermore, among the processes described in each of the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, specific names, and information including various data and parameters shown in the above document and drawings can be changed at will unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown.

[0084] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0085] Furthermore, the above embodiments can be combined as appropriate, provided that the processing content is not contradictory.

[0086] Although some embodiments of the present invention have been described in detail above with reference to the drawings, these are illustrative examples, and the present disclosure can be implemented in various other forms with modifications and improvements based on the knowledge of those skilled in the art, starting with the embodiments described in the disclosure section.

[0087] Furthermore, the terms "section, module, unit" mentioned above can be replaced with "means" or "circuit," etc. For example, the generation unit can be replaced with generation means or generation circuit. [Explanation of symbols]

[0088] 1. Information Processing Device 2 Communications Department 3. Input section 4 Output section 5. Control Unit 51 Generation part 52 Acquisition Department 53 Output control unit 6 Memory section 61. First pre-trained language model 62. Second pre-trained language model

Claims

1. A generation unit that inputs a first prompt to a first trained language model to generate a response to a user request and generates a response sentence, An acquisition unit inputs a second prompt, which includes search conditions for searching for the first information contained in the response sentence from information publicly available on the internet, into a second trained language model to obtain search results. An output control unit that updates the response statement based on the search results and outputs it to the user, An information processing device characterized by comprising:

2. The information processing apparatus according to claim 1, characterized in that the acquisition unit inputs the second prompt, which includes a plurality of search conditions, to the second trained language model to acquire the search results.

3. The information processing apparatus according to claim 1, characterized in that the output control unit updates the response statement by modifying the response statement with a predetermined number or more of the search information when the search results include multiple different search information.

4. The information processing apparatus according to claim 1, characterized in that the output control unit updates the response statement by adding search information that is not included in the response statement when the search result includes search information that is not included in the response statement.

5. The information processing apparatus according to claim 3, wherein the output control unit adds information to the response statement indicating that there are multiple different search pieces of information when the search results include multiple different search pieces of information.

6. The information processing apparatus according to claim 1, characterized in that the search conditions include conditions for searching the official homepage of a facility included in the response statement, or conditions for searching a platform with user-submitted posts about the facility.

7. The output control unit, The information processing device according to claim 1, characterized in that if the search results include an official website that provides the first information, the response statement is updated based on the information provided by the official website, and if the official website is not included, the response statement is updated based on the information provided by a platform among the multiple information providers that has user posts related to the first information.

8. An information processing method performed by an information processing device, A generation process involves inputting a first prompt, which causes a user request to generate a response, into a first trained language model to generate a response sentence, and An acquisition step is to input a second prompt, which includes search conditions for searching for the first information contained in the response sentence from information publicly available on the internet, into a second trained language model to obtain search results, An output control step which updates the response statement based on the search results and outputs it to the user, An information processing method characterized by including

9. A generation procedure that generates a response sentence by inputting a first prompt, which causes the user to generate a response to their request, into a first trained language model, and A procedure for obtaining search results by inputting a second prompt containing search conditions for searching for the first information contained in the response sentence from information publicly available on the internet into a second trained language model, and An output control procedure which updates the response statement based on the search results and outputs it to the user, An information processing program characterized by causing a computer to execute it.

Images