Information processing system, information processing method, and program

Abstract

This information processing system includes: an acquisition unit that acquires an input text entered by a user via a first GUI; a generation unit that generates, using a language model, a second GUI for entering an answer to a question for clarifying a requirement intended by the user based on the input text; and a display control unit that displays the second GUI.

Description

Information processing system, information processing method, and program

[0001] The present invention relates to a conversational AI (CAI) system.

[0002] Conversational AI (CAI) systems generally use text-based interfaces. However, to perform complex tasks such as travel planning and programming support, users need to read and write large amounts of text to instruct the system on requirements, understand the system's responses, and add additional instructions to the system as needed. This reduces interaction efficiency. To improve interaction efficiency, graphical user interfaces (GUIs) are used in conventional CAI systems.

[0003] As a conventional technology for CAI using GUIs, there exists a conventional technology that allows users to directly edit documents and code on a canvas, eliminating the need to re-enter them into prompts. Further, as a conventional technology for CAI using GUIs, Non-Patent Document 1 discloses a technology that converts long and complex LLM responses into graphical diagrams to support information search and question answering.

[0004] In the conventional CAI systems described above, a GUI is first implemented during system design in accordance with the tasks the system performs and the expected interactions with users. When the system is in use, users input information via the GUI, and the system displays outputs on the GUI.

[0005] Peiling Jiang, Jude Rayan, Steven P Dow, and Haijun Xia, "Graphologue: Exploring large language model responses with interactive diagrams," in Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, 2023, pp. 1-20.

[0006] In CAI systems, users may express requirements incompletely or inaccurately through instructions. This can lead to situations where the system's response fails to meet the user's requirements. To prevent this, an approach is being considered that generates clarifying questions (CQs) to clarify the user's intended requirements. However, when generating CQs using conventional CAI systems, some CQs require the user to expend effort to answer pre-configured GUIs, posing a challenge.

[0007] The present invention has been made in view of the above points, and aims to provide a technology that enables a conversational AI system to easily answer questions generated in order to clarify the user's intended requirements in response to an instruction.

[0008] An information processing system according to one aspect of the disclosed technology includes: an acquisition unit that acquires input text entered by a user via a first GUI; a generation unit that generates a second GUI for inputting answers to questions for clarifying the user's intended requirements based on the input text, using a language model; and a display control unit that displays the second GUI.

[0009] According to the disclosed technology, a technology is provided that enables conversational AI systems to easily answer questions generated to clarify the user's intended requirements in response to instructions.

[0010] This figure shows an example of the configuration of a CAI system in the prior art. This figure shows an example of a GUI. This is a flowchart for explaining the processing procedure in the prior art. This figure shows an example of the configuration of a CAI system in an embodiment of the present invention. This figure shows an example of the functional configuration of the information processing system 100. This is a flowchart for explaining the processing procedure in an embodiment of the present invention. This figure shows an example of a GUI. This figure shows an example of the hardware configuration of the information processing system 100.

[0011] Hereinafter, embodiments of the present invention (this embodiment) will be described with reference to the drawings. The embodiments described below are merely examples, and the embodiments to which the present invention is applied are not limited to these embodiments. Multiple features are described in the embodiments, but not all of these features are essential to the invention, and the multiple features may be combined arbitrarily.

[0012] First, the prior art and its problems will be described in more detail below, and then the technology related to this embodiment will be described in detail.

[0013] (Regarding the prior art) Figure 1 shows an example of the configuration of a CAI system in the prior art. As shown in Figure 1, this CAI system consists of an information processing system 10, a model 30, and GUI code 40. The information processing system 10 is one or more computers. In this example, the model 30 is an LLM (Large Language Model).

[0014] GUI code 40 may be implemented in advance. GUI code is code written in, for example, HTML, CSS, or JavaScript. User 20 refers to a person who accesses the information processing system 10 and interacts with model 30 via the GUI. Note that "User 20" can also be thought of as a user terminal operated by a user (person).

[0015] In the example shown in Figure 1, Model 30 and GUI code 40 are shown as being provided on an external device (e.g., a server on the Internet) of the information processing system 10, but both Model 30 and GUI code 40 may also be stored in an internal storage unit (not shown) of the information processing system 10.

[0016] In this CAI system, first, the display control unit, which constitutes the information processing system 10, renders the GUI using the GUI code 40 and displays it to the user 20 via the display unit.

[0017] Figure 2 shows an example of a GUI displayed on a display unit in the prior art. As shown in Figure 2, this GUI has an area for displaying the conversation history, a text box for the user to input instructions (the "Type a message" section), and a submit button for inputting the input text to the information processing system 10 (the "Submit" section).

[0018] The processing steps performed by a conventional CAI system will be explained according to the procedure in the flowchart of Figure 3. The user 20 and the information processing system 10 perform the processes S1 (step 1) to S4 shown in Figure 3 during each turn of the conversation, and repeat these processes until the conversation ends. Note that Figure 3 shows one turn as the process from start to end.

[0019] <S1: User Input> User 20 operates the GUI displayed on the display unit and inputs an instruction. For example, User 20 uses the keyboard to input an instruction (e.g., "Tell me a recommended smartphone") into the text box and clicks the send button with the mouse. As a result, the instruction is input into the information processing system 10.

[0020] <S2: Display of user input> The information processing system 10 displays the input instruction text in the conversation history area.

[0021] <S3: Response Generation> The information processing system 10 inputs the instruction text as input text to the model 30. The model 30 estimates an appropriate response text (output text) for the input text and returns the output text to the information processing system 10.

[0022] Furthermore, when inputting to Model 30, the information processing system 10 may provide Model 30 with pre-configured system prompts, etc., in addition to the instruction statement.

[0023] <S4: Display of Response> The information processing system 10 updates the GUI using the output text and the pre-implemented GUI code 40, and displays the output text in the conversation history area, for example, as shown in Figure 2.

[0024] (Regarding the challenges) GUIs, with their intuitive and user-friendly visual elements, need to be designed and implemented specifically for certain interactions. Therefore, with conventional CAIs that pre-design and use GUIs, it is difficult to prepare GUIs suitable for various tasks handled by the system, limiting the tasks that can be effectively performed.

[0025] For example, in CAI systems, users may express requirements incompletely or inaccurately through instructions. This can lead to situations where the system's response fails to meet the user's requirements. To prevent this, as mentioned earlier, an approach of generating clarifying questions (CQs) is being considered, but answering these questions requires effort from the user.

[0026] Because it is difficult to anticipate the content of questions that will clarify the requirements, traditional CAIs, which use GUIs designed in advance, have the challenge of not being able to solve the above problems.

[0027] The following provides a detailed explanation of the system configuration and operation necessary to address the above challenges.

[0028] (System Configuration) Figure 4 shows an example of the configuration of the CAI system in this embodiment. As shown in Figure 4, this CAI system has an information processing system 100, a model 30, and a first GUI code 40. The information processing system 100 is one or more computers.

[0029] In this example, Model 30 is assumed to be a Large Language Model (LLM). However, Model 30 is not limited to an LLM. For example, Model 30 could be "a language model that has similar functionality to an LLM but is not called an LLM." Note that "language model" includes both language models that are not called LLMs and LLMs themselves.

[0030] The first GUI code 40 is a pre-implemented GUI code. The first GUI code is, for example, code written in HTML, CSS, or JavaScript. User 20 (person) accesses the information processing system 100. Note that "User 20" can be thought of as a user terminal operated by the user (person).

[0031] In the example shown in Figure 4, the model 30 and the pre-implemented GUI code 40 are located outside the information processing system 100 (e.g., on a server on the internet), but both the model 30 and the pre-implemented GUI code 40 may also be stored in an internal storage unit (not shown) of the information processing system 100.

[0032] (Operation Overview) In this embodiment, unlike the conventional technology, the information processing system 100 uses Model 30 to further generate a second GUI suitable for interaction with the user 20 according to the context of the conversation. Specifically, in each system turn, the GUI code (e.g., HTML) is generated by Model 30 and immediately rendered by the information processing system 100, and the second GUI is displayed to the user 20. With this configuration, an appropriate GUI can be generated as needed according to the context of the conversation, so that the CAI system can effectively perform a variety of tasks.

[0033] More specifically, in this embodiment, to solve the problem of users having to expend effort to respond to CQ, the information processing system 100 provides interactive widgets such as radio buttons and a calendar along with the CQ, and dynamically generates a second GUI that allows users to easily input their responses. By adopting this configuration, the information processing system 100 can reduce the burden on the user and improve the user interaction efficiency and usability of the CAI.

[0034] (Example of the configuration of the information processing system 100) Figure 5 shows an example of the functional configuration of the information processing system 100 that performs the above operations. As shown in Figure 5, the information processing system 100 includes an acquisition unit 110, a generation unit 120, and a display control unit 130. The information processing system 100 may also include a determination unit 140. The determination unit 140 will be described later.

[0035] The acquisition unit 110 acquires the input text from the user 20. The generation unit 120 uses the model 30 to acquire a second GUI code and generates a second GUI by executing the second GUI code. The display control unit 130 executes a process to display the generated second GUI on the display unit. For example, the display control unit 130 may display the second GUI on a display provided by the information processing system 100 as an example of a display unit, or it may display the second GUI on a user terminal connected to the information processing system 100. The operation example of the system will be described in detail below.

[0036] (Example of operation) In the CAI system of this embodiment, the initial processing (processing before conversation takes place) is the same as in the prior art. First, the information processing system 100 renders the first GUI using the pre-implemented first GUI code 40 and displays it on the display unit. This GUI may be the same as in the prior art (shown in Figure 2).

[0037] The processes that follow the initial processing (processing before the conversation takes place) will be explained according to the steps in the flowchart in Figure 6. Figure 6 is a flowchart that focuses on the display of the second GUI, which is a GUI for facilitating the input of responses to CQ. In some cases, steps S11 to S14 shown in Figure 6 may be repeated multiple times before the final response is obtained, or in other cases, steps S11 to S14 shown in Figure 6 may be performed only once before the final response is obtained.

[0038] <S11: User Input> User 20 inputs an instruction text, which is the input text, via the first GUI. For example, User 20 uses the keyboard to input an instruction text (e.g., "Tell me a recommended smartphone") into the text box and clicks the send button with the mouse. As a result, the instruction text is input into the information processing system 100. The acquisition unit 110 acquires the input instruction text (input text).

[0039] <S12: Display of user input> The display control unit 130 in the information processing system 100 displays the instruction text in the conversation history area.

[0040] <S13: GUI Code Generation> In S13, the generation unit 120 generates GUI code that displays questions to clarify the requirements intended by the user. Examples 1, 2, and 3 are described as variations of the processing method here. Examples 2 and 3 will mainly be described in terms of differences from Example 1.

[0041] <S13: Example 1> First, let's explain Example 1 of S13. In Example 1, GUI code is generated in two steps: a step to generate CQ and a step to generate GUI code. Specifically, the processes of S13-1 and S13-2 are executed as follows.

[0042] S13-1: First, the generation unit 120 of the information processing system 100 inputs a system prompt to the model 30 along with an instruction text in order to cause the model 30 to estimate the CQ.

[0043] For example, by using a system prompt based on a one-shot example, you can instruct the system to generate a specified number of CQs in a single inference. A specific example of a system prompt for generating CQs is as follows: --System prompt for CQ generation starts here--- Generate questions to confirm the user's specific situation in order to provide useful and specific responses based on the user's input. Please generate EXACTLY {number_of_questions} questions. Output in the following format: - List the questions in a list format. - Make sure to generate exactly three most important questions. - Make sure each question is concise and less than 15 words. Below is an example of input and output: User: Please provide advice on a weight loss plan. Assistant: ["What is your current weight and target weight?", "Do you have any dietary restrictions or medical conditions?"] --System prompt for CQ generation ends here--- Upon receiving the system prompt and input text, Model 30 generates a CQ according to the instructions of the system prompt and outputs it to the information processing system 100. The generation unit 120 of the information processing system 100 acquires the CQ.

[0044] S13-2: Next, the generation unit 120 inputs a system prompt for GUI code generation and the CQ acquired from the model 30 into the model 30, thereby instructing generation of second GUI code (e.g., HTML, etc.) that creates a second GUI for allowing the user 20 to easily answer the CQ.

[0045] The system prompt includes a description that instructs to appropriately use input elements such as a calendar, check box, radio button, or slider.

[0046] Note that the user 20 may set the setting items of the GUI to be displayed by the second GUI code. For example, the user 20 designates any one or more of a font type, a font size, and a color theme as GUI setting items for the information processing system 100. The generation unit 120 includes the setting items designated by the user 20 in the system prompt for GUI code generation.

[0047] The following is a concrete example of a system prompt for GUI code generation. This is an example of a prompt based on a one-shot example. --- System prompt for GUI code generation From here --- Generate usable GUI components (eg, calendars , checkboxes , radio buttons , ranges ) to minimize cognitive load for user inputs . Output format : - JSON with key 'qas ', value : list of dicts . Each dict has 'question ' ( string ) and ' answer ' ( HTML string ). - Always specify units for numerical inputs (eg, 'What is your budget ? (in dollars ) '). - Exclude markdown or formatting , output JSON string only . Example input : user : " Please give me advice on a weight loss plan ." assistant :[" What is your target duration for weight loss ?" ," What is your current exercise routine ?" ," Do you have any dietary preferences or restrictions ?"]. Example output : {" qas ":["{ in - context - example -1}" ,"{ in - context - example -2}" ,"{ in - context -example -3}"]}." in-context-example-1 { " question ":" < label for =\" weight - goal \"> What is your target duration for weight loss?< / label >", " answer ":" < input type =\" range \" id =\" duration \" name =\" duration \" min =\"1\" max=\"12\" value =\"3\" oninput =\" document . getElementById (' duration -value '). innerText= this . value \">< span id =\" duration - value \" class =\" slider - value \" >3 (months )"} in-context-example-2 { " question ":" < label for =\" exercise - habits \"> What is your current exercise routine ?< / label >"," answer ":" < label >< input type =\" radio \" name =\" exercise - habits \" value =\" none \">I don 't exercise < / label >. <label >< input type =\" radio \" name =\" exercise - habits \" value =\"light \"> Light exercise (e.g., walking , yoga ) < / label > <label >< input type =\" radio \"name =\" exercise - habits \" value =\" moderate \"> Moderate exercise (e.g., jogging , cycling ) < / label > <label >< input type =\" radio \" name =\" exercise - habits \" value =\"intense \"> Intense exercise (eg, gym workouts , running ) "} in-context-example-3 { " question ":" < label for =\" diet - preferences \"> Do you have any dietary preferences or restrictions ?< / label > ", " answer ":" < textarea id =\" diet - preferences \" name =\" diet - preferences \" placeholder =\" eg, vegetarian , gluten -free , no sugar \" >< / textarea > < label for =\" lifestyle \"> What are your main time or lifestyle constraints ? (eg, busy work schedule , night shifts ) < textarea id =\" lifestyle \" name =\" lifestyle \" placeholder =\" Enter your constraints \">< / textarea > "} ---End of system prompt for GUI code generation--- Model 30 receives input (system prompt and CQ) from the information processing system 100, generates GUI code based on the instructions of the system prompt, and returns the GUI code to the information processing system 100. The generation unit 120 acquires the GUI code.

[0048] More specifically, the model 30 generates second GUI code in which a name and a value are assigned to each element (e.g., an answer to a question) in accordance with the instruction content in the system prompt (e.g., …. name = "exercise-habits" value = "none">I don't exercise…….). The generation unit 120 of the information processing system 100 uses these names and values when converting the content of the user's response using the second GUI into text. The generated GUI code is included in the utterance text of the conversation context, and is taken into consideration in the subsequent system turn.

[0049] <S13: Example 2> In Example 2 of S13, the model 30 is caused to generate the CQ and the second GUI code in a single inference. Specifically, the generation unit 120 of the information processing system 100 inputs, to the model 30, an instruction sentence as input text, a "system prompt for CQ generation", and a "system prompt for second GUI code generation". Note that the "system prompt for CQ generation" and the "system prompt for second GUI code generation" may be regarded as a single system prompt.

[0050] The "system prompt for CQ generation" in Example 2 may be the same as the "system prompt for CQ generation" used in Example 1. When the "system prompt for CQ generation" used in Example 1 is used as the "system prompt for CQ generation" in Example 2, as the "system prompt for second GUI code generation" in Example 2, for example, it is possible to use a prompt obtained by adding, to the "system prompt for second GUI code generation" in Example 1, an instruction to use the inference result from the "system prompt for CQ generation" as the input CQ.

[0051] <S13: Example 3> Example 3 is a modified version of Example 1. In Example 3, after the generation unit 120 obtains the CQ from the model 30 in Example 1, it inputs a system prompt for GUI code generation, the CQ obtained from the model 30, and input text (= instruction text) into the model 30.

[0052] <S14: Display> The generation unit 120 generates (renders) the GUI by executing the GUI code received from the model 30. The display control unit 130 displays the generated GUI to the user 20. Specifically, for example, as shown in Figure 7, the GUI of the system utterance on the chat interface is displayed.

[0053] In the second GUI example shown in Figure 7, multiple options are displayed, for example, using radio buttons, to make it easier for user 20 to respond to the CQ, allowing them to select one option.

[0054] When user 20 operates the GUI and clicks the submit button, the form data is converted into a JSON string (e.g., "os":"OS-B","price":"700", etc.) based on the name and value of each element, and input into the information processing system 100 as the user's spoken text. This spoken text is acquired by the acquisition unit 110.

[0055] (Regarding the determination unit 140) In the explanation so far, we have assumed that a second GUI is created, but it is also possible to determine whether or not to create the GUI and create the GUI only if it is determined that the GUI should be created. When such a determination is made, the information processing system 100 includes the determination unit 140 shown in Figure 5. Below, we will explain examples of the operation of the information processing system 100 including the determination unit 140. Here, we will explain three examples: Example A, Example B, and Example C.

[0056] <Example A> Example A assumes the execution of Example 1 or Example 3 of S13 described above. In S13-1, the generation unit 120 inputs the "system prompt for CQ generation" and "input text" to the model 30 and obtains the CQ from the model 30. Subsequently, the determination unit 140 makes a query by sending a system prompt to the model 30 asking whether a GUI for CQ display is necessary. The determination unit 140 receives a response to the query from the model 30.

[0057] If the above answer is "GUI required", the determination unit 140 instructs the generation unit 120 to proceed to S13-2. If the above answer is "GUI not required", the determination unit 140 instructs the display control unit 130 to display the CQ acquired in S13-1.

[0058] <Example B> In Example B, we also assume that the case described above is performed in Example 1 or Example 3 of S13. In S13-1, the generation unit 120 inputs to the model 30 a "system prompt for generating CQ" and "input text", in addition to a "system prompt to inquire whether a GUI for displaying CQ is necessary".

[0059] If Model 30 determines that a GUI is needed, it returns a response indicating "GUI needed" along with CQ to the information processing system 100. The determination unit 140, having determined that a GUI is needed based on "GUI needed," instructs the generation unit 120 to proceed to S13-2.

[0060] If Model 30 determines that a GUI is not needed, it returns a response indicating "GUI not needed" along with CQ to the information processing system 100. The determination unit 140, having determined that a GUI is not needed based on "GUI not needed," instructs the display control unit 130 to display CQ.

[0061] <Example C> Example C assumes the execution of Example 2 of S13 described above. The generation unit 120 inputs a "system prompt for generating CQ", a "system prompt for generating GUI code", and "input text", in addition to a "system prompt to inquire whether a GUI for displaying CQ is necessary" to the model 30.

[0062] If Model 30 determines that a GUI is required, it returns a response indicating that a GUI is required, along with a GUI code, to the information processing system 100. The determination unit 140, having determined that a GUI is required based on the "GUI required" response, instructs the generation unit 120 and the display control unit 130 to execute the GUI code and display the GUI.

[0063] If Model 30 determines that a GUI is not needed, it returns a response indicating "GUI not needed" along with CQ to the information processing system 100. The determination unit 140, having determined that a GUI is not needed based on "GUI not needed," instructs the display control unit 130 to display CQ.

[0064] (Example Hardware Configuration) The information processing system 100 described in this embodiment can be realized, for example, by having a computer execute a program. This computer may be a physical computer or a virtual machine on the cloud. The information processing system 100 may also consist of multiple information processing devices that can be connected via a network.

[0065] In other words, the information processing system 100 can be realized by using hardware resources such as the CPU and memory built into the computer to execute a program corresponding to the processing performed by the information processing system 100. The above program can be recorded on a computer-readable recording medium (such as portable memory), saved, and distributed. It is also possible to provide the above program via a network such as the Internet or email.

[0066] Figure 8 shows an example of the hardware configuration of the computer described above. The computer in Figure 8 has a drive device 1000, an auxiliary storage device 1002, a memory device 1003, a CPU 1004, an interface device 1005, a display device 1006, an input device 1007, an output device 1008, etc., all of which are interconnected by bus B. The computer may also be equipped with a GPU.

[0067] The program that enables processing on the computer is provided on a recording medium 1001, such as a CD-ROM or memory card. When the recording medium 1001 containing the program is set in the drive device 1000, the program is installed from the recording medium 1001 to the auxiliary storage device 1002 via the drive device 1000. However, the program does not necessarily have to be installed from the recording medium 1001; it may also be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program as well as necessary files and data.

[0068] The memory device 1003 reads and stores a program from the auxiliary storage device 1002 when a program startup command is received. The CPU 1004 implements functions related to the information processing system 100 according to the program stored in the memory device 1003. The interface device 1005 is used as an interface for connecting to a network, etc. The display device 1006 displays a GUI (Graphical User Interface) etc., based on the program. The input device 1007 consists of a keyboard and mouse, buttons, or a touch panel, etc., and is used to input various operation commands. The output device 1008 outputs the calculation results.

[0069] Furthermore, the functions of the elements disclosed herein may be implemented using circuits or processing circuitry that include general-purpose processors, special-purpose processors, integrated circuits, ASICs (Application Specific Integrated Circuits), FPGAs (Field Programmable Gate Arrays), conventional circuits, and / or combinations thereof that are programmed using one or more programs stored in one or more memories, or otherwise configured to perform the disclosed functions. A processor is considered processing circuitry or circuitry because it includes transistors and other circuits. A processor may be a programmed processor that executes programs stored in memory. In this disclosure, a circuit, unit, or means is hardware that performs the enumerated functions, or hardware programmed to perform the enumerated functions. Hardware may be any hardware disclosed herein that is programmed or configured to perform the enumerated functions.

[0070] The system includes memory for storing a computer program containing computer instructions. These computer instructions provide logic and routines that enable hardware (e.g., processing circuitry or circuitry) to perform the methods disclosed herein. The computer program can be implemented in commonly known forms, such as computer-readable storage media, computer program products, memory devices, recording media such as CD-ROMs and DVDs, and / or memory in FPGAs and ASICs.

[0071] (Effects of the Embodiment) As described above, the technology described in this embodiment makes it possible to easily answer questions to clarify the requirements for an instruction in a conversational AI system.

[0072] In other words, when generating system questions to clarify unclear points in the requirements included in the prompts entered by the user into the system, the efficiency of the interaction can be improved by simultaneously generating user-friendly GUI elements (such as radio buttons or a calendar) that make it easy for the user to answer.

[0073] Furthermore, in this embodiment, when Model 30 generates a response, it generates not only the response text but also GUI code to display the response and accept user input. By rendering the generated GUI code, an appropriate GUI is realized according to the context of the conversation. In this way, by using a GUI that is suited to the conversation context, user interaction can be made more efficient in the various tasks handled by the CAI.

[0074] The following additional information is disclosed regarding the embodiments described above.

[0075] <Notes> (Note 1) An information processing system comprising: memory; and at least one processor connected to the memory, wherein the processor obtains input text entered by a user via a first GUI; generates a second GUI for inputting answers to questions for clarifying the user's intended requirements based on the input text, using a language model; and displays the second GUI. (Note 2) The information processing system according to Note 1, wherein the processor inputs a system prompt for generating the questions and the input text into the language model, obtains the questions from the language model; inputs a system prompt for generating code for generating the second GUI and the questions into the language model, and obtains the code from the language model. (Note 3) The information processing system according to Note 1, wherein the processor inputs a system prompt for generating the questions, a system prompt for generating code for generating the second GUI, and the input text into the language model, and obtains the code from the language model. (Note 4) An information processing system according to any one of Note 1 to 3, wherein the first GUI is generated based on pre-implemented code, and the second GUI is generated based on code generated by the generation unit. (Note 5) An information processing system according to any one of Note 1 to 4, wherein the second GUI includes a calendar, checkboxes, radio buttons, or a slider. (Note 6) An information processing system according to any one of Note 1 to 5, wherein the processor determines whether or not to generate the second GUI. (Note 7) An information processing method to be executed by an information processing system, comprising: an acquisition step of acquiring input text entered by a user via a first GUI; a generation step of generating a second GUI for inputting answers to questions for clarifying the requirements intended by the user based on the input text, using a language model; and a display control step of displaying the second GUI.(Appendix 8) A non-temporary storage medium storing a program for causing a computer to perform each step of the information processing method described in Appendix 7.

[0076] Although this embodiment has been described above, the present invention is not limited to this specific embodiment, and various modifications and changes are possible within the scope of the gist of the invention as described in the claims.

[0077] 10, 100 Information Processing System 20 User 30 Model 40 Pre-implemented GUI Code 110 Acquisition Unit 120 Generation Unit 130 Display Control Unit 140 Determination Unit 1000 Drive Device 1001 Recording Medium 1002 Auxiliary Storage Device 1003 Memory Device 1004 CPU 1005 Interface Device 1006 Display Device 1007 Input Device 1008 Output Device

Claims

1. An information processing system comprising: an acquisition unit that acquires input text entered by a user via a first GUI; a generation unit that generates a second GUI for inputting answers to questions that clarify the requirements intended by the user based on the input text, using a language model; and a display control unit that displays the second GUI.

2. The information processing system according to claim 1, wherein the generation unit inputs a system prompt for generating the question and the input text to the language model, obtains the question from the language model, inputs a system prompt for generating code for generating the second GUI and the question to the language model, and obtains the code from the language model.

3. The information processing system according to claim 1, wherein the generation unit inputs a system prompt for generating the question, a system prompt for generating code for generating the second GUI, and the input text into the language model, and obtains the code from the language model.

4. The information processing system according to claim 1, wherein the first GUI is generated based on pre-implemented code, and the second GUI is generated based on code generated by the generation unit.

5. The information processing system according to claim 1, wherein the second GUI includes a calendar, checkboxes, radio buttons, or a slider.

6. The information processing system according to claim 1, further comprising a determination unit for determining whether or not to generate the second GUI.

7. An information processing method to be executed by an information processing system, comprising: an acquisition step of acquiring input text entered by a user via a first GUI; a generation step of generating a second GUI for inputting answers to questions for clarifying the requirements intended by the user based on the input text, using a language model; and a display control step of displaying the second GUI.

8. A program for causing a computer to perform each step of the information processing method described in claim 7.

Images