Information processing device, processing method, and program

The information processing apparatus assists in creating response summaries and texts by using AI to generate and select appropriate candidates, enhancing response efficiency and accuracy through mutually exclusive and independent options with proofreading support.

JP2026093284AActive Publication Date: 2026-06-08RAKUTEN GROUP INC

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
RAKUTEN GROUP INC
Filing Date
2024-11-27
Publication Date
2026-06-08

AI Technical Summary

Technical Problem

Responders face challenges in creating appropriate responses to inquiries due to the wide range of inquiry content, requiring significant time and effort.

Method used

An information processing apparatus that utilizes a short phrase generation AI to generate mutually exclusive summary candidates, a selection unit to choose appropriate candidates, and a text generation AI to create a response text, supported by a presentation unit to present the response, with optional proofreading by a proofreading AI.

Benefits of technology

Facilitates efficient creation of response summaries and texts by selecting from candidate lists, allowing for detailed and accurate responses with reduced effort, and includes proofreading for improved accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026093284000001_ABST
    Figure 2026093284000001_ABST
Patent Text Reader

Abstract

The present invention provides an information processing device that can assist in creating a response by selecting a summary of the response from a list of possible summaries that could serve as a valid response to an inquiry. [Solution] The inquiry receiving unit 101 of the information processing device 100 receives the inquiry text. The summary candidate acquisition unit 102 acquires multiple summary candidates by giving a summary candidate prompt to the short phrase generation AI, instructing it to generate multiple summary candidates, including mutually exclusive summary candidates, that can serve as the summary of the answer to the inquiry text. The selection unit 103 selects one or more from the multiple summary candidates, and selects not to select multiple mutually exclusive summary candidates. The answer text acquisition unit 104 acquires an answer text by giving an answer text prompt to the text generation AI, instructing it to generate an answer text based on the summary candidates. The presentation unit 105 presents the acquired answer text to the respondent.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

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

Background Art

[0002] Conventionally, with respect to sellers who sell items in the electronic market or online auctions, there is known a technique for assisting in creating responses to inquiries from potential purchasers or purchasers. For example, Patent Document 1 discloses a technique of inputting the content of a question and information on an item into an AI (artificial intelligence) agent, causing the AI agent to generate a response, and supporting the creation of a response by the seller of the item.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] On the other hand, the content of inquiries covers a wide range, such as questions about items, questions about delivery, claims, cancellations, etc. For the responder, it takes time and effort to give appropriate instructions when seeking support for creating a response.

[0005] The present invention solves the above problems, and an object thereof is to provide an information processing apparatus, a processing method, and a program capable of assisting in creating a response by selecting the gist of the response from candidates for the gist that can be established as the gist of the response to an inquiry.

Means for Solving the Problems

[0006] The information processing apparatus according to the present invention is The inquiry reception department receives inquiries from those making inquiries, A summary candidate acquisition unit obtains a plurality of summary candidate summaries, which are summaries that can serve as a summary of the response to the received inquiry, and which include mutually exclusive summary candidate summaries, by providing a summary candidate prompt to a short phrase generation AI that generates short phrases and causing it to generate such a plurality of summary candidate summaries. A selection unit that selects one or more of the obtained abstract candidates, wherein the selection unit selects such that multiple abstract candidates are not selected from mutually exclusive abstract candidates. The response text acquisition unit obtains the response text by providing a response text prompt to the text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate, A presentation unit that presents the obtained response text to the respondent who is providing the response to the inquiry text, It is characterized by being equipped with [the following features]. [Effects of the Invention]

[0007] According to the present invention, it is possible to provide an information processing device, a processing method, and a program that can support the creation of an answer by selecting an answer summary from a list of candidates for an answer summary that can serve as an answer summary to an inquiry. [Brief explanation of the drawing]

[0008] [Figure 1] This diagram shows the relationship between the information processing device, the respondent terminal, the electronic market server, and the inquirer terminal according to this embodiment. [Figure 2] This figure shows the hardware configuration of the information processing device according to the embodiment. [Figure 3] This figure shows the functional configuration of the information processing device according to the embodiment. [Figure 4] This figure shows an example of an inquiry management page according to the embodiment. [Figure 5] This figure shows an example of a response creation support dialog according to the embodiment. [Figure 6]This figure shows an example of a response creation support dialog according to the embodiment. [Figure 7] This figure shows an example of a response creation support dialog according to the embodiment. [Figure 8] This figure shows an example of an inquiry management page according to the embodiment. [Figure 9] This is a diagram illustrating the response creation support process performed by the information processing device according to the embodiment. [Modes for carrying out the invention]

[0009] (1. Overall structure) As shown in Figure 1, the information processing device 100 according to an embodiment of the present invention is connected to a respondent terminal 200 and an electronic market server 300 via a computer communication network 500 such as the Internet. The respondent terminal 200 is also connected to the electronic market server 300 via the computer communication network 500, and the electronic market server 300 is connected to an inquirer terminal 400 via the computer communication network 500.

[0010] The information processing device 100 is a device that assists respondents using the respondent terminal 200 in creating answers to inquiries they have received. For example, the information processing device 100 provides the respondent terminal 200 with an answer in response to a request from the respondent terminal 200, and accepts edits to the answer provided by the respondent terminal 200. The information processing device 100 also obtains information from the electronic market server 300 regarding communication between shops selling items in the electronic market and shoppers who have purchased or are considering purchasing items.

[0011] The respondent terminal 200 is a terminal used by a respondent who answers inquiries. For example, it is a device used by a shop that sells items in an electronic market managed by the electronic market server 300. For example, a shop that uses the respondent terminal 200 receives an inquiry made to the shop in the electronic market via the electronic market server 300, creates an answer using the information processing device 100 for the received inquiry, and answers the shopper who is the inquirer. The respondent terminal 200 is, for example, a smartphone, a tablet computer, or the like.

[0012] The electronic market server 300 is a server that manages an electronic market and performs processes related to the sale and purchase of items. In addition, the electronic market server 300 mediates and manages the communication between a shop that sells items in the electronic market and a shopper who purchases or is considering purchasing an item. For example, the electronic market server 300 receives an inquiry from the inquirer terminal 400 to the shop and notifies the received inquiry to the respondent terminal 200. Also, the electronic market server 300 receives an answer to the inquiry from the respondent terminal 200 and notifies the received answer to the inquirer terminal 400.

[0013] The inquirer terminal 400 is a terminal used by an inquirer who makes an inquiry to a respondent. For example, it is a device used by a shopper who purchases or is considering purchasing an item in an electronic market managed by the electronic market server 300. For example, a shopper who uses the inquirer terminal 400 makes an inquiry to a shop via the electronic market server 300 and receives an answer from the shop to the inquiry. The inquirer terminal 400 is, for example, a smartphone, a tablet computer, or the like.

[0014] (2. Hardware Configuration of Information Processing Device) FIG. 2 is a block diagram showing the hardware configuration of the information processing device 100.

[0015] As shown in FIG. 2, the information processing apparatus 100 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM 13, a recording medium 14, an output device 15, a communication device 16, and an input device 17. Each component is connected by a bus 18.

[0016] The CPU 11 controls the operation of the entire information processing apparatus 100, is connected to each component, and exchanges control signals and data.

[0017] In the ROM 12, operating programs and various data necessary for controlling the operation of the entire information processing apparatus 100 are recorded.

[0018] The RAM 13 is for temporarily recording data and programs, and holds programs and data read from the recording medium 14, as well as other data necessary for communication.

[0019] The recording medium 14 is composed of a hard disk, a flash memory, etc., and records data processed by the information processing apparatus 100.

[0020] The output device 15 includes a display device such as an LCD (Liquid Crystal Display) and an audio output device such as a speaker. The output device 15 outputs, for example, data output from the CPU 11 under the control of the CPU 11.

[0021] The communication device 16 includes a communication interface for connecting the information processing apparatus 100 to a computer communication network such as the Internet, and exchanges data with other information processing apparatuses via the communication device 16.

[0022] The input device 17 includes input devices such as buttons, keyboards, touch panels, and microphones. The input device 17 receives operation inputs from the user of the information processing apparatus 100 and outputs signals corresponding to the received operation inputs to the CPU 11.

[0023] (3. Functional configuration of the information processing device of the embodiment) The functional configuration of the information processing device 100 will be explained using Figure 3.

[0024] Functionally, the information processing device 100 comprises an inquiry reception unit 101, a summary candidate acquisition unit 102, a selection unit 103, a response text acquisition unit 104, a presentation unit 105, and a calibration unit 106. In this embodiment, the CPU 11 and the communication device 16 function as the inquiry reception unit 101, the summary candidate acquisition unit 102, the selection unit 103, the response text acquisition unit 104, the presentation unit 105, and the calibration unit 106.

[0025] In the following, the information processing device 100 will provide the respondent terminal 200 with an inquiry management page for managing inquiries made to shops in the electronic marketplace.

[0026] Figure 4 shows an example of an inquiry management page. The inquiry management page 600 in Figure 4 is used by "Shop S1" to manage inquiries received from shoppers using the electronic marketplace. The inquiry management page 600 includes an area 601 that displays a list 601-1 of inquiries made to the shop, an area 602 that displays the content of the selected inquiry from list 601-1 and the response text to the inquiry, a button 603 for requesting assistance in creating the response text, and a button 604 for sending the created response text to the shopper who made the inquiry. In the inquiry management page 600 in Figure 4, inquiry 601-2 has been selected from list 601-1, and the content of the selected inquiry 601-2 is displayed in area 602. Area 602 includes a field 602-1 for entering the response text.

[0027] Inquiry reception unit 101 receives inquiry messages from callers.

[0028] An inquiry is a sentence that describes the content of the inquiry made by the inquirer to the respondent. For example, the inquiry is the sentence that describes the content of the inquiry included in List 601-1 in Figure 4.

[0029] For example, if "Shop S1" selects button 603 on the inquiry management page 600 in Figure 4, the inquiry reception unit 101 receives the currently selected inquiry 601-2, which reads "When will this item be in stock?".

[0030] Furthermore, when button 603 is selected on the inquiry management page 600 in Figure 4, the information processing device 100 displays a response creation support dialog on the screen of the respondent terminal 200 to assist in creating a response to the inquiry.

[0031] Figure 5 shows an example of a response creation support dialog. In addition to the function to create a response, the response creation support dialog also provides a function to proofread the response. The response creation support dialog 700-1 in Figure 5 includes an item 701 for using the function to create a response, an item 702 for using the function to proofread the response, a field 703 that displays mutually exclusive summary candidates, a field 704 that displays independent summary candidates, a button 705 for generating a response based on the information selected in fields 703 and 704, a field 706 that displays the generated response, and a button 707 for displaying the response displayed in field 706 in field 602-1 of the response management page in Figure 4. The response creation support dialog 700-1 in Figure 5 shows the state where item 701 for using the function to create a response is selected. The summary candidates included in fields 703 and 704 are summary candidates obtained by the summary candidate acquisition unit 102.

[0032] The summary candidate acquisition unit 102 acquires multiple summary candidates by providing a summary candidate prompt to the short phrase generation AI that generates short phrases, instructing it to generate multiple summary candidates, including mutually exclusive summary candidates, that can serve as a summary of the answer to the received inquiry.

[0033] A candidate summary is a short phrase that represents a possible summary of the answer to a received inquiry. A candidate summary that represents a possible answer is one that is appropriate as a written response to the inquiry and that allows for a dialogue with the inquiry. The short phrase representing the candidate summary is a phrase that expresses the answer's essence concisely with as few characters as possible, and is not suitable for use as the final answer. Furthermore, mutually exclusive candidate summaries are those that, when one candidate is selected, prevent the selection of other candidates.

[0034] A summary candidate prompt is a prompt given to the short phrase generation AI, and it is a prompt that indicates the content of the instructions for the short phrases that the short phrase generation AI will generate. A template for the summary candidate prompt is prepared in advance, and the summary candidate acquisition unit 102 creates the summary candidate prompt by inputting the received inquiry text into the template. The template for the summary candidate prompt includes, for example, instructions to generate multiple summary candidates that could serve as the summary of the answer to the received inquiry text, instructions to represent the summary candidates to be generated with phrases that express the summary of the answer concisely with as few characters as possible, but are not suitable to be used as the answer text as is, and instructions to include candidates among the multiple summary candidates to be generated that, when one candidate is selected, the other candidates are not selected, and so on.

[0035] A short phrase generation AI is a generative AI that generates short phrases in response to summary candidate prompts, and is, for example, fine-tuned to be suitable for generating short phrases. Here, a generative AI is an AI technology that automatically generates various types of content, and for example, generates text data in response to prompts by providing a large-scale language model with prompts indicating the content of the instructions. The short phrase generation AI may be provided by the information processing device 100, or by another information processing device that can communicate with the information processing device 100, and the information processing device 100 may utilize the short phrase generation AI provided by the other information processing device.

[0036] Here, the summary candidate acquisition unit 102 acquires summary candidates that include supplementary fields by providing the short phrase generation AI with a summary candidate prompt that further includes a statement that the AI ​​should generate summary candidates that include supplementary fields for the respondent to input information.

[0037] For example, the template for the summary candidate prompt further includes instructions to provide a supplementary field for entering information for one of the multiple summary candidates to be generated. When the summary candidate acquisition unit 102 inputs the created summary candidate prompt to the short phrase generation AI, the short phrase generation AI will provide the following summary candidates as mutually exclusive summary candidates for the answer to the inquiry sentence "When will this item be in stock?", such as "Scheduled to be in stock [ ]month[ Retrieve "day", "arrival date undecided, arrival planned", and "no expected arrival". Summary candidate "arrival planned [ ]month[ The "[Day]" section includes supplementary fields 709 and 710 for entering the date. The abstract candidate acquisition unit 102 then displays the acquired abstract candidates in field 703, as shown in Figure 5, using radio buttons 708 that, when one is selected, automatically deselect the other abstract candidates.

[0038] Furthermore, the abstract candidate acquisition unit 102 acquires multiple abstract candidates by providing the short phrase generation AI with an abstract candidate prompt that further includes a statement that multiple abstract candidates, including mutually non-exclusive and independent abstract candidates, should be generated.

[0039] An independent abstract candidate is an abstract candidate for which multiple selections can be made simultaneously.

[0040] For example, the template for the summary candidate prompt includes instructions to include multiple summary candidates that can be selected simultaneously. When the summary candidate acquisition unit 102 inputs the created summary candidate prompt to the short phrase generation AI, the short phrase generation AI provides summary candidates for the answer to the inquiry sentence "When will this item be in stock?", with independent summary candidates such as "Person in charge unavailable", "We will contact you after checking with the manufacturer", and "Alternative products are [ ]” and “(Free entry)[ Get the summary candidate “Alternative products are [ ]” contains supplementary field 712 for entering information on alternative products, summary candidate “(free entry) [ The section 704 includes a supplementary field 713 for respondents to freely enter information. The summary candidate acquisition unit 102 then displays the acquired summary candidates in the field 704, as shown in Figure 5, using checkboxes 711 that allow multiple selections to be made simultaneously.

[0041] The selection unit 103 selects one or more from the multiple abstract candidates obtained, and does not select multiple abstract candidates that are mutually exclusive. Specifically, the selection unit 103 makes the selection by having the respondent select mutually exclusive abstract candidates.

[0042] For example, the selection unit 103 selects one or more abstract candidates from the multiple abstract candidates displayed in columns 703 and 704 in Figure 5, and selects only one from the mutually exclusive abstract candidates in column 703. The selection unit 103 may select multiple abstract candidates from the independent abstract candidates in column 704.

[0043] Here, when information is entered into the supplementary field, the selection unit 103 selects a candidate summary that includes the supplementary field into which the information was entered.

[0044] For example, if "Shop S1" is in column 703, and information is entered in supplementary columns 709 and 710 as shown in Figure 6, the selection unit 103 will select summary candidate "Scheduled arrival [ 12 ]month[ 15Select "Day". Also, when "Shop S1" is entered in the supplement field 713 in field 704, as shown in Figure 6, a check mark will appear in the checkbox, and the selection unit 103 will select the summary candidate "Send thanks for the inquiry."

[0045] The response text acquisition unit 104 acquires the response text by providing a response text prompt to the text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate.

[0046] A response prompt is a prompt given to the text generation AI, and it indicates the content of the instructions for the text that the text generation AI should generate. A template for the response prompt is prepared in advance, and the response acquisition unit 104 creates the response prompt by inputting the content of the selected summary candidate into the template. The template for the response prompt includes, for example, an instruction to generate a response to the received inquiry, using the input summary candidate as the summary of the response.

[0047] A text generation AI is a generation AI that generates text in response to a response prompt, and is, for example, finely tuned to be suitable for text generation. The text generation AI may be provided by the information processing device 100, or by another information processing device that can communicate with the information processing device 100, and the information processing device 100 may utilize the text generation AI provided by the other information processing device.

[0048] For example, if "Shop S1" selects button 705 in the response creation support dialog 700-1 in Figure 6, the response acquisition unit 104 inputs the created response prompt to the text generation AI and obtains a response to the inquiry "When will this item be in stock?".

[0049] The presentation unit 105 presents the acquired response text to the respondent who is providing the response to the inquiry. Here, presentation to the respondent by the presentation unit 105 means transmitting the information to the respondent terminal 200.

[0050] For example, if the response acquisition unit 104 acquires a response from the text generation AI corresponding to the input response prompt, the presentation unit 105 transmits the information of the acquired response to the respondent terminal 200 of "Shop S1". On the screen of the respondent terminal 200 of "Shop S1", the acquired response is displayed in column 706, as shown in Figure 6.

[0051] Furthermore, when "Shop S1" selects button 707 in the response creation support dialog 700-1 in Figure 6, the response text from field 706 is displayed in field 602-1 on the inquiry management page 600.

[0052] Furthermore, when "Shop S1" selects item 702 in the response creation support dialog 700-1 shown in Figure 6 to use the function for proofreading the response, the information processing device 100 displays the modified response creation support dialog 700-2 on the screen of the respondent terminal 200, as shown in Figure 7.

[0053] The response creation support dialog 700-2 in Figure 7 includes a field 714 where the generated response is displayed, a button 715 for proofreading the response displayed in field 714, a field 716 where the proofread response is displayed, and a button 717 for displaying the response displayed in field 716 in field 602-1 of the response management page in Figure 4. The response displayed in field 714 is the response displayed in field 706 of the response creation support dialog 700-1 in Figure 6, and the response displayed in field 714 is editable. For example, suppose the area indicated by the dotted line 718 has been edited by "Shop S1".

[0054] When the response submitted by the respondent is edited, the proofreading unit 106 has the edited response proofread by a proofreading AI that proofreads text, and obtains the proofread response.

[0055] A proofreading AI is a generative AI that proofreads and outputs input text, and is, for example, finely tuned to be suitable for proofreading text. The proofreading AI may be provided by the information processing device 100, or by another information processing device that can communicate with the information processing device 100, and the information processing device 100 may utilize the proofreading AI provided by the other information processing device.

[0056] For example, if "Shop S1" edits the answer displayed in column 714 and selects button 715, the proofreading unit 106 inputs a prompt to the proofreading AI instructing it to proofread the edited answer in column 714, and obtains the proofreading result.

[0057] The presentation unit 105 presents the respondent with the difference between the edited response and the proofread response.

[0058] For example, if the proofreading unit 106 obtains a corrected response from the proofreading AI, the presentation unit 105 transmits the information of the corrected response to the respondent terminal 200 of "Shop S1," and the corrected response is displayed in column 716 on the screen of the respondent terminal 200 of "Shop S1," as shown in Figure 7. Here, the presentation unit 105 also transmits information indicating the difference between the edited response in column 714 and the corrected response in column 716 using strikethrough lines 719 and underlines 720.

[0059] Furthermore, when "Shop S1" selects button 717 in the response creation support dialog 700-2 in Figure 7, the corrected response is displayed in column 602-1 of the inquiry management page 600, as shown in Figure 8. In this case, the difference between the edited response and the corrected response is not displayed in column 602-1; only the response with the corrections reflected is displayed. "Shop S1" can edit the response displayed in column 602-1. Also, when "Shop S1" selects button 604, the response displayed in column 602-1 is notified to "Shopper U1," who made inquiry 601-2.

[0060] (4. Support process for creating answers using the information processing device of the embodiment) The response creation support process performed by the information processing device 100 in this embodiment will be explained with reference to Figure 9. When the information processing device 100 receives a request from the respondent terminal 200 for assistance in creating a response to an inquiry, it starts the response creation support process shown in Figure 9. For example, when button 603 is selected and item 701 is selected on the inquiry management page 600 in Figure 4, the information processing device 100 starts the response creation support process shown in Figure 9.

[0061] The inquiry reception unit 101 receives the inquiry from the inquirer (step S101).

[0062] For example, the inquiry reception unit 101 receives the inquiry text "When will this item be in stock?" for inquiry 601-2, which is selected on the inquiry management page 600 in Figure 4.

[0063] The summary candidate acquisition unit 102 acquires multiple summary candidates by providing the short phrase generation AI with a summary candidate prompt that indicates the generation of multiple summary candidates, including mutually exclusive summary candidates, which can serve as a summary of the answer to the received inquiry (step S102).

[0064] For example, the summary candidate acquisition unit 102 inputs a summary candidate prompt to the short phrase generation AI, which includes instructions to generate multiple summary candidates that could serve as a summary of the answer to the inquiry sentence "When will this item be in stock?", instructions to represent the generated summary candidates with phrases that concisely express the gist of the answer in as few characters as possible but are not suitable for use as the answer sentence as is, instructions to include among the multiple summary candidates that, when one candidate is selected, prevents the other candidates from being selected, instructions to include among the multiple summary candidates that allow for multiple selections simultaneously, and instructions to provide a supplementary field for inputting information for one of the multiple summary candidates, thereby acquiring multiple summary candidates.

[0065] The selection unit 103 selects one or more of the acquired abstract candidates, and does not select multiple abstract candidates that are mutually exclusive (step S103).

[0066] For example, as shown in Figure 6, if "Shop S1" enters information in supplementary fields 709 and 710 in field 703 and enters information in supplementary field 713 in field 704, the selection unit 103 will select summary candidate "Scheduled arrival [ 12 ]month[ 15 Select "Date" and the summary option "Send a thank you for the inquiry."

[0067] The response text acquisition unit 104 determines whether or not there has been an instruction to acquire a response text (step S104). If the response text acquisition unit 104 determines that there has been an instruction to acquire a response text (step S104; YES), it acquires a response text by giving the text generation AI a response text prompt to generate a response text based on the selected summary candidate (step S105). On the other hand, if the response text acquisition unit 104 determines that there has been no instruction to acquire a response text (step S104; NO), it simply waits.

[0068] For example, when "Shop S1" selects button 705 in the response text creation support dialog 700-1 in Figure 6, the response text acquisition unit 104 generates a summary candidate for the response text to the inquiry "When will this product be in stock?", which is "Expected arrival date [ 12 ]month[ 15 Create a prompt that includes instructions to generate the summary of the response, including the phrase "[Day]" and the suggested summary "Thank you for your inquiry.", input it into the text generation AI, and retrieve the response output from the text generation AI. Meanwhile, if "Shop S1" does not select button 705, wait.

[0069] The presentation unit 105 presents the acquired response text to the respondent who is providing the response to the inquiry text (step S106).

[0070] For example, the display unit 105 transmits the information of the acquired response text to the respondent terminal 200 of "Shop S1". On the screen of the respondent terminal 200 of "Shop S1", the acquired response text is displayed in column 706, as shown in Figure 6.

[0071] According to this embodiment, multiple summary candidates are generated, including mutually exclusive summary candidates, that can serve as the gist of an answer to an inquiry. An answer is then generated based on the summary candidate selected from the multiple summary candidates, and the answer to the inquiry is presented to the respondent. This assists in creating an answer to an inquiry, allowing the respondent to easily create an answer.

[0072] Furthermore, according to this embodiment, the respondent is given mutually exclusive summary candidates to select from. This allows for the generation and presentation of a response text that contains the summary intended by the respondent.

[0073] Furthermore, according to this embodiment, the response text edited by the respondent is proofread, and the difference between the edited response text and the proofread response text is presented to the respondent. This allows the respondent to confirm a more accurate response text and to easily compare the edited response text with the proofread response text.

[0074] Furthermore, according to this embodiment, the summary candidates include supplementary fields for the respondent to input information, and when information is entered in the supplementary fields, the summary candidate including the entered supplementary fields is selected. This makes it possible to include more specific content in the generated response text and reduces the effort required for the respondent to make a selection.

[0075] Furthermore, according to this embodiment, in addition to mutually exclusive summary candidates, independent summary candidates can be generated, and response sentences based on both the mutually exclusive and independent summary candidates can be presented to the respondent. This makes it possible to generate response sentences that specify the summary in more detail.

[0076] (5. Variant) Although embodiments of the present invention have been described above, various modifications and applications are possible when implementing the present invention.

[0077] In the above embodiment, the information processing device 100 is said to acquire information regarding communication between the responding shop and the inquiring shopper from the electronic market server 300, but it is not limited to this. The information processing device 100 may be connected to the inquiring terminal in a communicative manner, and the inquiry receiving unit 101 may directly receive inquiries from the inquiring terminal 400.

[0078] Furthermore, in the above embodiment, the respondent is defined as a shop that sells items in the electronic market managed by the electronic market server 300, and the inquirer is defined as a shopper who purchases or considers purchasing items in the electronic market managed by the electronic market server 300. However, the respondent and inquirer are not limited to these definitions. The respondent is simply a person who answers an inquiry, and the inquirer is simply a person who makes an inquiry to the respondent.

[0079] Furthermore, in the above embodiment, the selection unit 103 selects by having the respondent select mutually exclusive summary candidates, but it is not limited to this. The selection unit 103 may also select mutually exclusive summary candidates based on the content of the inquiry.

[0080] For example, the selection unit 103 inputs a prompt to the generation AI that includes an instruction to select from mutually exclusive summary candidates acquired by the summary candidate acquisition unit 102 the one that shows the summary of the most frequently occurring answer to an inquiry in the electronic market, and selects the output of the generation AI as a summary candidate.

[0081] This configuration allows for the automatic selection of potential abstracts.

[0082] Furthermore, in the above embodiment, the short phrase generation AI and the text generation AI may be implemented by a single generation AI. Moreover, the short phrase generation AI, the text generation AI, and the proofreading AI may be implemented by a single generation AI.

[0083] For example, the summary candidate acquisition unit 102 gives the generation AI a summary candidate prompt to acquire multiple summary candidates, and the answer text acquisition unit 104 gives the same generation AI an answer text prompt to acquire an answer text. Furthermore, the proofreading unit 106 gives the same generation AI a prompt that includes an instruction to proofread the edited answer text to acquire the proofreading result.

[0084] This configuration eliminates the need to prepare separate AI systems for short phrase generation, text generation, and proofreading. The system can obtain the necessary information through instructions included in the prompt and easily present a response.

[0085] Furthermore, in the above embodiment, the selection unit 103 selects only one from the mutually exclusive abstract candidates in column 703, but this is not limited to that. The selection unit 103 may select one or more from the acquired multiple abstract candidates, but may not select multiple from the mutually exclusive abstract candidates; in other words, it may not select any of the mutually exclusive abstract candidates.

[0086] For example, mutually exclusive summary candidates are displayed as selectable checkboxes rather than radio buttons, as shown in column 703 of Figure 5. When a checkbox for one of the mutually exclusive summary candidates is checked, the checkboxes for the other summary candidates become unselectable. Furthermore, button 705 can be selected even if none of the checkboxes for the mutually exclusive summary candidates are selected. If the selection unit 103 does not select any of the mutually exclusive summary candidates, the response text acquisition unit 104 creates a response text prompt that includes instructions to generate a response text based on the one or more selected independent summary candidates and the fact that no mutually exclusive summary candidates were selected, and provides this prompt to the text generation AI to acquire the response text.

[0087] Furthermore, in the above embodiment, the calibration unit 106 obtains the calibration result when "Shop S1" selects button 715, but this is not limited to this. For example, the calibration unit 106 may be configured to automatically obtain the calibration result when the answer text in column 714 is edited.

[0088] Furthermore, by applying an operating program that defines the operation of the information processing device 100 according to the above embodiment to an existing personal computer or information terminal device, it is also possible to make the personal computer or information terminal device function as the information processing device 100 according to the embodiment.

[0089] (Note) [1] The inquiry reception department receives inquiries from those making inquiries, A summary candidate acquisition unit obtains a plurality of summary candidate summaries, which are summaries that can serve as a summary of the response to the received inquiry, and which include mutually exclusive summary candidate summaries, by providing a summary candidate prompt to a short phrase generation AI that generates short phrases and causing it to generate such a plurality of summary candidate summaries. A selection unit that selects one or more of the obtained abstract candidates, wherein the selection unit selects such that multiple abstract candidates are not selected from mutually exclusive abstract candidates. The response text acquisition unit obtains the response text by providing a response text prompt to the text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate, A presentation unit that presents the obtained response text to the respondent who is providing the response to the inquiry text, An information processing device characterized by comprising:

[0090] [2] The selection unit selects the mutually exclusive summary candidates by having the respondent select one. The information processing apparatus according to [1], characterized in that

[0091] [3] The selection unit selects the mutually exclusive summary candidates based on the content of the inquiry statement. The information processing apparatus according to [1], characterized in that

[0092] [4] The system further includes a proofreading unit that, once the respondent has edited the response text provided, has a proofreading AI proofread the edited response text and obtains the proofread response text. The display unit presents the difference between the edited response and the proofread response to the respondent. An information processing apparatus according to any one of [1] to [3], characterized by the above.

[0093] [5] The aforementioned short phrase generation AI and the aforementioned sentence generation AI are realized by a single generation AI. An information processing apparatus according to any one of [1] to [4], characterized by the following:

[0094] [6] The aforementioned short phrase generation AI, the aforementioned text generation AI, and the aforementioned proofreading AI are all realized by a single generation AI. The information processing apparatus according to [4], characterized in that

[0095] [7] The summary candidate acquisition unit acquires a summary candidate including the supplementary field by providing the short phrase generation AI with a summary candidate prompt that further includes a statement that the AI ​​should generate a summary candidate including a supplementary field for the respondent to input information, thereby acquiring a summary candidate including the supplementary field. When information is entered into the supplementary field, the selection unit selects a candidate summary that includes the supplementary field into which the information was entered. The information processing apparatus according to [2], characterized in that

[0096] [8] The summary candidate acquisition unit acquires the multiple summary candidates by providing the short phrase generation AI with a summary candidate prompt that further includes a statement that the AI ​​should generate multiple summary candidates, including mutually non-exclusive and independent summary candidates. An information processing device according to any one of [1] to [7], characterized by the following:

[0097] [9] A processing method executed by an information processing device, The inquiry reception step involves receiving an inquiry from the person making the inquiry, A summary candidate acquisition step involves obtaining a plurality of summary candidates by providing a summary candidate prompt to a short phrase generation AI that generates short phrases, which prompts the AI ​​to generate a plurality of summary candidates, including mutually exclusive summary candidates, that can serve as a summary of the answer to the received inquiry, and causing the AI ​​to generate such a plurality of summary candidates. A selection unit that selects one or more from the multiple abstract candidates obtained, the selection step being such that multiple abstract candidates are not selected from mutually exclusive abstract candidates, The response text acquisition step involves obtaining the response text by providing a response text prompt to a text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate, and causing it to generate the response text. A presentation step in which the obtained response text is presented to the respondent who is to respond to the inquiry text, A processing method characterized by comprising:

[0098]

[10] Computers, The inquiry reception department receives inquiries from those making inquiries. A summary candidate acquisition unit obtains multiple summary candidates by providing a summary candidate prompt to a short phrase generation AI that generates short phrases, which prompts the AI ​​to generate multiple summary candidates, including mutually exclusive summary candidates, that can serve as a summary of the answer to the received inquiry. A selection unit that selects one or more of the obtained abstract candidates, wherein the selection unit does not select more than one abstract candidate that is mutually exclusive. The response text acquisition unit obtains the response text by providing a response text prompt to the text generation AI, which generates the text, instructing it to generate a response text based on the selected summary candidate. A display unit that presents the obtained response text to the respondent who is providing the response to the inquiry text. A program characterized by being designed to function as such.

[0099] The above program may be recorded on a non-transitory recording medium. Non-transitory recording media can be distributed and sold independently of computers. Here, a non-transitory recording medium refers to a tangible recording medium. Examples of non-transitory recording media include compact disks, flexible disks, hard disks, magneto-optical disks, digital video disks, magnetic tapes, and semiconductor memory. A transient recording medium refers to the transmission medium (propagating signal) itself. Examples of transient recording media include electrical signals, optical signals, and electromagnetic waves. A temporary storage area is an area for temporarily storing data or programs, such as volatile memory like RAM (Random Access Memory).

[0100] This invention allows for various embodiments and modifications without departing from the broad spirit and scope of the invention. Furthermore, the embodiments described above are for illustrative purposes only and do not limit the scope of the invention. In other words, the scope of the invention is indicated not by the embodiments, but by the claims. Various modifications made within the scope of the claims and the equivalent scope of the meaning of the invention are considered to be within the scope of this invention. [Industrial applicability]

[0101] According to the present invention, it is possible to provide an information processing device, a processing method, and a program that can support the creation of an answer by selecting an answer summary from a list of candidates for an answer summary that can serve as an answer summary to an inquiry. [Explanation of Symbols]

[0102] 11 CPU 12 ROM 13 RAM 14 Recording media 15 Output Devices 16 Communication devices 17 Input Devices 18 bus 100 Information Processing Devices 101 Inquiry Reception Department 102 Abstract Candidate Acquisition Department 103 Selection Section 104 Answer sentence acquisition part 105 Presentation section 106 Proofreading Department 200 respondent terminals 300 Electronic Market Servers 400 Inquirer terminal 500 Computer Communication Networks 600 Inquiry Management Page 601,602 areas 601-1 List 601-2 Inquiry Buttons 603, 604, 705, 707, 715, 717 Columns 602-1, 703, 704, 706, 714, 716 700-1, 700-2 Answer text creation support dialog 701,702 items 708 Radio buttons 709,710,712,713 Replenishment column 711 checkboxes 718 dotted line 719 Strikethrough 720 Underline

Claims

1. The inquiry reception department receives inquiries from those making inquiries, A summary candidate acquisition unit obtains a plurality of summary candidate summaries, which are summaries that can serve as a summary of the response to the received inquiry, and which include mutually exclusive summary candidate summaries, by providing a summary candidate prompt to a short phrase generation AI that generates short phrases and causing it to generate the plurality of summary candidate summaries. A selection unit that selects one or more of the acquired abstract candidates, wherein the selection unit ensures that no more than one abstract candidate is selected from mutually exclusive abstract candidates. The response text acquisition unit obtains the response text by providing a response text prompt to the text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate, A presentation unit that presents the obtained response text to the respondent who is providing the response to the inquiry text, An information processing device characterized by comprising:

2. The selection unit selects the mutually exclusive summary candidates by having the respondent select one. The information processing apparatus according to feature 1.

3. The selection unit selects the mutually exclusive summary candidates based on the content of the inquiry statement. The information processing apparatus according to feature 1.

4. The system further includes a proofreading unit that, once the respondent has edited the response text provided, has a proofreading AI proofread the edited response text and obtains the proofread response text. The display unit presents the difference between the edited response and the proofread response to the respondent. The information processing apparatus according to any one of claims 1 to 3.

5. The aforementioned short phrase generation AI and the aforementioned sentence generation AI are realized by a single generation AI. The information processing apparatus according to any one of claims 1 to 3.

6. The aforementioned short phrase generation AI, the aforementioned text generation AI, and the aforementioned proofreading AI are all realized by a single generation AI. The information processing apparatus according to feature 4.

7. The summary candidate acquisition unit acquires a summary candidate including the supplementary field by providing the short phrase generation AI with a summary candidate prompt that further includes a statement that the AI ​​should generate a summary candidate including a supplementary field for the respondent to input information, thereby acquiring a summary candidate including the supplementary field. When information is entered into the supplementary field, the selection unit selects a candidate summary that includes the supplementary field into which the information was entered. The information processing apparatus according to feature 2.

8. The summary candidate acquisition unit acquires the multiple summary candidates by providing the short phrase generation AI with a summary candidate prompt that further includes a statement that the AI ​​should generate multiple summary candidates, including mutually non-exclusive and independent summary candidates. The information processing apparatus according to any one of claims 1 to 3.

9. A processing method executed by an information processing device, The inquiry reception step involves receiving an inquiry from the person making the inquiry, A summary candidate acquisition step involves providing a short phrase generation AI with a summary candidate prompt indicating that it should generate multiple summary candidates, including mutually exclusive summary candidates, which are all possible summaries of the response to the received inquiry, thereby obtaining the multiple summary candidates. A selection unit that selects one or more of the obtained abstract candidates, the selection step being such that multiple abstract candidates are not selected from mutually exclusive abstract candidates, The response text acquisition step involves obtaining the response text by providing a response text prompt to a text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate, and causing it to generate the response text. A presentation step in which the obtained response text is presented to the respondent who is to respond to the inquiry text, A processing method characterized by comprising:

10. Computers, The inquiry reception department receives inquiries from those making inquiries. A summary candidate acquisition unit obtains the multiple summary candidates by providing a summary candidate prompt to a short phrase generation AI that generates short phrases, which prompts the AI ​​to generate multiple summary candidates, including mutually exclusive summary candidates, that can serve as a summary of the answer to the received inquiry. A selection unit that selects one or more of the obtained abstract candidates, wherein the selection unit ensures that no more than one abstract candidate is selected from mutually exclusive abstract candidates. The response text acquisition unit obtains the response text by providing a response text prompt to the text generation AI that generates text, instructing it to generate a response text based on the selected summary candidate. A display unit that presents the obtained response text to the respondent who is providing the response to the inquiry text. A program characterized by being designed to function as such.