Design assistance system

Abstract

Provided is a design assistance system suitable for grasping the order of creation or editing of elements.　A drawing creation assistance device 100 inputs, to a large-scale language model, a first request including inquiry information for inquiring first execution information for editing an editing element, and acquires first execution information output from the large-scale language model in response to the first request. Then the drawing creation assistance device 100 inputs, to the large-scale language model, a second request including inquiry information for inquiring second execution information for editing the editing element with respect to a result obtained by editing the editing element on the basis of the acquired first execution information, and execution result information related to the result, and acquires second execution information output from the large-scale language model in response to the second request.

Description

Design Support System

[0001] The present invention relates to a design support system, and more particularly to a design support system suitable for grasping the order in which elements are created or edited.

[0002] 2. Description of the Related Art Conventionally, a technique for supporting the creation of architectural drawings using AI (Artificial Intelligence) is known, for example, from the technique described in Patent Document 1.

[0003] The technology described in Patent Document 1 acquires an edited element that has been changed among editable edit elements in an architectural drawing, and estimates related elements and judgment necessity information from the acquired edited element using a trained model that has been trained based on information about the editable edit element in the architectural drawing, information about other edit elements and their contents that have been changed due to the change in the edited element, and judgment necessity information regarding whether human judgment is required to change the other edited elements.If it is determined based on the estimated judgment necessity information that human judgment is required to change the related element, the other edited elements are displayed in a specific manner based on the estimated related elements.

[0004] Patent No. 7341580 (

[0088] )

[0005] However, in the technique described in Patent Document 1, the work efficiency differs depending on the order in which elements are created or edited, and it is not clear in what order elements should be created or edited.

[0006] Therefore, the present invention has been made in consideration of the unresolved problems of the conventional technology, and aims to provide a design support system that is suitable for understanding the order in which elements are created or edited.

[0007] [Invention 1] In order to achieve the above object, a design support system of Invention 1 comprises: first request input means for inputting a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element in response to a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information related to the result; and second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0008] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element or editing a second element, the first implementation information acquisition means acquires the first implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element in response to a result of creating the first element or editing the second element based on the acquired first implementation information, and implementation result information related to the result, and the second implementation information acquisition means acquires the second implementation information output from the large-scale language model in response to the second request.

[0009] Here, the first inquiry information acquisition means may, for example, input inquiry information from an input device or the like, acquire or receive inquiry information from an external terminal or the like, read inquiry information from a storage device or storage medium or the like, or generate or calculate inquiry information by information processing or the like. Therefore, acquisition includes at least input, acquisition, reception, readout (including search), generation, and calculation. The same concept of acquisition applies hereinafter.

[0010] The implementation result information may include, for example, element information related to the first element created or the second element edited. The element information may be configured as, for example, information for identifying the first element or the second element (e.g., link information such as a name, number, ID, code, or URL). The element information may be configured as, for example, characters, numbers, figures, codes, symbols, images, sounds, or other information. The element information may be configured as keywords related to the first element or the second element (e.g., one or more keywords indicating part of the name of the first element or the second element). The same applies to the design support systems of Inventions 2 to 4.

[0011] Furthermore, this system may be realized as a single device, terminal, or other equipment, or as a network system in which multiple devices, terminals, or other equipment are connected so that they can communicate with each other. In the latter case, each component may belong to any of the multiple devices, etc., as long as they are connected so that they can communicate with each other. The same applies to the design support systems of Inventions 2 to 8 below.

[0012] [Invention 2] Furthermore, a design support system of Invention 2 comprises first request input means for inputting a first request to a large-scale language model, the first request including implementation result information related to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the result; first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including inquiry information for inquiring about second implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information related to the result; and second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0013] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including implementation result information on a result of creating the first element or editing the second element and including a request to generate first implementation information for creating the first element or editing the second element based on the result, the first implementation information acquisition means acquires the first implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the acquired first implementation information, and implementation result information on the result, and the second implementation information acquisition means acquires the second implementation information output from the large-scale language model in response to the second request.

[0014] [Invention 3] Furthermore, a design support system of Invention 3 comprises first request input means for inputting a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including implementation result information on a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the result; and second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0015] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element or editing a second element, the first implementation information acquisition means acquires first implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including implementation result information on a result of creating the first element or editing the second element based on the acquired first implementation information, and the second request including a request to generate second implementation information for creating the first element or editing the second element based on the result, and the second implementation information acquisition means acquires second implementation information output from the large-scale language model in response to the second request.

[0016] [Invention 4] Furthermore, a design support system of Invention 4 comprises first request input means for inputting a first request to a large-scale language model, the first request including implementation result information related to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the result; first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including implementation result information related to a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the result; and second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0017] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including implementation result information on a result of creating the first element or editing the second element and including a request to generate first implementation information for creating the first element or editing the second element based on the result, the first implementation information acquisition means acquires the first implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including implementation result information on a result of creating the first element or editing the second element based on the acquired first implementation information and including a request to generate second implementation information for creating the first element or editing the second element based on the result, and the second implementation information acquisition means acquires the second implementation information output from the large-scale language model in response to the second request.

[0018] [Invention 5] Furthermore, a design support system of Invention 5 comprises first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information, to a large-scale language model; first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request including inquiry information inquiring about implementation information for creating the first element or editing the second element, in response to a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the result; and second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0019] With this configuration, the first request input means inputs a first request including inquiry information inquiring about implementation information for creating a first element or editing a second element and its design information to the large-scale language model, the first implementation information acquisition means acquires implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request including inquiry information inquiring about implementation information for creating a first element or editing a second element in response to a result of creating the first element or editing the second element based on the acquired implementation information to the large-scale language model and design information including the result, and the second implementation information acquisition means acquires implementation information output from the large-scale language model in response to the second request.

[0020] [Invention 6] Furthermore, a design support system of Invention 6 comprises first request input means for inputting a first request to a large-scale language model, the first request including design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, and including a request to generate implementation information for creating the first element or editing the second element for the result; first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including inquiry information inquiring about implementation information for creating the first element or editing the second element for the result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the result; and second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0021] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including design information including a result of creating a first element or editing a second element, and including a request to generate implementation information for creating the first element or editing the second element based on the result, the first implementation information acquisition means acquires the implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including query information inquiring about implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the acquired implementation information, and design information including the result, and the second implementation information acquisition means acquires the implementation information output from the large-scale language model in response to the second request,

[0022] [Invention 7] Furthermore, a design support system of Invention 7 comprises first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information and the design information into a large-scale language model; first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including the design information including a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, the second request including a request to generate implementation information for creating the first element or editing the second element based on the result; and second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0023] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including inquiry information inquiring about implementation information for creating a first element or editing a second element and its design information, the first implementation information acquisition means acquires implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including design information including results of creating the first element or editing the second element based on the acquired implementation information, and the second request including a request to generate implementation information for creating the first element or editing the second element based on the results, and the second implementation information acquisition means acquires the implementation information output from the large-scale language model in response to the second request.

[0024] [Invention 8] Furthermore, a design support system of Invention 8 comprises first request input means for inputting a first request to a large-scale language model, the first request including design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, and including a request to generate implementation information for creating the first element or editing the second element based on the result; first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; second request input means for inputting a second request to the large-scale language model, the second request including the design information including a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and including a request to generate implementation information for creating the first element or editing the second element based on the result; and second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0025] With this configuration, the first request input means inputs a first request to the large-scale language model, the first request including design information including a result of creating a first element or editing a second element, and a request to generate implementation information for creating the first element or editing the second element based on the result, the first implementation information acquisition means acquires implementation information output from the large-scale language model in response to the first request, the second request input means inputs a second request to the large-scale language model, the second request including design information including a result of creating the first element or editing the second element based on the acquired implementation information, and a request to generate implementation information for creating the first element or editing the second element based on the result, and the second implementation information acquisition means acquires implementation information output from the large-scale language model in response to the second request.

[0026] [Invention 9] Furthermore, the design support system of Invention 9 is the design support system of any one of Inventions 1 to 8, wherein the first element is an element that requires human judgment for setting, and the setting affects other editable elements in the design information, and the second element is an element that requires human judgment for changing, and the change affects other editable elements in the design information.

[0027] [Invention 10] Furthermore, in the design support system of invention 10, in any one of the design support systems of inventions 1 to 8, the design information is design information for designing a building.

[0028] As described above, according to the design support system of Invention 1, first implementation information and second implementation information can be obtained in response to an inquiry, and second implementation information can be obtained for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be grasped.

[0029] Furthermore, according to the design support system of Invention 2, second implementation information can be obtained in response to an inquiry, and second implementation information can be obtained for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be understood.

[0030] Furthermore, according to the design support system of Invention 3, first implementation information can be obtained in response to an inquiry, and second implementation information can be obtained for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be understood.

[0031] Furthermore, according to the design support system of Invention 4, it is possible to obtain second implementation information for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be grasped.

[0032] Furthermore, according to the design support system of Invention 5, it is possible to obtain implementation information in response to an inquiry, and to obtain implementation information for creating or editing the results of creation or editing based on the implementation information, so that it is possible to grasp the order in which elements are created or edited.

[0033] Furthermore, according to the design support system of Invention 6, it is possible to obtain implementation information in response to an inquiry, and to obtain implementation information for creating or editing the results of creation or editing based on the implementation information, so that it is possible to grasp the order in which elements are created or edited.

[0034] Furthermore, according to the design support system of Invention 7, it is possible to obtain implementation information in response to an inquiry, and to obtain implementation information for creating or editing the results of creation or editing based on the implementation information, so that it is possible to grasp the order in which elements are created or edited.

[0035] Furthermore, according to the design support system of Invention 8, it is possible to obtain implementation information for creating or editing the results of creation or editing based on implementation information, so that the order in which elements are created or edited can be understood.

[0036] 1 is a diagram showing the hardware configuration of a drawing creation support device 100. FIG. 2 is a diagram showing the structure of learning data. FIG. 3 is a diagram showing the structure of CAD data of a plan drawing. FIG. 4 is a flowchart showing a trained model generation process. FIG. 5 is a block diagram showing the process of generating and using a trained model. FIG. 6 is a flowchart showing an implementation information estimation process. FIG. 7 is a diagram showing the structure of learning data. FIG. 8 is a flowchart showing an implementation information estimation process. FIG. 9 is a flowchart showing an implementation information estimation process.

[0037] First Embodiment A first embodiment of the present invention will now be described with reference to Figures 1 to 7.

[0038] First, the configuration of this embodiment will be described. FIG. 1 is a diagram showing the hardware configuration of a drawing creation support device 100.

[0039] As shown in FIG. 1, the drawing creation support device 100 comprises a CPU (Central Processing Unit) 30 that controls calculations and the entire system based on a control program, a ROM (Read Only Memory) 32 that stores the control program and the like for the CPU 30 in advance in a predetermined area, a RAM (Random Access Memory) 34 that stores data read from the ROM 32 and the like and calculation results required in the calculation process of the CPU 30, and an I / F (Interface) 38 that mediates the input and output of data to and from external devices, and these are connected to each other and capable of sending and receiving data by a bus 39, which is a signal line for transferring data.

[0040] The I / F 38 is connected to external devices such as an input device 40 consisting of a keyboard, mouse, etc. that can input data as a human interface, a storage device 42 that stores data, tables, etc. as files, and a display device 44 that displays a screen based on an image signal.

[0041] CAD (Computer Aided Design) software and BIM (Building Information Modeling) software (hereinafter collectively referred to as "CAD software") are installed in the storage device 42. CAD software is software that assists designers in creating drawings in response to their operations. When a request is made to start the CAD software, the CPU 30 starts a program for the CAD software stored in a predetermined area of ​​the ROM 32 and executes processing in accordance with the program. The designer can start the CAD software to create plan drawings, detailed floor plans, and other architectural drawings.

[0042] Next, we will explain the data structure of the storage device 42. Figure 2 is a diagram showing the structure of the learning data.

[0043] As shown in FIG. 2, the storage device 42 stores learning data that serves as training data in AI learning.

[0044] 2, the learning data includes query information 400 inquiring about implementation information 404 for creating a first element that can be created in CAD data or BIM data (hereinafter collectively referred to as "CAD data") or editing a second element that can be edited in the CAD data, implementation result information 402 regarding the results of creating the first element or editing the second element, and implementation information 404 for creating the first element or editing the second element. Hereinafter, the first element or the second element will be referred to as an "edited element," and creating the first element or editing the second element will be referred to as "editing."

[0045] In the first line of FIG. 2, "Install a piano with a width of 120 mm" is registered as inquiry information 400, and "Change the width of the toilet to 870 mm or less" is registered as implementation information 404. This indicates that if a designer makes an inquiry such as "I would like to install a piano with a width of 120 mm," the AI ​​should respond with "Please change the width of the toilet to 870 mm or less." This information can be created based on past editing history. That is, the implementation information 404 is extracted from the editing history data in which the designer edited the toilet to 870 mm or less in order to install a piano with a width of 120 mm, and is registered in the learning data in association with the inquiry information 400.

[0046] Here, the edit element "toilet" is a third element that requires human judgment to change, and the edit element "piano" is a fourth element that is affected by or impacts the change to the third element "toilet." Whether an edit element requires human judgment to change can be determined, for example, using the techniques described in Patent Documents 1 (Patent Documents 1)

[0076] to

[0089] . That is, edit elements that require human judgment to change may be set by human judgment. Alternatively, edit elements that have two or more change patterns before and after the change may be extracted based on the CAD data before and after the change, and the extracted edit elements may be set as edit elements that require human judgment to change. Based on the results, judgment necessity data is created and stored in the storage device 42, in which the edit elements are associated with information on whether or not the change requires human judgment. Then, edit elements determined to require human judgment to change based on the judgment necessity data in the storage device 42 are treated as third elements in the edit history data. On the other hand, the fourth element may include edit elements or other elements (e.g., furniture to be installed in the room or other virtual elements). The edit history data can be created by, for example, recording a change made by a designer to the third element in relation to the fourth element as edit history data.

[0047] The fourth line in FIG. 2 indicates the edit to be made after the edit on the first line. The fourth line registers "Instructions for toilet design" as inquiry information 400, "Toilet width: 870 mm, Hand wash basin width: 501 mm or more" as implementation result information 402, and "Change the hand wash basin width to 500 mm or less" as implementation information 404. This indicates that as a result of the designer editing the edit element "toilet," "the toilet width is 870 mm, the hand wash basin width is 501 mm or more." Therefore, if the designer next inquires, "Please give instructions for the toilet design," the AI ​​should respond, "Change the hand wash basin width to 500 mm or less." This information can be created based on past edit history. That is, the implementation result information 402 and implementation information 404 are extracted from the editing history data in which the designer edited the toilet width to 870 mm or less and the hand basin width to 500 mm or less, and are registered in the learning data in association with the inquiry information 400.

[0048] Here, the implementation information 404 in the first line is the first implementation information for editing the editing elements, and the implementation information 404 in the fourth line is the second implementation information for editing the editing elements based on the results of editing related to the first implementation information (implementation information 404 in the first line).

[0049] The seventh line in FIG. 2 indicates the edit to be made after the edit on the fourth line. The seventh line registers "Instructions for toilet design" as inquiry information 400, "The width of the hand wash basin is 500 mm or less, and the depth is 600 mm or more" as implementation result information 402, and "Ensure 400 mm or more of space on the left wall of the toilet" as implementation information 404. This indicates that the result of the designer editing the edit element "hand wash basin" is "The width of the hand wash basin is 500 mm or less, and the depth is 600 mm or more." Therefore, if the designer next inquires, "Please give instructions for the toilet design," the AI ​​should respond, "Ensure 400 mm or more of space on the left wall of the toilet." This information can be created based on past edit history. That is, the implementation result information 402 and implementation information 404 are extracted from the editing history data in which the designer edited the width of the hand wash basin to 500 mm or less and edited the space on the left wall of the toilet to 400 mm or more, and are registered in the learning data in correspondence with the inquiry information 400.

[0050] Here, the edit elements "hand wash basin" and "left wall of toilet" are related to the third and fourth elements. Also, the implementation information 404 on the fourth line is the second implementation information for editing the edit elements, and the implementation information 404 on the seventh line is the fourth implementation information for editing the edit elements based on the results of editing related to the second implementation information (the implementation information 404 on the fourth line).

[0051] In FIG. 2, the second, fifth, and eighth lines and the third, sixth, and ninth lines each indicate a series of edits, similar to the relationship between the first, fourth, and seventh lines.

[0052] 3 is a diagram showing the structure of CAD data of a rectangular plan view. As shown in FIG. 3, the storage device 42 stores CAD data of a rectangular plan view.

[0053] A rectangular plan is a detailed drawing of a building's cross section. CAD data for the plan is created by a designer using CAD software. The designer creates the plan in the CAD software by adding, editing, or deleting editable elements in the plan. In the example of Figure 3, the floor, wall, and ceiling edit elements are arranged in the area labeled "internal corridor," and the floor, wall, and ceiling edit elements are arranged in the area labeled "windbreak."

[0054] The storage device 42 stores CAD data of other architectural drawings and data of the trained model. The trained model is trained based on the training data of FIG. 2. Since the editing history data is used to create the training data, the storage device 42 stores a large amount of editing history data created in the past. This editing history data is data in which each editing element in the architectural drawing conforms to specifications.

[0055] Next, the operation of this embodiment will be described. Fig. 4 is a flowchart showing the trained model generation process.

[0056] 5 is a block diagram showing the steps of generating and using a trained model. The trained model generation process is a process executed to generate a trained model, and when executed by the CPU 30, the process proceeds to step S100, where a training data analysis process is executed, as shown in FIG. 4. In the training data analysis process, training data is read from the storage device 42, and inquiry information 400, implementation result information 402, and implementation information 404 are extracted from the read training data.

[0057] Next, the process proceeds to step S102. In step S102, as shown in FIG. 5, a training dataset is generated based on the information extracted in step S100. The process then proceeds to step S104, where the generated training dataset is input into a training program and a trained model is generated by the training program. The training program includes pre-training parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-training parameters. Trained models 1 to 3 are then output as the training results.

[0058] Trained model 1 is trained based on information inquiring about first implementation information for editing edit elements and the first implementation information. In the example of FIG. 2, training is performed based on the inquiry information 400 and implementation information 404 in the first to third lines. Trained model 1 includes trained parameters in which pre-training parameters have been updated through training, and an inference program. The inference program inputs inquiry information inquiring about the first implementation information, estimates the first implementation information from the input inquiry information based on the trained parameters, and outputs the estimated first implementation information. Note that the relationship between the input inquiry information and the output first implementation information is determined by AI training, and therefore, although it shows a similar trend to the content of past training data, there is ambiguity in that it does not necessarily match exactly. However, this ambiguity can be reduced by adjusting the amount of training data and the training accuracy. The same applies to trained models 2 and 3.

[0059] The trained model 2 is trained based on information inquiring about second implementation information for editing the edit element based on the results of editing related to the first implementation information, information related to the results, and the second implementation information. In the example of FIG. 2, training is performed based on the inquiry information 400, implementation result information 402, and implementation information 404 in lines 4 to 6. The trained model 2 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs inquiry information inquiring about second implementation information for editing the edit element based on the results of editing the edit element based on the first implementation information, and implementation result information related to the results, estimates second implementation information from the input inquiry information and implementation result information based on the trained parameters, and outputs the estimated second implementation information.

[0060] The trained model 3 is trained based on information inquiring about third implementation information for editing the edit element based on the results of editing related to the second implementation information, information related to the results, and the third implementation information. In the example of FIG. 2, training is performed based on the inquiry information 400, implementation result information 402, and implementation information 404 on lines 7 to 9. The trained model 3 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs inquiry information inquiring about third implementation information for editing the edit element based on the results of editing the edit element based on the second implementation information, and implementation result information related to the results, and estimates third implementation information from the input inquiry information and implementation result information based on the trained parameters, and outputs the estimated third implementation information.

[0061] Next, the process proceeds to step S106, where the trained models 1 to 3 generated in step S104 are stored in the storage device 42, and the series of processes ends.

[0062] 6 is a flowchart showing the implementation information estimation process. The implementation information estimation process is executed in response to a request from the designer, and when executed by the CPU 30, the process first proceeds to step S200 as shown in FIG.

[0063] In step S200, inquiry information inquiring about the first implementation information is acquired. The inquiry information may be acquired by voice input from a microphone (not shown) or by input from the input device 40. The same applies to steps S208 and S216.

[0064] Next, the process proceeds to step S204, where the first implementation information is estimated from the inquiry information acquired in step S200 using the trained model 1 in the storage device 42. If multiple pieces of inquiry information are acquired, the first implementation information is estimated for each piece of inquiry information. The estimation is performed by inputting the inquiry information into the trained model 1 and acquiring the first implementation information output from the trained model 1.

[0065] Next, the process proceeds to step S206, where the first implementation information estimated by the trained model 1 is displayed on the display device 44, the process proceeds to step S208, where inquiry information inquiring about the second implementation information is obtained, and the process proceeds to step S210.

[0066] In step S210, implementation result information is acquired regarding the results of the designer editing the edit element based on the first implementation information displayed on the display device 44. The implementation result information may be acquired from CAD software or CAD data, or may be acquired by inputting it from the input device 40. The same applies to step S218 below.

[0067] Next, the process proceeds to step S212, where the second implementation information is estimated from the inquiry information and implementation result information acquired in steps S208 and S210 using the trained model 2 in the storage device 42. If multiple pieces of inquiry information are acquired, the second implementation information is estimated for each piece of inquiry information. The estimation is performed by inputting the inquiry information and implementation result information into the trained model 2 and acquiring the second implementation information output from the trained model 2.

[0068] Next, the process proceeds to step S214, where the second implementation information estimated by the trained model 2 is displayed on the display device 44, and the process proceeds to step S216.

[0069] In step S216, inquiry information inquiring about the third implementation information is obtained, and then the process proceeds to step S218, where implementation result information regarding the results of the designer editing the editing elements based on the second implementation information displayed on the display device 44 is obtained, and then the process proceeds to step S220.

[0070] In step S220, the third implementation information is estimated from the inquiry information and implementation result information acquired in steps S216 and S218 using the trained model 3 in the storage device 42. If multiple pieces of inquiry information are acquired, the third implementation information is estimated for each piece of inquiry information. The estimation is performed by inputting the inquiry information and implementation result information into the trained model 3 and acquiring the third implementation information output from the trained model 3.

[0071] Next, the process proceeds to step S222, where the third implementation information estimated by the trained model 3 is displayed on the display device 44, and the series of processes is terminated.

[0072] [When a Piano is to be Carried In] Next, an operation when a piano is to be carried in will be described.

[0073] FIG. 7 shows a plan view of a room assuming the delivery of a piano. While editing the plan view of FIG. 7 in CAD software, a designer makes a query such as, "I would like to install a piano with a width of 120 mm." Steps S200 to S206 are then performed to acquire query information indicating the query, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is estimated and displayed. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information, and makes a query such as, "Please give me instructions regarding the toilet design." Steps S208 to S214 are then performed to acquire query information indicating the query and implementation result information indicating the editing results, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is estimated and displayed. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information and makes an inquiry saying, "Please give me instructions on the toilet design." Steps S216 to S222 are then carried out to obtain inquiry information indicating the inquiry and implementation result information indicating the edited results, and the third implementation information is estimated and displayed: "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications." The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0074] Here, the editing purposes "to prevent interference with the toilet wall," "to maintain the convenience of the toilet," and "to add barrier-free specifications" may be included in the implementation information 404 in Fig. 2 and learned, or may be learned based on other learning data. The purposes may be set manually in the implementation information 404 or other learning data, or may be set based on comments or the like entered by the designer during design.

[0075] Next, the effects of this embodiment will be described. In this embodiment, inquiry information inquiring about first implementation information is acquired, the first implementation information is estimated from the acquired inquiry information using the information inquiring about the first implementation information and a trained model 1 that has been trained based on the first implementation information, inquiry information inquiring about second implementation information for editing the edit elements based on the results of editing the edit elements based on the estimated first implementation information and implementation result information related to the results are acquired, and the second implementation information is estimated from the acquired inquiry information and implementation result information using information inquiring about second implementation information for editing the edit elements based on the results of editing related to the first implementation information, information related to the results, and a trained model 2 that has been trained based on the second implementation information.

[0076] This makes it possible to obtain first implementation information and second implementation information in response to an inquiry, and to obtain second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information, thereby making it possible to grasp the order in which the edited elements are edited.

[0077] Furthermore, in this embodiment, inquiry information inquiring about second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information and implementation result information related to the results are obtained, and second implementation information is estimated from the acquired inquiry information and implementation result information using information inquiring about second implementation information for editing the edited elements based on the results of editing related to the first implementation information, information related to the results, and a trained model 2 that has been trained based on the second implementation information, and inquiry information inquiring about third implementation information for editing the edited elements based on the estimated second implementation information and implementation result information related to the results are obtained, and third implementation information is estimated from the acquired inquiry information and implementation result information using information inquiring about third implementation information for editing the edited elements based on the results of editing related to the second implementation information, information related to the results, and a trained model 3 that has been trained based on the third implementation information.

[0078] This makes it possible to obtain second implementation information and third implementation information in response to an inquiry, and to obtain third implementation information for editing the edited elements based on the results of editing the edited elements based on the second implementation information, thereby making it possible to grasp the order in which the edited elements are edited.

[0079] Furthermore, in this embodiment, implementation information for editing a third element is estimated in relation to a third element whose change requires human judgment and a fourth element that is affected or affects the change to the third element.

[0080] This allows the order of editing to be understood, taking into account the relationship between the third and fourth elements.

[0081] Second Embodiment Next, a second embodiment of the present invention will be described. Fig. 8 is a diagram showing this embodiment. Figs. 2, 4, 5 and 7 are also used.

[0082] This embodiment differs from the first embodiment in that the second implementation information and the third implementation information are not queried. Only the differences from the first embodiment will be described below, and descriptions of overlapping parts will be omitted.

[0083] First, the operation of this embodiment will be described. When the trained model generation process is executed by the CPU 30, the process proceeds to step S100, where a training data analysis process is executed, as shown in Figure 4. In the training data analysis process, the training data is read from the storage device 42, and inquiry information 400, implementation result information 402, and implementation information 404 are extracted from the read training data.

[0084] Next, the process proceeds to step S102. In step S102, as shown in FIG. 5, a training dataset is generated based on the information extracted in step S100. The process then proceeds to step S104, where the generated training dataset is input into a training program and a trained model is generated by the training program. The training program includes pre-training parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-training parameters. Trained models 1 to 3 are then output as the training results.

[0085] Trained model 1 is the same as trained model 1 in the first embodiment. Trained model 2 is trained based on information regarding the results of editing related to the first implementation information and second implementation information for editing the edit elements based on the results. In the example of Figure 2, training is performed based on the implementation result information 402 and implementation information 404 in lines 4 to 6. Trained model 2 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs implementation result information regarding the results of editing the edit elements based on the first implementation information, estimates second implementation information from the input implementation result information based on the trained parameters, and outputs the estimated second implementation information.

[0086] The trained model 3 is trained based on information regarding the results of editing related to the second implementation information and third implementation information for editing the edit element based on the results. In the example of Figure 2, training is performed based on the implementation result information 402 and implementation information 404 on lines 7 to 9. The trained model 3 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs implementation result information regarding the results of editing the edit element based on the second implementation information, estimates third implementation information from the input implementation result information based on the trained parameters, and outputs the estimated third implementation information.

[0087] Next, the process proceeds to step S106, where the trained models 1 to 3 generated in step S104 are stored in the storage device 42, and the series of processes ends.

[0088] 8 is a flowchart showing the implementation information estimation process. When the implementation information estimation process is executed by the CPU 30, the process first proceeds to step S200, as shown in FIG. 8. The processes of steps S200 to S206 are the same as those in the first embodiment, and when the process of step S206 is completed, the process proceeds to step S210.

[0089] In step S210, implementation result information regarding the results of the designer editing the edit element based on the displayed first implementation information is obtained, and then the process proceeds to step S224, where second implementation information is estimated from the implementation result information obtained in step S210 using the trained model 2 in the memory device 42, the process proceeds to step S214, where the estimated second implementation information is displayed, and the process proceeds to step S218.

[0090] In step S218, implementation result information regarding the results of the designer editing the edit element based on the displayed second implementation information is obtained, and then the process proceeds to step S226, where third implementation information is estimated from the implementation result information acquired in step S218 using the trained model 3 in the memory device 42, and the process proceeds to step S222, where the estimated third implementation information is displayed, and the series of processes is terminated.

[0091] [When a Piano is to be Carried In] Next, an operation when a piano is to be carried in will be described.

[0092] 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 mm." Steps S200 to S206 are then carried out to acquire inquiry information indicating the inquiry, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is estimated and displayed. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information, and steps S210, S224, and S214 are carried out to acquire implementation result information indicating the editing results, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is estimated and displayed. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information, and through steps S218, S226, and S222, implementation result information showing the edit results is obtained, and the third implementation information, "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications," is estimated and displayed.The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0093] Next, the effects of this embodiment will be described. In this embodiment, inquiry information inquiring about first implementation information is acquired, the first implementation information is estimated from the acquired inquiry information using the information inquiring about the first implementation information and a trained model 1 that has been trained based on the first implementation information, implementation result information regarding the results of editing the edit element based on the estimated first implementation information is acquired, and the second implementation information is estimated from the acquired implementation result information using a trained model 2 that has been trained based on information regarding the results of editing related to the first implementation information and second implementation information for editing the edit element based on the results.

[0094] This allows first implementation information to be obtained in response to an inquiry, and second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information, so that the order in which the edited elements are edited can be understood.

[0095] Furthermore, in this embodiment, implementation result information regarding the results of editing the editing elements based on the first implementation information is obtained, and second implementation information is estimated from the obtained implementation result information using a trained model 2 that has been trained based on information regarding the results of editing related to the first implementation information and second implementation information for editing the editing elements based on the results, and implementation result information regarding the results of editing the editing elements based on the estimated second implementation information is obtained, and third implementation information is estimated from the obtained implementation result information using a trained model 3 that has been trained based on information regarding the results of editing related to the second implementation information and third implementation information for editing the editing elements based on the results.

[0096] This makes it possible to obtain third implementation information for editing the edit elements based on the results of editing the edit elements based on the second implementation information, thereby making it possible to grasp the order in which the edit elements are edited.

[0097] [Third embodiment] Next, a third embodiment of the present invention will be described. Figures 9 and 10 are diagrams showing this embodiment. Figures 4, 5 and 7 are also used.

[0098] This embodiment differs from the first embodiment in that CAD data is provided instead of implementation result information. Only the differences from the first embodiment will be described below, and descriptions of overlapping parts will be omitted.

[0099] First, we will explain the data structure of the storage device 42. Fig. 9 is a diagram showing the structure of the learning data.

[0100] As shown in FIG. 9, the storage device 42 stores learning data that serves as training data in AI learning.

[0101] As shown in FIG. 9, the learning data includes inquiry information 400, CAD data 406 (file name) including the results of editing the edit elements, and implementation information 404.

[0102] The first line in FIG. 9 registers "Install a piano with a width of 120 mm" as inquiry information 400, "file01.dxf" as CAD data 406, and "Change the width of the toilet to 870 mm or less" as implementation information 404. "file01.dxf" is the file name of the CAD data immediately before editing to change the toilet width to 870 mm or less. This indicates that if a designer inquires, "I would like to install a piano with a width of 120 mm," the AI ​​should respond, "Please change the toilet width to 870 mm or less." This information can be created based on past editing history. That is, from the editing history data in which the designer edited the toilet to change its width to 870 mm or less in order to install a piano with a width of 120 mm, the implementation information 404 is extracted and the CAD data 406 is identified, and the data is associated with the inquiry information 400 and registered in the learning data.

[0103] The fourth line in FIG. 9 indicates the edit to be made after the edit on the first line. The fourth line registers "Instructions for toilet design" as inquiry information 400, "file04.dxf" as CAD data 406, and "Change the width of the hand basin to 500 mm or less" as implementation information 404. "file04.dxf" is the CAD data edited to change the toilet width to 870 mm or less. This indicates that if a designer inquires, "Please give instructions for the toilet design," the AI ​​should respond, "Please change the width of the hand basin to 500 mm or less." This information can be created based on past editing history. That is, from the editing history data in which the designer made edits to change the toilet width to 870 mm or less and edits to change the width of the hand basin to 500 mm or less, the implementation information 404 is extracted and the CAD data 406 is identified, and the edits are associated with the inquiry information 400 and registered in the learning data.

[0104] The seventh line in FIG. 9 indicates the edit to be made after the edit on the fourth line. The seventh line registers "Instructions regarding toilet design" as inquiry information 400, "file07.dxf" as CAD data 406, and "Ensure at least 400 mm of space on the left wall of the toilet" as implementation information 404. "file07.dxf" registers CAD data edited to change the width of the hand basin to 500 mm or less. This indicates that when a designer inquires, "Please provide instructions regarding the toilet design," the AI ​​should respond, "Ensure at least 400 mm of space on the left wall of the toilet." This information can be created based on past edit history. That is, the designer performs edits to change the width of the hand wash basin to 500 mm or less and the space on the left wall of the toilet to 400 mm or more, and from the edit history data, the implementation information 404 is extracted and the CAD data 406 is identified, and the data is registered in the learning data in association with the inquiry information 400.

[0105] In FIG. 9, the second, fifth, and eighth lines and the third, sixth, and ninth lines each indicate a series of edits, similar to the relationship between the first, fourth, and seventh lines.

[0106] Next, the operation of this embodiment will be described. When the trained model generation process is executed by the CPU 30, the process proceeds to step S100, where a training data analysis process is executed, as shown in FIG. 4. In the training data analysis process, the training data is read from the storage device 42, and the inquiry information 400, the CAD data 406, and the implementation information 404 are extracted from the read training data.

[0107] Next, the process proceeds to step S102. In step S102, as shown in FIG. 5, a training dataset is generated based on the information extracted in step S100. The process then proceeds to step S104, where the generated training dataset is input into a training program and a trained model is generated by the training program. The training program includes pre-training parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-training parameters. Trained models 1 to 3 are then output as the training results.

[0108] The trained model 1 is trained based on information inquiring about first implementation information for editing an edit element, CAD data immediately before editing the edit element, and the first implementation information. In the example of FIG. 9 , training is performed based on the inquiry information 400, CAD data 406, and implementation information 404 in the first to third lines. The trained model 1 includes trained parameters in which pre-training parameters have been updated by training, and an inference program. The inference program inputs the inquiry information inquiring about the first implementation information and the CAD data immediately before editing, estimates the first implementation information from the input inquiry information and CAD data based on the trained parameters, and outputs the estimated first implementation information.

[0109] The trained model 2 is trained based on information querying for second implementation information for editing the edit element based on the results of editing related to the first implementation information, CAD data including the results, and the second implementation information. In the example of FIG. 9 , training is performed based on the inquiry information 400, CAD data 406, and implementation information 404 in lines 4 to 6. The trained model 2 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs inquiry information querying for second implementation information for editing the edit element based on the results of editing the edit element based on the first implementation information, and CAD data including the results, estimates the second implementation information from the input inquiry information and CAD data based on the trained parameters, and outputs the estimated second implementation information.

[0110] The trained model 3 is trained based on information querying third implementation information for editing the edit element based on the results of editing related to the second implementation information, CAD data including the results, and the third implementation information. In the example of FIG. 9 , training is performed based on the inquiry information 400, CAD data 406, and implementation information 404 on lines 7 to 9. The trained model 3 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs inquiry information querying third implementation information for editing the edit element based on the results of editing the edit element based on the second implementation information, and CAD data including the results, estimates the third implementation information from the input inquiry information and CAD data based on the trained parameters, and outputs the estimated third implementation information.

[0111] Next, the process proceeds to step S106, where the trained models 1 to 3 generated in step S104 are stored in the storage device 42, and the series of processes ends.

[0112] 10 is a flowchart showing the implementation information estimation process. When the implementation information estimation process is executed by the CPU 30, as shown in FIG.

[0113] In step S300, inquiry information inquiring about the first implementation information is acquired, and the process proceeds to step S302 to acquire the CAD data immediately before editing, and the process proceeds to step S304.

[0114] In step S304, using the trained model 1 in the memory device 42, first implementation information is estimated from the inquiry information and CAD data acquired in steps S300 and S302, and the process proceeds to step S306, where the first implementation information estimated by the trained model 1 is displayed on the display device 44, and the process proceeds to step S308.

[0115] In step S308, inquiry information inquiring about the second implementation information is obtained, and then the process proceeds to step S310, where CAD data including the results of the designer editing the editing elements based on the first implementation information displayed on the display device 44 is obtained, and the process proceeds to step S312.

[0116] In step S312, using the trained model 2 in the memory device 42, second implementation information is estimated from the inquiry information and CAD data acquired in steps S308 and S310, and the process proceeds to step S314, where the second implementation information estimated by the trained model 2 is displayed on the display device 44, and the process proceeds to step S316.

[0117] In step S316, inquiry information inquiring about the third implementation information is obtained, and then the process proceeds to step S318, where CAD data including the results of the designer editing the editing elements based on the second implementation information displayed on the display device 44 is obtained, and the process proceeds to step S320.

[0118] In step S320, the third implementation information is estimated from the inquiry information and CAD data acquired in steps S316 and S318 using the trained model 3 in the storage device 42. If multiple pieces of inquiry information are acquired, the third implementation information is estimated for each piece of inquiry information, and the process proceeds to step S322, where the third implementation information estimated by the trained model 3 is displayed on the display device 44, and the series of processes is terminated.

[0119] [When a Piano is to be Carried In] Next, an operation when a piano is to be carried in will be described.

[0120] While editing the plan of Figure 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 [mm]." Steps S300 to S306 are then performed to obtain the inquiry information indicating the inquiry and the CAD data immediately before the editing, and the first implementation information, "Please change the width of the toilet to 870 [mm] or less to prevent interference with the toilet wall," is estimated and displayed. Next, the designer changes the width of the toilet to 870 [mm] based on the first implementation information, and makes an inquiry such as "Please give me instructions regarding the toilet design." Steps S308 to S314 are then performed to obtain the inquiry information indicating the inquiry and the CAD data including the editing results, and the second implementation information, "Please change the width of the hand basin to 500 [mm] or less to maintain the convenience of the toilet," is estimated and displayed. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information and makes an inquiry saying, "Please give me instructions on the toilet design." Steps S316 to S322 are then carried out to obtain CAD data including the inquiry information indicating the inquiry and the edited results, and the third implementation information is estimated and displayed: "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications." The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0121] Next, the effects of this embodiment will be described. In this embodiment, inquiry information inquiring about first implementation information and CAD data immediately before editing are acquired, the first implementation information is estimated from the acquired inquiry information and CAD data using a trained model 1 that has been trained based on the information inquiring about the first implementation information, the CAD data, and the first implementation information, inquiry information inquiring about second implementation information for editing the edit elements based on the results of editing the edit elements based on the estimated first implementation information and CAD data including the results, and the second implementation information is estimated from the acquired inquiry information and CAD data using a trained model 2 that has been trained based on information inquiring about second implementation information for editing the edit elements based on the results of editing related to the first implementation information, the CAD data including the results, and the second implementation information.

[0122] This makes it possible to obtain first implementation information and second implementation information in response to an inquiry, and to obtain second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information, thereby making it possible to grasp the order in which the edited elements are edited.

[0123] Furthermore, in this embodiment, inquiry information inquiring about second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information and CAD data including the results are obtained, and the second implementation information is estimated from the acquired inquiry information and CAD data using information inquiring about second implementation information for editing the edited elements based on the results of editing related to the first implementation information, CAD data including the results, and a trained model 2 trained based on the second implementation information, and inquiry information inquiring about third implementation information for editing the edited elements based on the results of editing related to the estimated second implementation information and CAD data including the results are obtained, and the third implementation information is estimated from the acquired inquiry information and CAD data using information inquiring about third implementation information for editing the edited elements based on the results of editing related to the second implementation information, CAD data including the results, and a trained model 3 trained based on the third implementation information.

[0124] This makes it possible to obtain second implementation information and third implementation information in response to an inquiry, and to obtain third implementation information for editing the edited elements based on the results of editing the edited elements based on the second implementation information, thereby making it possible to grasp the order in which the edited elements are edited.

[0125] [Fourth embodiment] Next, a fourth embodiment of the present invention will be described. Fig. 11 is a diagram showing this embodiment. Figs. 4, 5, 7 and 9 are also used.

[0126] This embodiment differs from the third embodiment in that the second implementation information and the third implementation information are not queried. Hereinafter, only the differences from the first and third embodiments will be described, and the description of the overlapping parts will be omitted.

[0127] First, the operation of this embodiment will be described. When the trained model generation process is executed by the CPU 30, the process proceeds to step S100, where a training data analysis process is executed, as shown in FIG. 4. In the training data analysis process, the training data is read from the storage device 42, and the inquiry information 400, the CAD data 406, and the implementation information 404 are extracted from the read training data.

[0128] Next, the process proceeds to step S102. In step S102, as shown in FIG. 5, a training dataset is generated based on the information extracted in step S100. The process then proceeds to step S104, where the generated training dataset is input into a training program and a trained model is generated by the training program. The training program includes pre-training parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-training parameters. Trained models 1 to 3 are then output as the training results.

[0129] Trained model 1 is the same as trained model 1 in the third embodiment. Trained model 2 is trained based on CAD data including the results of editing related to the first implementation information and second implementation information for editing the edit elements based on the results. In the example of FIG. 9 , training is performed based on the CAD data 406 and implementation information 404 in the fourth to sixth lines. Trained model 2 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs CAD data including the results of editing the edit elements based on the first implementation information, estimates second implementation information from the input CAD data based on the trained parameters, and outputs the estimated second implementation information.

[0130] The trained model 3 is trained based on CAD data including the results of editing related to the second implementation information and third implementation information for editing the edit elements based on the results. In the example of Figure 9, training is performed based on the CAD data 406 and implementation information 404 on lines 7 to 9. The trained model 3 includes trained parameters in which the pre-learning parameters have been updated by training, and an inference program. The inference program inputs CAD data including the results of editing the edit elements based on the second implementation information, estimates third implementation information from the input CAD data based on the trained parameters, and outputs the estimated third implementation information.

[0131] Next, the process proceeds to step S106, where the trained models 1 to 3 generated in step S104 are stored in the storage device 42, and the series of processes ends.

[0132] 11 is a flowchart showing the implementation information estimation process. When the implementation information estimation process is executed by the CPU 30, the process first proceeds to step S300, as shown in FIG. 11. The processes of steps S300 to S306 are the same as those in the first embodiment, and when the process of step S306 is completed, the process proceeds to step S310.

[0133] In step S310, CAD data including the results of the designer editing the editing elements based on the displayed first implementation information is acquired, and the process proceeds to step S324, where second implementation information is estimated from the CAD data acquired in step S310 using the trained model 2 in the memory device 42, and the process proceeds to step S314, where the estimated second implementation information is displayed, and the process proceeds to step S318.

[0134] In step S318, CAD data including the results of the designer editing the edit elements based on the displayed second implementation information is acquired, and the process proceeds to step S326, where third implementation information is estimated from the CAD data acquired in step S318 using the trained model 3 in the memory device 42, and the process proceeds to step S322, where the estimated third implementation information is displayed, and the series of processes is terminated.

[0135] [When a Piano is to be Carried In] Next, an operation when a piano is to be carried in will be described.

[0136] While editing the plan of Figure 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 [mm]." Steps S300 to S306 are then performed to obtain the inquiry information and the CAD data immediately before editing, and the first implementation information, "Please change the width of the toilet to 870 [mm] or less to prevent interference with the toilet wall," is estimated and displayed. Next, the designer changes the width of the toilet to 870 [mm] based on the first implementation information. Steps S310, S324, and S314 are then performed to obtain the CAD data including the editing results, and the second implementation information, "Please change the width of the hand basin to 500 [mm] or less to maintain the convenience of the toilet," is estimated and displayed. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information, and CAD data including the edited results is acquired through steps S318, S326, and S322, and the third implementation information, "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications," is estimated and displayed.The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0137] Next, the effects of this embodiment will be described. In this embodiment, inquiry information inquiring about first implementation information and CAD data immediately before editing are acquired, the first implementation information is estimated from the acquired inquiry information and CAD data using a trained model 1 that has been trained based on the information inquiring about the first implementation information, the CAD data, and the first implementation information, CAD data including the results of editing the edit elements based on the estimated first implementation information is acquired, and the second implementation information is estimated from the acquired CAD data using a trained model 2 that has been trained based on the CAD data including the results of editing related to the first implementation information and second implementation information for editing the edit elements based on the results.

[0138] This allows first implementation information to be obtained in response to an inquiry, and second implementation information for editing the edited elements based on the results of editing the edited elements based on the first implementation information, so that the order in which the edited elements are edited can be understood.

[0139] Furthermore, in this embodiment, CAD data including the results of editing the editing elements based on the first implementation information is acquired, and second implementation information is estimated from the acquired CAD data using a trained model 2 that has been trained based on the CAD data including the results of editing related to the first implementation information and second implementation information for editing the editing elements based on the results, and CAD data including the results of editing related to the estimated second implementation information is acquired, and third implementation information is estimated from the acquired CAD data using a trained model 3 that has been trained based on the CAD data including the results of editing related to the second implementation information and third implementation information for editing the editing elements based on the results.

[0140] This makes it possible to obtain third implementation information for editing the edit elements based on the results of editing the edit elements based on the second implementation information, thereby making it possible to grasp the order in which the edit elements are edited.

[0141] [Variant] In the above-described first embodiment and its variant, the trained model 2 is one that has been trained based on inquiry information, implementation result information, and implementation information, but this is not limited to this, and one that has been trained based on implementation result information and implementation information can also be used.

[0142] Furthermore, in the above-described first and second embodiments and their modifications, the trained models 2 and 3 can be configured as a single trained model.

[0143] In addition, in the above first and second embodiments and their variations, the trained models 1 to 3 can be configured as one or more trained models that have been trained based on inquiry information, implementation result information, and implementation information.

[0144] Furthermore, in the above-mentioned third embodiment and its variant, the trained model 2 is one that has been trained based on inquiry information, CAD data, and implementation information, but this is not limited to this, and one that has been trained based on CAD data and implementation information can also be used.

[0145] Furthermore, in the third and fourth embodiments and their variations, the trained models 2 and 3 can be configured as a single trained model.

[0146] In addition, in the above third and fourth embodiments and their variations, the trained models 1 to 3 can be configured as one or more trained models that have been trained based on inquiry information, CAD data, and implementation information.

[0147] Furthermore, in the above first to fourth embodiments and their variations, three stages of implementation information (first implementation information → second implementation information → third implementation information) were estimated, but this is not limited to this, and one or two stages of implementation information or four or more stages of implementation information can be estimated.

[0148] In the first to fourth embodiments and their modifications, the implementation information indicating one creation or editing is estimated, but the invention is not limited to this. It is also possible to estimate implementation information indicating multiple creations or editings (multiple options). The designer selects and performs one of the multiple creations or editings, and then estimates the implementation information indicating the next creation or editing based on the result.

[0149] In the first to fourth embodiments and their modifications, the first implementation information is queried, but this is not limiting, and a configuration in which the first implementation information is not queried may be adopted. In this case, the processing from step S204 or S302 onward of the processing shown in the flowcharts of Figures 6, 8, 10, or 11 may be executed in response to some trigger (for example, a request from the designer).

[0150] In addition, in the first to fourth embodiments and their modifications, a trained model is used, but this is not limiting, and a configuration using a large language model can also be adopted. Specifically, for example, the following configuration can be adopted.

[0151] The drawing creation support device 100 is connected via a network to a large-scale language model server having a large-scale language model. The large-scale language model server has a hardware configuration similar to that of the drawing creation support device 100. The large-scale language model is a deep learning model that pre-trains a language model, which models human spoken language based on its occurrence probability, from a massive amount of data. Upon receiving a request, the large-scale language model server uses the large-scale language model to statistically estimate the probability of generating the next word from the sentence included in the received request and transmits the estimation result to the requestor. For example, known technologies described in the internet sites "https: / / chatgpt-lab.com / n / n418d3aa56f0b" and "https: / / agirobots.com / chatgpt-mechanism-and-problem / " can be used as the large-scale language model.

[0152] The following inventions A1 to A8 can be applied to the drawing creation support device 100.

[0153] [Invention A1] A system for implementing a method for implementing a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; first implementation information acquisition means acquiring first implementation information output from the large-scale language model in response to the first request; second request input means inputting a second request to the large-scale language model including inquiry information inquiring about second implementation information for creating the first element or editing the second element in response to a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information related to the result; and second implementation information acquisition means acquiring second implementation information output from the large-scale language model in response to the second request.

[0154] This makes it possible to obtain first implementation information and second implementation information in response to an inquiry, and to obtain second implementation information for creating or editing the results of creation or editing based on the first implementation information, thereby making it possible to understand the order in which elements are created or edited.

[0155] [Invention A2] The system comprises: a first request input means for inputting a first request to a large-scale language model, the first request including implementation result information relating to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the result; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information relating to the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0156] This allows second implementation information to be obtained in response to an inquiry, and second implementation information for creating or editing the results of creation or editing based on the first implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0157] The following describes embodiments A1 and A2 as modifications of the first embodiment, in which the processes of steps S204, S212, and S220 are replaced with processes for obtaining implementation information from a large-scale language model server.

[0158] While editing the plan of FIG. 7 in CAD software, a designer makes a query such as, "I would like to install a piano with a width of 120 mm." Steps S200 and S204 are then performed to acquire query information indicating the query, and a request is sent to the large-scale language model server. The request includes, for example, query information such as, "What editing should I do and how to install a piano with a width of 120 mm?" In response to this request, the large-scale language model server outputs first implementation information. Upon receiving the first implementation information, the drawing creation support device 100 displays, in step S206, the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the wall of the toilet." Next, the designer changes the width of the toilet to 870 mm based on the first implementation information and makes a query such as, "Please give me instructions on the design of the toilet." Steps S208 to S212 are then performed to acquire query information indicating the query and implementation result information indicating the editing results, and a request is sent to the large-scale language model server. The request includes, for example, inquiry information such as "What should I edit next in the toilet design and how?" and the acquired implementation result information. In response to this request, the large-scale language model server outputs second implementation information. When the drawing creation support device 100 receives the second implementation information, it displays, via step S214, the second implementation information, "To maintain the convenience of the toilet, please change the width of the hand wash basin to 500 mm or less." Next, the designer changes the width of the hand wash basin to 500 mm based on the second implementation information and makes an inquiry such as "Please give me instructions on the toilet design." In steps S216 to S220, inquiry information indicating the inquiry and implementation result information indicating the editing results are acquired, and the request is sent to the large-scale language model server. The request includes, for example, inquiry information such as "What should I edit next in the toilet design and how?" and the acquired implementation result information. In response to this request, the large-scale language model server outputs third implementation information.When the drawing creation support device 100 receives the third implementation information, it goes through step S222 and displays the third implementation information, "To add barrier-free specifications, please ensure there is at least 400 mm of space on the left wall of the toilet." Then, the designer changes the space on the left wall of the toilet to 400 mm based on the third implementation information.

[0159] In this case, step S204 corresponds to the first request input means of invention 1 or the first implementation information acquisition means of invention 1, step S212 corresponds to the second request input means of invention 1 or the second implementation information acquisition means of invention 1, and the CAD data corresponds to the design information of inventions 1, 2, or 10. Also, step S212 corresponds to the first request input means of invention 2 or the first implementation information acquisition means of invention 2, and step S220 corresponds to the second request input means of invention 2 or the second implementation information acquisition means of invention 2.

[0160] [Invention A3] A system includes a first request input means for inputting a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including implementation result information on a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0161] This allows first implementation information to be obtained in response to an inquiry, and second implementation information to be obtained for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be understood.

[0162] [Invention A4] The system comprises a first request input means for inputting a first request to a large-scale language model, the first request including implementation result information relating to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the result; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including implementation result information relating to a result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

[0163] This makes it possible to obtain second implementation information for creating or editing the results of creation or editing based on the first implementation information, thereby making it possible to understand the order in which elements are created or edited.

[0164] The following describes embodiments A3 and A4 as modifications of the second embodiment, in which the processes of steps S204, S224, and S226 are replaced with processes for acquiring implementation information from a large-scale language model server.

[0165] While editing the plan of FIG. 7 in CAD software, the designer makes a query such as, "I would like to install a piano with a width of 120 mm." In response to this query, inquiry information indicating the query is acquired through steps S200 and S204, and a request is sent to the large-scale language model server. The request includes, for example, inquiry information such as, "What should I edit and how should I edit to install a piano with a width of 120 mm?" In response to this request, the large-scale language model server outputs first implementation information. Upon receiving the first implementation information, the drawing creation support device 100 displays, through step S206, the first implementation information: "Please change the width of the toilet to 870 mm or less to prevent interference with the wall of the toilet." Next, the designer changes the width of the toilet to 870 mm based on the first implementation information. In response to this query, inquiry information indicating the editing results is acquired through steps S210 and S224, and a request is sent to the large-scale language model server. The request includes, for example, the acquired implementation result information and a request such as "Please generate implementation information for editing edit elements in the toilet design." In response to this request, the large-scale language model server outputs second implementation information. Upon receiving the second implementation information, the drawing creation support device 100 displays, via step S214, the second implementation information, "To maintain the convenience of the toilet, please change the width of the hand wash basin to 500 mm or less." Next, the designer changes the width of the hand wash basin to 500 mm based on the second implementation information. Then, via steps S218 and S226, implementation result information indicating the editing results is acquired, and a request is sent to the large-scale language model server. The request includes, for example, the acquired implementation result information and a request such as "Please generate implementation information for editing edit elements in the toilet design." In response to this request, the large-scale language model server outputs third implementation information. When the drawing creation support device 100 receives the third implementation information, it goes through step S222 and displays the third implementation information, "To add barrier-free specifications, please ensure there is at least 400 mm of space on the left wall of the toilet." Then, the designer changes the space on the left wall of the toilet to 400 mm based on the third implementation information.

[0166] In this case, step S204 corresponds to the first request input means or the first implementation information acquisition means of invention 3, and step S224 corresponds to the second request input means or the second implementation information acquisition means of invention 3. Also, step S224 corresponds to the first request input means or the first implementation information acquisition means of invention 4, and step S226 corresponds to the second request input means or the second implementation information acquisition means of invention 4.

[0167] [Invention A5] A system includes a first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information, into a large-scale language model; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request including inquiry information inquiring about implementation information for creating the first element or editing the second element in response to a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the result; and a second implementation information acquisition means for acquiring the implementation information output from the large-scale language model in response to the second request.

[0168] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0169] [Invention A6] A system includes a first request input means for inputting a first request to a large-scale language model, the first request including design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, and including a request to generate implementation information for creating the first element or editing the second element based on the result; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including inquiry information inquiring about implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the result; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0170] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0171] The following describes embodiments A5 and A6 as modifications of the third embodiment, in which the processes of steps S304, S312, and S320 are replaced with processes for obtaining implementation information from a large-scale language model server.

[0172] While editing the plan of FIG. 7 in the CAD software, the designer makes a query such as, "I would like to install a piano with a width of 120 mm." Steps S300 to S304 are then performed to acquire query information indicating the query and the CAD data immediately prior to editing, and a request is sent to the large-scale language model server. The request includes, for example, query information such as, "What should I edit and how to edit to install a piano with a width of 120 mm?" and the acquired CAD data. In response to this request, the large-scale language model server outputs first implementation information. Upon receiving the first implementation information, the drawing creation support device 100 displays, in step S306, the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the wall of the toilet." Next, the designer changes the width of the toilet to 870 mm based on the first implementation information and makes an inquiry, "Please give me instructions on the toilet design." Then, through steps S308 to S312, CAD data including inquiry information indicating the inquiry and the edited results are acquired, and a request is sent to the large-scale language model server. The request includes, for example, inquiry information such as "What should I edit next and how in the toilet design?" and the acquired CAD data. In response to this request, the large-scale language model server outputs second implementation information. Upon receiving the second implementation information, the drawing creation support device 100 displays, through step S314, the second implementation information: "To maintain the convenience of the toilet, please change the width of the hand wash basin to 500 mm or less." Next, the designer changes the width of the hand wash basin to 500 mm based on the second implementation information and makes an inquiry, "Please give me instructions on the toilet design." Then, through steps S316 to S320, CAD data including inquiry information indicating the inquiry and the edited results are acquired, and a request is sent to the large-scale language model server. The request includes, for example, inquiry information such as "What should I edit next and how in the toilet design?" and the acquired CAD data. In response to this request, the large-scale language model server outputs third implementation information.When the drawing creation support device 100 receives the third implementation information, it goes through step S322 and displays the third implementation information, "To add barrier-free specifications, please ensure there is at least 400 mm of space on the left wall of the toilet." Then, the designer changes the space on the left wall of the toilet to 400 mm based on the third implementation information.

[0173] In this case, step S304 corresponds to the first request input means or the first implementation information acquisition means of invention 5, and step S312 corresponds to the second request input means or the second implementation information acquisition means of invention 5. Also, step S312 corresponds to the first request input means or the first implementation information acquisition means of invention 6, and step S320 corresponds to the second request input means or the second implementation information acquisition means of invention 6.

[0174] [Invention A7] A system includes a first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information and the design information into a large-scale language model; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including the design information including a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, the second request including a request to generate implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring the implementation information output from the large-scale language model in response to the second request.

[0175] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0176] [Invention A8] A system including a first request input means for inputting a first request to a large-scale language model, the first request including design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, the first request including a request to generate implementation information for creating the first element or editing the second element based on the result; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including the design information including a result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, the second request including a request to generate implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

[0177] This makes it possible to obtain the implementation information for creating or editing the result of creating or editing based on the implementation information, and therefore makes it possible to grasp the order in which elements are created or edited.

[0178] The following describes embodiments A7 and A8 as modifications of the fourth embodiment, in which the processes of steps S304, S324, and S326 are replaced with processes for acquiring implementation information from a large-scale language model server.

[0179] While editing the plan of FIG. 7 in CAD software, a designer makes a query such as, "I would like to install a piano with a width of 120 mm." Steps S300 to S304 are then performed to acquire query information indicating the query and CAD data immediately before editing, and a request is sent to the large-scale language model server. The request includes, for example, query information such as, "What should I edit and how to edit to install a piano with a width of 120 mm?" and the acquired CAD data. In response to this request, the large-scale language model server outputs first implementation information. Upon receiving the first implementation information, the drawing creation support device 100 displays, in step S306, the first implementation information: "Please change the width of the toilet to 870 mm or less to prevent interference with the wall of the toilet." Next, the designer changes the width of the toilet to 870 mm based on the first implementation information. Steps S310 and S324 are then performed to acquire CAD data including the edited results, and a request is sent to the large-scale language model server. The request includes, for example, the acquired CAD data and a request such as "Please generate implementation information for editing edit elements in the toilet design." In response to this request, the large-scale language model server outputs second implementation information. Upon receiving the second implementation information, the drawing creation support device 100 displays, via step S314, the second implementation information, "To maintain the convenience of the toilet, please change the width of the hand wash basin to 500 mm or less." Next, the designer changes the width of the hand wash basin to 500 mm based on the second implementation information. Then, via steps S318 and S326, CAD data including the editing results is acquired, and a request is sent to the large-scale language model server. The request includes, for example, the acquired CAD data and a request such as "Please generate implementation information for editing edit elements in the toilet design." In response to this request, the large-scale language model server outputs third implementation information. When the drawing creation support device 100 receives the third implementation information, it goes through step S322 and displays the third implementation information, "Please ensure there is at least 400 mm of space on the left wall of the toilet to add barrier-free specifications."Then, based on the third implementation information, the designer changes the space on the left wall of the toilet to 400 mm.

[0180] In this case, step S304 corresponds to the first request input means or the first implementation information acquisition means of invention 7, and step S324 corresponds to the second request input means or the second implementation information acquisition means of invention 7. Also, step S324 corresponds to the first request input means or the first implementation information acquisition means of invention 8, and step S326 corresponds to the second request input means or the second implementation information acquisition means of invention 8.

[0181] Furthermore, in the first to fourth embodiments and their modifications, a trained model or a large-scale language model is used, but this is not limited to this, and the following inventions B1 to B8 can be applied to the drawing creation support device 100.

[0182] [Invention B1] A first inquiry information acquisition means acquires inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first search means searches a first storage means that stores first implementation information in association with the inquiry information inquiring about first implementation information for creating the first element or editing the second element, for the first implementation information corresponding to the inquiry information acquired by the first inquiry information acquisition means; and a second inquiry information acquisition means that acquires inquiry information inquiring about second implementation information for creating the first element or editing the second element based on a result of creating the first element or editing the second element searched for by the first search means, and implementation result information related to the result. and a second search means for searching for the second implementation information corresponding to the inquiry information and implementation result information acquired by the second inquiry information acquisition means from a second storage means that stores the second implementation information in association with inquiry information inquiring about second implementation information for creating the first element or editing the second element based on the results of creating or editing the first implementation information and implementation result information related to the results.

[0183] This makes it possible to obtain first implementation information and second implementation information in response to an inquiry, and to obtain second implementation information for creating or editing the results of creation or editing based on the first implementation information, thereby making it possible to understand the order in which elements are created or edited.

[0184] [Invention B2] An implementation result information acquisition means for acquiring implementation result information relating to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first search means for searching, in association with the implementation result information relating to the result of creating the first element or editing the second element, for first implementation information for creating the first element or editing the second element in response to the result, from a first storage means for storing first implementation information for creating the first element or editing the second element in response to the result; an inquiry information acquisition means for acquiring inquiry information inquiring about second implementation information for creating the first element or editing the second element in response to the result of creating the first element or editing the second element based on the first implementation information searched by the first search means, and implementation result information relating to the result; and a second search means for searching for the second implementation information corresponding to the inquiry information and implementation result information acquired by the inquiry information acquisition means from a second storage means that stores the second implementation information in association with inquiry information inquiring about second implementation information for creating the first element or editing the second element based on the results of creating or editing the first implementation information and implementation result information related to the results.

[0185] This allows second implementation information to be obtained in response to an inquiry, and second implementation information for creating or editing the results of creation or editing based on the first implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0186] The following describes embodiments B1 and B2 as variations of the first embodiment. The storage device 42 stores implementation information tables 1 to 3. Implementation information table 1 stores inquiry information inquiring about first implementation information for editing edit elements, and the first implementation information in association with the inquiry information. Implementation information table 2 stores inquiry information inquiring about second implementation information for editing edit elements based on the results of editing related to the first implementation information, and implementation result information related to the results, and the second implementation information in association with the inquiry information. Implementation information table 3 stores inquiry information inquiring about third implementation information for editing edit elements based on the results of editing related to the second implementation information, and implementation result information related to the results, and the third implementation information in association with the inquiry information. The processes of steps S204, S212, and S220 are replaced by processes for acquiring implementation information from implementation information tables 1 to 3.

[0187] 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 mm." Steps S200 to S206 are then carried out to obtain inquiry information indicating the inquiry, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is obtained and displayed from Implementation Information Table 1. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information, and makes an inquiry such as "Please give me instructions regarding the toilet design." Steps S208 to S214 are carried out to obtain inquiry information indicating the inquiry and implementation result information indicating the editing results, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is obtained and displayed from Implementation Information Table 2. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information and makes an inquiry saying, "Please give me instructions on the toilet design." Steps S216 to S222 are then carried out to obtain inquiry information indicating the inquiry and implementation result information indicating the edited results, and the third implementation information, "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications," is obtained and displayed from the implementation information table 3. The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0188] [Invention B3] The system includes an inquiry information acquisition means for acquiring inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first search means for searching, from a first storage means that stores first implementation information in association with the inquiry information inquiring about first implementation information for creating the first element or editing the second element, for the first implementation information corresponding to the inquiry information acquired by the inquiry information acquisition means; an implementation result information acquisition means for acquiring implementation result information regarding a result of creating the first element or editing the second element based on the first implementation information searched by the first search means; and a second search means for searching, from a second storage means that stores second implementation information in association with the implementation result information regarding a result of creating or editing related to the first implementation information, for the second implementation information corresponding to the implementation result information acquired by the implementation result information acquisition means.

[0189] This allows first implementation information to be obtained in response to an inquiry, and second implementation information to be obtained for creating or editing the results of creation or editing based on the first implementation information, so that the order in which elements are created or edited can be understood.

[0190] [Invention B4] The system includes a first implementation result information acquisition means for acquiring implementation result information relating to a result of creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first search means for searching, in association with the implementation result information relating to the result of creating the first element or editing the second element, first implementation information corresponding to the implementation result information acquired by the first implementation result information acquisition means from a first storage means that stores first implementation information for creating the first element or editing the second element based on the result; a second implementation result information acquisition means for acquiring implementation result information relating to the result of creating or editing the first element or editing the second element based on the first implementation information searched by the first search means; and a second search means for searching, in association with the implementation result information relating to the result of creating or editing the first implementation information, second implementation information for creating the first element or editing the second element based on the result.

[0191] This makes it possible to obtain second implementation information for creating or editing the results of creation or editing based on the first implementation information, thereby making it possible to understand the order in which elements are created or edited.

[0192] As a modification of the second embodiment, embodiments B3 and B4 will be described. The storage device 42 stores implementation information tables 1 to 3. Implementation information table 1 stores inquiry information inquiring about first implementation information for editing edit elements, and the first implementation information in association with the inquiry information. Implementation information table 2 stores implementation result information relating to the results of editing related to the first implementation information, and second implementation information for editing edit elements in association with the results. Implementation information table 3 stores implementation result information relating to the results of editing related to the second implementation information, and third implementation information for editing edit elements in association with the results. The processes of steps S204, S224, and S226 are replaced by processes for acquiring implementation information from implementation information tables 1 to 3.

[0193] 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 mm." Steps S200 to S206 are then carried out to obtain inquiry information indicating the inquiry, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is obtained and displayed from Implementation Information Table 1. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information, and steps S210, S224, and S214 are carried out to obtain implementation result information indicating the editing results, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is obtained and displayed from Implementation Information Table 2. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information, and through steps S218, S226, and S222, implementation result information showing the editing results is obtained, and the third implementation information "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications" is obtained and displayed from implementation information table 3. Then, the designer changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0194] [Invention B5] The system comprises a first inquiry information acquisition means for acquiring inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information; a first search means for searching, from a storage means that stores inquiry information inquiring about implementation information for creating the first element or editing the second element and the implementation information in association with the design information, for implementation information corresponding to the inquiry information and design information acquired by the first inquiry information acquisition means; a second inquiry information acquisition means for acquiring, in response to a result of creating the first element or editing the second element based on the implementation information searched by the first search means, inquiry information inquiring about implementation information for creating the first element or editing the second element, and the design information including the result; and a second search means for searching, from the storage means, the implementation information corresponding to the inquiry information and design information acquired by the second inquiry information acquisition means.

[0195] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0196] [Invention B6] A system includes a design information acquisition means for acquiring design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information; a first search means for searching for implementation information corresponding to the design information acquired by the design information acquisition means from a first storage means that stores implementation information for creating the first element or editing the second element in association with the design information; a query information acquisition means for acquiring inquiry information inquiring about implementation information for creating the first element or editing the second element in response to a result of creating the first element or editing the second element based on the implementation information searched by the first search means, and the design information including the result; and a second search means for searching for the implementation information corresponding to the inquiry information and design information acquired by the query information acquisition means from a second storage means that stores inquiry information inquiring about implementation information for creating the first element or editing the second element and the implementation information in association with the design information.

[0197] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0198] The following describes embodiments B5 and B6 as variations of the third embodiment. The storage device 42 stores implementation information tables 1 to 3. Implementation information table 1 stores query information inquiring about first implementation information for editing an edit element, CAD data immediately before editing the edit element, and the first implementation information, all associated with each other. Implementation information table 2 stores query information inquiring about second implementation information for editing the edit element based on the results of editing related to the first implementation information, CAD data containing the results, and the second implementation information, all associated with each other. Implementation information table 3 stores query information inquiring about third implementation information for editing the edit element based on the results of editing related to the second implementation information, CAD data containing the results, and the third implementation information, all associated with each other. The processes of steps S304, S312, and S320 are then replaced with processes for acquiring implementation information from implementation information tables 1 to 3.

[0199] 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 mm." Steps S300 to S306 are then executed to obtain the inquiry information and the CAD data immediately before the editing, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is obtained and displayed from Implementation Information Table 1. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information and makes an inquiry such as "Please give instructions regarding the toilet design." Steps S308 to S314 are then executed to obtain the inquiry information and the CAD data including the editing results, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is obtained and displayed from Implementation Information Table 2. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information and makes an inquiry saying, "Please give me instructions regarding the toilet design." Steps S316 to S322 are then carried out to obtain CAD data including the inquiry information indicating the inquiry and the edited results, and the third implementation information, "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications," is obtained and displayed from the implementation information table 3. The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0200] [Invention B7] The system comprises an inquiry information acquisition means for acquiring inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information; a first search means for searching, from a first storage means that stores the inquiry information inquiring about implementation information for creating the first element or editing the second element and the design information in association with the inquiry information and the design information, for the implementation information corresponding to the inquiry information and design information acquired by the inquiry information acquisition means; a design information acquisition means for acquiring the design information including a result of creating the first element or editing the second element based on the implementation information searched by the first search means; and a second search means for searching, from a second storage means that stores the implementation information for creating the first element or editing the second element in association with the design information, for the implementation information corresponding to the design information acquired by the design information acquisition means.

[0201] This allows implementation information to be obtained in response to an inquiry, and implementation information for creating or editing the results of creation or editing based on the implementation information can be obtained, so the order in which elements are created or edited can be understood.

[0202] [Invention B8] A design system including: a first design information acquisition means for acquiring design information including a result of creating a first element that can be created in the design information or editing a second element that can be edited in the design information; a first search means for searching for implementation information corresponding to the design information acquired by the first design information acquisition means from a storage means that stores implementation information for creating the first element or editing the second element in association with the design information; a second design information acquisition means for acquiring the design information including a result of creating the first element or editing the second element based on the implementation information searched by the first search means; and a second search means for searching for the implementation information corresponding to the design information acquired by the second design information acquisition means from the storage means.

[0203] This makes it possible to obtain the implementation information for creating or editing the result of creating or editing based on the implementation information, and therefore makes it possible to grasp the order in which elements are created or edited.

[0204] The following describes embodiments B7 and B8 as variations of the fourth embodiment. The storage device 42 stores implementation information tables 1 to 3. Implementation information table 1 stores query information for inquiring about first implementation information for editing an edit element, CAD data immediately before editing the edit element, and the first implementation information, all associated with each other. Implementation information table 2 stores CAD data containing the results of editing related to the first implementation information, and second implementation information for editing the edit element in association with the results. Implementation information table 3 stores CAD data containing the results of editing related to the second implementation information, and third implementation information for editing the edit element in association with the results. The processes of steps S304, S324, and S326 are replaced by processes for acquiring implementation information from implementation information tables 1 to 3.

[0205] 7 in the CAD software, the designer makes an inquiry such as "I would like to install a piano with a width of 120 mm." Steps S300 to S306 are then carried out to obtain the inquiry information and the CAD data immediately prior to the editing, and the first implementation information, "Please change the width of the toilet to 870 mm or less to prevent interference with the toilet wall," is obtained and displayed from Implementation Information Table 1. Next, the designer changes the width of the toilet to 870 mm based on the first implementation information, and steps S310, S324, and S314 are carried out to obtain CAD data including the results of the editing, and the second implementation information, "Please change the width of the hand basin to 500 mm or less to maintain the convenience of the toilet," is obtained and displayed from Implementation Information Table 2. Next, the designer changes the width of the hand wash basin to 500 [mm] based on the second implementation information, and CAD data including the edited results is obtained through steps S318, S326, and S322, and the third implementation information "Please ensure there is at least 400 [mm] of space on the left wall of the toilet to add barrier-free specifications" is obtained and displayed from implementation information table 3. The designer then changes the space on the left wall of the toilet to 400 [mm] based on the third implementation information.

[0206] Furthermore, in inventions B1, B2 and their variations, the implementation information table 2 employs a table in which inquiry information, implementation result information and implementation information are registered in correspondence with each other, but this is not limited to this, and it is also possible to employ a table in which implementation result information and implementation information are registered in correspondence with each other.

[0207] In addition, in the inventions B1 to B4 and their modifications, the working information tables 2 and 3 can be configured as one working information table.

[0208] In addition, in inventions B1 to B4 and their modifications, implementation information tables 1 to 3 can be configured as one or more implementation information tables in which inquiry information, implementation result information, and implementation information are registered in association with each other.

[0209] Furthermore, in inventions B5, B6 and their variations, the implementation information table 2 is one in which inquiry information, CAD data and implementation information are registered in association with each other, but this is not limited to this, and it is also possible to use one in which CAD data and implementation information are registered in association with each other.

[0210] In addition, in the inventions B5 to B8 and their modifications, the working information tables 2 and 3 can be configured as one working information table.

[0211] In addition, in inventions B5 to B8 and their modifications, the implementation information tables 1 to 3 can be configured as one or more implementation information tables in which inquiry information, CAD data, and implementation information are registered in association with each other.

[0212] Furthermore, in inventions B1 to B8 and their variations, the storage means stores the implementation information by any means and at any time, and may store the implementation information in advance, or may store the implementation information by external input or the like when the drawing creation support device 100 is operating without storing the implementation information in advance.

[0213] Furthermore, in the first to fourth embodiments and their modifications, a configuration including any one of an estimation process using a trained model, an acquisition process from a large-scale language model, and a search process using a table is employed. However, this is not limited to this, and a configuration including multiple of these processes can be employed. In this case, it is possible to control which process is prioritized. For example, it is possible to employ (1) a configuration in which, among multiple processes, a process with a low current load is prioritized; (2) a configuration in which, among multiple processes, a process whose results have been adopted by designers to a high degree (referring to the number of times, percentage, or other degree of adoption) is prioritized; or (3) a configuration in which, among multiple processes, a process whose results have been used by designers to a high degree (referring to the number of times, percentage, or other degree of adoption) is prioritized.

[0214] Furthermore, in the first to fourth embodiments and their variations, the data structures of Figures 2 and 9 are used as learning data, but this is not limited to this, and information regarding other attributes related to the first and second elements (e.g., ID, specifications, coordinates in the drawing, position in the building) can also be included.

[0215] Furthermore, in the above first to fourth embodiments and their variations, implementation information for editing a third element is estimated in relation to a third element whose change requires human judgment and a fourth element that is affected or affects the change to the third element. However, this is not limited to this, and it is also possible to estimate implementation information for editing an edit element whose change does not require human judgment, an edit element that is not constrained by other edit elements, or any other edit element, or for creating or editing a virtual element.

[0216] In the first to fourth embodiments and their modifications, the device is implemented as a single device, but the present invention is not limited to this and can also be implemented as a network system. As an example of a network system, some or all of the functions of the drawing creation support device 100 can be configured as a virtual server on a server that provides cloud computing services.

[0217] Furthermore, in the above first to fourth embodiments and their variations, the drawing creation support device 100 is configured to use the storage device 42, but this is not limited to this, and it can also be configured to use an external storage device such as a database server.

[0218] Furthermore, in the above-described first to fourth embodiments and their variations, when executing the processes shown in the flowcharts of Figures 4, 6, 8, 10 and 11, a program pre-stored in ROM 32 is executed, but this is not limiting, and the program showing these procedures may be read into RAM 34 from a storage medium on which the program is stored and executed.

[0219] Here, storage media refers to semiconductor storage media such as RAM and ROM, magnetic storage storage media such as FD and HD, optically readable storage media such as CD, CDV, LD and DVD, and magnetic storage / optically readable storage media such as MO, and includes any storage media that can be read by a computer, regardless of the reading method, whether electronic, magnetic, optical, etc.

[0220] Furthermore, the first to fourth embodiments and their modifications can be applied to each other. Furthermore, the present invention is not limited to the first to fourth embodiments and their modifications, and can be applied to other cases without departing from the spirit of the present invention. For example, the present invention can be applied to a wide range of design cases, such as automobile design, machine design, and circuit design.

[0221] 100...Drawing creation support device, 30...CPU, 32...ROM, 34...RAM, 38...I / F, 39...bus, 40...input device, 42...storage device, 44...display device, 400...inquiry information, 402...implementation result information, 404...implementation information, 406...CAD data

Claims

1. A design support system comprising: a request input means for inputting a request including query information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and implementation result information related to the result, into a large-scale language model; and an implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the request.

2. A design support system comprising: a request input means for inputting a request to a large-scale language model, the request including implementation result information relating to the results of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and a request for generating implementation information for creating the first element or editing the second element in response to the results; and an implementation information acquisition means for acquiring the implementation information output from the large-scale language model in response to the request.

3. A design support system comprising: a request input means for inputting a request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information including the result, into a large-scale language model; and an implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the request.

4. A design support system comprising: a request input means for inputting a request to a large-scale language model, the request including a request to generate implementation information for creating the first element or editing the second element in the design information, the request including design information including the results of creating a first element that can be created in the design information or editing a second element that can be edited in the design information; and an implementation information acquisition means for acquiring the implementation information output from the large-scale language model in response to the request.

5. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element in response to the result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information related to the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

6. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including implementation result information regarding the result of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the result; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including inquiry information inquiring about second implementation information for creating the first element or editing the second element based on the result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and implementation result information regarding the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

7. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including inquiry information inquiring about first implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including implementation result information regarding the result of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

8. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including implementation result information relating to the results of creating a first element that can be created in design information or editing a second element that can be edited in the design information, and including a request to generate first implementation information for creating the first element or editing the second element based on the results; a first implementation information acquisition means for acquiring first implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including implementation result information relating to the results of creating the first element or editing the second element based on the first implementation information acquired by the first implementation information acquisition means, and including a request to generate second implementation information for creating the first element or editing the second element based on the results; and a second implementation information acquisition means for acquiring second implementation information output from the large-scale language model in response to the second request.

9. A design support system comprising: a first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information, and the design information into a large-scale language model; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request including inquiry information inquiring about implementation information for creating the first element or editing the second element in response to the result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the result; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

10. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including design information including the results of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, and including a request to generate implementation information for creating the first element or editing the second element based on the results; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including query information inquiring about implementation information for creating the first element or editing the second element based on the results of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and the design information including the results; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

11. A design support system comprising: a first request input means for inputting a first request including inquiry information inquiring about implementation information for creating a first element that can be created in design information or editing a second element that can be edited in the design information and the design information into a large-scale language model; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request into the large-scale language model, the second request including the design information including the result of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and including a request to generate implementation information for creating the first element or editing the second element based on the result; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

12. A design support system comprising: a first request input means for inputting a first request to a large-scale language model, the first request including design information containing the results of creating a first element that can be created in the design information or editing a second element that can be edited in the design information, and including a request to generate implementation information for creating the first element or editing the second element based on the results; a first implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the first request; a second request input means for inputting a second request to the large-scale language model, the second request including the design information containing the results of creating the first element or editing the second element based on the implementation information acquired by the first implementation information acquisition means, and including a request to generate implementation information for creating the first element or editing the second element based on the results; and a second implementation information acquisition means for acquiring implementation information output from the large-scale language model in response to the second request.

13. A design support system according to any one of claims 1 to 12, wherein the first element is an element that requires human judgment to set, and the setting affects other editable elements in the design information, and the second element is an element that requires human judgment to change, and the change affects other editable elements in the design information.

14. A design support system according to any one of claims 1 to 12, wherein the design information is design information for designing a building.

Images