Architectural drawing creation support system

Abstract

20 26 20 48 55 23 J un 2 02 6 A B S T R A C T 2 0 2 6 2 0 4 8 5 5 2 3 J u n 2 0 2 6

Description

RELATED APPLICATIONS

[0001] The present application is a divisional application from Australian Patent Application No. 2023305021 filed on 4 September 2023, the entire disclosure of which is incorporated herein by this reference. This application also claims priority of JP Application No. 2023110781, filed July 05, 2023, now allowed as JP7341581B1, which is incorporated herein by reference in their entirety. TECHNICAL FIELD [0001a] The present invention relates to a system for supporting creation of an architectural drawing, and more particularly to an architectural drawing creation support system suitable for adapting an editable editing element in an architectural drawing to the specifications of a finish schedule. BACKGROUND ART

[0002] Conventionally, as a technique of supporting creation of an architectural drawing using artificial intelligence (AI), for example, a technique described in Patent Literature 1 has been known.

[0003] The technique described in Patent Literature 1 includes a table storage unit that extracts a legal determination element related to a numerical value from laws and regulations and stores a legal determination table including the legal determination element for each of the 2026204855   23 Jun 2026 laws and regulations, a substitution element extraction unit that extracts a substitution element including a numerical value substitutable for the legal determination element from initial floor plan data using AI, an appropriateness determination unit that substitutes the substitution element extracted by the substitution element extraction unit into the law determination table stored in the table storage unit, and determines that the substitution element is appropriate in a case where the condition of the law determination table is satisfied and determines that the substitution element is inappropriate in a case where the condition is not satisfied, and a determination result creation unit that records a determination result obtained by the appropriateness determination unit and creates a floor plan sheet including the determination result. [0003a] A reference herein to a patent document or any other matter identified as prior art, is not to be taken as an admission that the document or other matter was known or that the information it contains was part of the common general knowledge as at the priority date of any of the claims. CITATION LIST PATENT LITERATURE

[0004] Patent Literature 1: JP-A-2022-100026 SUMMARY OF INVENTION PROBLEMS TO BE SOLVED BY INVENTION

[0005] In building design, a sectional detail and a detail plan are created based on a finish schedule, but dimensions and other specifications described in the sectional detail must match 2026204855   23 Jun 2026 those described in the finish schedule. In addition, tags described in the detail plan must match tags described in the finish schedule. Since a designer performs these processes while confirming the finish schedule, an error is likely to occur.

[0006] However, in the technique described in Patent Literature 1, it is possible to determine whether or not an editable editing element in an architectural drawing conforms to construction-related laws and regulations, but it is not possible to adapt the editing element to the specifications of the finish schedule.

[0007] Thus, the present invention has been made in view of such unsolved problems of the prior art, and an aspect of the present invention is to provide an architectural drawing creation support system suitable for adapting an editable editing element in an architectural drawing to the specifications of a finish schedule. SOLUTION TO PROBLEMS [0007a] An aspect of the invention provides an architectural drawing creation support system for generating a large language model for an architectural drawing by learning based on information regarding finish element described in the finish schedule and information regarding matching element matching specifications of the finish element and described in the architectural drawing and / or other architectural drawings, comprising request input means that inputs, to the large language model, a request including finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing and including a request for generating matching element information regarding a matching element matching specifications of the finish element and described or to be described in the architectural 2026204855   23 Jun 2026 drawing; and element information acquisition means that acquires the matching element information output from the large language model in response to the request, wherein the large language model is generated by a data set of the finish schedule being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the prelearning parameter; wherein the large language model includes a learned parameter updated from the pre-learning parameter by learning and an inference program; and wherein the inference program receives the input finish schedule data and editing element, and based on the learned parameter, estimates the matching element; a storage device configured to store the matching element information on the large language model. [0007b] Another aspect of invention provides an architectural drawing creation support system for updating a large language model, comprising request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element, and element information acquisition means that acquires the relevant element information output from the large language model in response to the request, wherein the large language model is generated by a learning data set being input to a learning program, the learning program includes a prelearning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter; wherein the large language includes a learned parameter updated from the pre-learning parameter by learning and an inference program; and wherein the inference program receives the input of the changed editing element among the editable editing elements in the architectural drawing, and based on the learned parameter, estimates the other editing elements to be changed with the change of 2026204855   23 Jun 2026 the editing element; a storage device configured to store the relevant element information on the large language model. [0007c] A further aspect of the invention provides an architectural drawing creation support system for updating a large language model, comprising request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating determination necessity information regarding whether to require person’s determination for a change in other editing elements to be changed with a change of the editing element; and determination necessity information acquisition means that acquires the determination necessity information output from the large language model in response to the request, wherein the large language model is generated by a learning data set being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter; wherein the large language model includes a learned parameter updated from the pre-learning parameter by learning and an inference program; and wherein the inference program receives the input of the changed editing element among the editable editing elements in the architectural drawing, estimates the determination necessity information from the input editing element based on the learned parameter; a storage device configured to store the determination necessity information on the large language model. [0007d] A yet further aspect of the invention provides an architectural drawing creation support system for updating a large language model, comprising request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural 2026204855   23 Jun 2026 drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element and contents of the other editing elements and determination necessity information regarding whether to require person’s determination for a change in the other editing elements; and element information acquisition means that acquires the relevant element information and the determination necessity information output from the large language learned model in response to the request, wherein the large language is generated by a learning data set being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the prelearning parameter; wherein the large language model includes a learned parameter updated from the pre-learning parameter by learning and an inference program; and wherein the inference program receives the input of the changed editing element among the editable editing elements in the architectural drawing, estimates other editing elements to be changed with the change of the editing element and the determination necessity information from the input editing element based on the learned parameter; a storage device configured to store the change element information and the determination necessity information on the large language model.

[0008] [First Example of Invention] An architectural drawing creation support system of a first example of the invention includes request input means that inputs, to a large language model, a request including finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing and including a request for generating matching element information regarding a matching element matching specifications of the finish element and described or to 2026204855   23 Jun 2026 be described in the architectural drawing, and element information acquisition means that acquires the matching element information output from the large language model in response to the request.

[0009] Here, the finish element information may be configured as, for example, information (e.g., link information such as a name, a number, an ID, a code, and a URL) for identifying the finish element. Further, the finish element information may be configured as, for example, a character, a number, a figure, a code, a symbol, an image, sound, and other types of information. In addition, the finish element information may be configured as a keyword (e.g., one or more keywords indicating part of the name of the finish element) related to the finish element. Hereinafter, the same also applies to other types of element information.

[0010] The present system may be implemented as a single device, a terminal, or other devices, or may be implemented as a network system in which a plurality of devices, terminals, and other devices are communicably connected. In the latter case, each component may belong to, e.g., any of a plurality of devices as long as the components are communicably connected. Hereinafter, the same also applies to architectural drawing creation support systems of fourth, seventh, ninth, twelfth, fifteenth, eighteenth, and twentieth examples of the invention.

[0011] [Second Example of Invention] An architectural drawing creation support system according to a second example of the invention is the architectural drawing creation support system according to the first example of the invention, which further includes output means that outputs information regarding the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means. 2026204855   23 Jun 2026 [Third Example of Invention] An architectural drawing creation support system according to a third example of the invention is the architectural drawing creation support system according to any one of the first and second examples of the invention, in which the request includes the finish element information regarding the finish element described or to be described in the finish schedule and the editing element information regarding the editable editing element in the architectural drawing, and includes a request for generating matching element information regarding the matching element matching the specifications of the finish element and described or to be described in the architectural drawing and the contents of the matching element, and the architectural drawing creation support system further includes change means that changes the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means.

[0013] [Fourth Example of Invention] An architectural drawing creation support system of a fourth example of the invention includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element, and element information acquisition means that acquires the relevant element information output from the large language model in response to the request.

[0014] [Fifth Example of Invention] An architectural drawing creation support system according to a fifth example of the invention is the architectural drawing creation support system according to the fourth example 2026204855   23 Jun 2026 of the invention, which further includes output means that outputs information regarding the other editing elements related to the relevant element information acquired by the element information acquisition means based on the relevant element information.

[0015] [Sixth Example of Invention] An architectural drawing creation support system according to a sixth example of the invention is the architectural drawing creation support system according to any one of the fourth and fifth examples of the invention, in which the request includes the change element information regarding the editing element having been changed among the editable editing elements in the architectural drawing, and includes a request for generating the relevant element information regarding the other editing elements to be changed with the change of the editing element and contents of the other editing elements, and the architectural drawing creation support system further includes change means that changes the other editing elements related to the relevant element information acquired by the element information acquisition means based on the relevant element information.

[0016] [Seventh Example of Invention] An architectural drawing creation support system according to a seventh example of the invention includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating determination necessity information regarding whether to require person’s determination for a change in other editing elements to be changed with a change of the editing element, and determination necessity information acquisition means that acquires the determination necessity information output from the large language model in response to the request. 2026204855   23 Jun 2026 [Eighth Example of Invention] An architectural drawing creation support system according to an eighth example of the invention is the architectural drawing creation support system according to the seventh example of the invention, which further includes output means that outputs information regarding whether to require person’s determination for the change in the other editing elements related to the determination necessity information acquired by the determination necessity information acquisition means based on the determination necessity information.

[0018] [Ninth Example of Invention] An architectural drawing creation support system according to a ninth example of the invention includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element and the contents of the other editing elements and determination necessity information regarding whether to require person’s determination for a change in the other editing elements, and element information acquisition means that acquires the relevant element information and the determination necessity information output from the large language model in response to the request.

[0019] [Tenth Example of Invention] An architectural drawing creation support system according to a tenth example of the invention is the architectural drawing creation support system according to the ninth example of the invention, which further includes output means that outputs information regarding the other editing elements based on the relevant element information acquired by the element 2026204855   23 Jun 2026 information acquisition means in a case where it is determined that person's determination is required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means.

[0020] [Eleventh Example of Invention] An architectural drawing creation support system according to an eleventh example of the invention is the architectural drawing creation support system according to any one of the ninth and tenth examples of the invention, which further includes change means that changes the other editing elements based on the relevant element information acquired by the element information acquisition means in a case where it is determined that person's determination is not required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means.

[0021] [Twelfth Example of Invention] An architectural drawing creation support system according to a twelfth example of the invention includes element information acquisition means that acquires finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing, and estimation means that estimates matching element information regarding a matching element matching specifications of a finish element described in the finish schedule and described in the architectural drawing from the finish element information and the editing element information acquired by the element information acquisition means using a learned model obtained by learning based on information regarding the finish element described in the finish schedule and the matching element information. 2026204855   23 Jun 2026 For example, the element information acquisition means may receive the input of the element information from an input device, may capture or receive the element information from an external terminal, may read the element information from a storage device or a storage medium, or may generate or calculate the element information by information processing. Thus, acquisition includes at least inputting, capturing, receiving, reading (including searching), generating, and calculating. Hereinafter, the same also applies to the concept of acquisition.

[0023] [Thirteenth Example of Invention] An architectural drawing creation support system according to a thirteenth example of the invention is the architectural drawing creation support system according to the twelfth example of the invention, which further includes output means that outputs information regarding the editing element related to the editing element information acquired by the element information acquisition means based on the matching element information estimated by the estimation means.

[0024] [Fourteenth Example of Invention] An architectural drawing creation support system according to a fourteenth example of the invention is the architectural drawing creation support system according to any one of the twelfth and thirteenth examples of the invention, in which the learned model is obtained by learning based on the information regarding the finish element described in the finish schedule and matching element information regarding the matching element matching the specifications of the finish element and described in the architectural drawing and contents of the matching element, and the architectural drawing creation support system further includes change means that changes the editing element related to the editing element information acquired by the 2026204855   23 Jun 2026 element information acquisition means based on the matching element information estimated by the estimation means.

[0025] [Fifteenth Example of Invention] An architectural drawing creation support system according to a fifteenth example of the invention includes element information acquisition means that acquires change element information regarding an editing element having been changed among editable editing elements in an architectural drawing, and estimation means that estimates relevant element information regarding other editing elements to be changed or changed with a change of an editable editing element in the architectural drawing from the change element information acquired by the element information acquisition means using a learned model obtained by learning based on information regarding the editable editing element in the architectural drawing and the relevant element information.

[0026] [Sixteenth Example of Invention] An architectural drawing creation support system according to a sixteenth example of the invention is the architectural drawing creation support system according to the fifteenth example of the invention, which further includes output means that outputs information regarding the other editing elements related to the relevant element information estimated by the estimation means based on the relevant element information.

[0027] [Seventeenth Example of Invention] An architectural drawing creation support system according to a seventeenth example of the invention is the architectural drawing creation support system according to any one of the fifteenth and sixteenth examples of the invention, in which the learned model is obtained by learning based on the information regarding the editable editing element in the architectural 2026204855   23 Jun 2026 drawing and relevant element information regarding the other editing elements to be changed or changed with the change of the editing element and the contents of the other editing elements, and the architectural drawing creation support system further includes change means that changes the other editing elements related to the relevant element information estimated by the estimation means based on the relevant element information.

[0028] [Eighteenth Example of Invention] An architectural drawing creation support system according to an eighteenth example of the invention includes element information acquisition means that acquires change element information regarding an editing element having been changed among editable editing elements in an architectural drawing, and estimation means that estimates determination necessity information regarding whether to require person’s determination for a change in other editing elements to be changed or changed with a change of an editable editing element in the architectural drawing from the change element information acquired by the element information acquisition means using a learned model obtained by learning based on information regarding the editable editing element in the architectural drawing and the determination necessity information.

[0029] [Nineteenth Example of Invention] An architectural drawing creation support system according to a nineteenth example of the invention is the architectural drawing creation support system according to the eighteenth example of the invention, which further includes output means that outputs information regarding whether to require person’s determination for the change in the other editing elements related to the determination necessity information estimated by the estimation means based on the determination necessity information.

[0030] 2026204855   23 Jun 2026 [Twentieth Example of Invention] An architectural drawing creation support system according to a twentieth example of the invention includes element information acquisition means that acquires change element information regarding an editing element having been changed among editable editing elements in an architectural drawing; and estimation means that estimates relevant element information regarding other editing elements to be changed or changed with a change of an editable editing element in the architectural drawing and the contents of the other editing elements and determination necessity information regarding whether to require person’s determination for a change in the other editing elements from the change element information acquired by the element information acquisition means using a learned model obtained by learning based on information regarding the editable editing element in the architectural drawing, the relevant element information, and the determination necessity information.

[0031] [Twenty-First Example of Invention] An architectural drawing creation support system according to a twenty-first example of the invention is the architectural drawing creation support system according to the twentieth example of the invention, which further includes output means that outputs information regarding the other editing elements based on the relevant element information estimated by the estimation means in a case where it is determined that person's determination is required to change the other editing elements related to the relevant element information based on the determination necessity information estimated by the estimation means.

[0032] [Twenty-Second Example of Invention] An architectural drawing creation support system according to a twenty-second example of the invention is the architectural drawing creation support system according to any one of the twentieth and twenty-first examples of the invention, which further includes change 2026204855   23 Jun 2026 means that changes the other editing elements related to the relevant element information based on the relevant element information estimated by the estimation means in a case where it is determined that person's determination is not required to change the other editing elements based on the determination necessity information estimated by the estimation means. EFFECTS OF INVENTION

[0033] As described above, according to the architectural drawing creation support system of the first example of the invention, the matching element information regarding the matching element corresponding to the finish element described in the finish schedule is obtained, and therefore, it can be expected that the editing element matches the specifications of the finish schedule.

[0034] Further, according to the architectural drawing creation support system of the third example of the invention, the matching element information regarding the matching element corresponding to the finish element described in the finish schedule and the contents of the matching element is obtained, and therefore, it can be expected that the editing element and the contents of the editing element match the specifications of the finish schedule.

[0035] Further, according to the architectural drawing creation support system of the fourth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element is obtained, and therefore, the other editing elements to be changed can also be changed.

[0036] Further, according to the architectural drawing creation support system of the sixth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element and the contents of the other editing elements 2026204855   23 Jun 2026 is obtained, and therefore, the other editing elements to be changed and the contents of the other editing elements can also be changed.

[0037] Further, according to the architectural drawing creation support system of the seventh example of the invention, the determination necessity information regarding whether to require person’s determination for the change in the other editing elements related to the changed editing element is obtained, and therefore, it can be grasped whether to change the other editing elements to be changed by a person.

[0038] Further, according to the architectural drawing creation support system of the ninth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element and the contents of the other editing elements and the determination necessity information regarding whether to require person’s determination for the change in the other editing elements are obtained, and therefore, it can be grasped whether to change the other editing elements to be changed and the contents of the other editing elements by a person.

[0039] Further, according to the architectural drawing creation support system of the twelfth example of the invention, the matching element information regarding the matching element corresponding to the finish element described in the finish schedule is estimated, and therefore, it can be expected that the editing element matches the specifications of the finish schedule.

[0040] Further, according to the architectural drawing creation support system of the fourteenth example of the invention, the matching element information regarding the matching element corresponding to the finish element described in the finish schedule and the contents 2026204855   23 Jun 2026 of the matching element is estimated, and therefore, it can be expected that the editing element and the contents of the editing element match the specifications of the finish schedule.

[0041] Further, according to the architectural drawing creation support system of the fifteenth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element is estimated, and therefore, the other editing elements to be changed can also be changed.

[0042] Further, according to the architectural drawing creation support system of the seventeenth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element and the contents of the other editing elements is estimated, and therefore, the other editing elements to be changed and the contents of the other editing elements can also be changed.

[0043] Further, according to the architectural drawing creation support system of the eighteenth example of the invention, the determination necessity information regarding whether to require person’s determination for the change in the other editing elements related to the changed editing element is estimated, and therefore, it can be grasped whether to change the other editing elements to be changed by a person.

[0044] Further, according to the architectural drawing creation support system of the twentieth example of the invention, the relevant element information regarding the other editing elements related to the changed editing element and the contents of the other editing elements and the determination necessity information regarding whether to require person’s determination for the change in the other editing elements is estimated, and therefore, it can be grasped whether to change the other editing elements to be changed and the contents of the 2026204855   23 Jun 2026 other editing elements by a person. BRIEF DESCRIPTION OF DRAWINGS

[0045] Fig. 1 is a diagram illustrating a hardware configuration of a drawing creation support device 100; Fig. 2 is a table illustrating the structure of finish schedule data; Fig. 3 is a view illustrating the structure of CAD data on a sectional detail; Fig. 4 is a view illustrating the structure of CAD data on a detail plan; Fig. 5 is a wall type table; Fig. 6 is a flowchart illustrating learned model generation processing; Fig. 7 is a block diagram illustrating a process of generating and using a learned model; Fig. 8 is a flowchart illustrating matching element estimation processing; Fig. 9 is a sectional detail before an editing element is changed; Fig. 10 is a sectional detail after the editing element has been changed; Fig. 11 is a detail plan before the editing element is changed; Fig. 12 is a detail plan after the editing element has been changed; Fig. 13 is a block diagram illustrating a process of generating and using a learned model; Fig. 14 is a flowchart illustrating relevant element estimation processing; Fig. 15 is a block diagram illustrating a process of generating and using a learned model; and Fig. 16 is a flowchart illustrating relevant element estimation processing. DESCRIPTION OF EMBODIMENTS 2026204855   23 Jun 2026 [First Embodiment] Hereinafter, a first embodiment of the present invention will be described. Figs. 1 to 12 are figures illustrating the present embodiment.

[0047] First, the configuration of the present embodiment will be described. Fig. 1 is a diagram illustrating a hardware configuration of a drawing creation support device 100.

[0048] As illustrated in Fig. 1, the drawing creation support device 100 includes a central processing unit (CPU) 30 that controls calculation and the entire system based on a control program, a read only memory (ROM) 32 that saves, e.g., the control program of the CPU 30 in advance in a predetermined area, a random access memory (RAM) 34 that saves data read from, e.g., the ROM 32 and a calculation result required in a calculation process by the CPU 30, and an interface (I / F) 38 that mediates input and output of data to and from an external device, and these components are connected to each other via a bus 39, which is a signal line for transferring data, so as to transmit and receive data to and from each other.

[0049] An input device 40 including, e.g., a keyboard and a mouse capable of inputting data as human interfaces, a storage device 42 that stores, e.g., data and a table as a file, and a display device 44 that displays a screen based on an image signal are connected to the I / F 38 as external devices.

[0050] Computer aided design (CAD) software or building information modeling (BIM) software (hereinafter, these types of software will be collectively referred to as “CAD software”) is installed in the storage device 42. The CAD software is software that supports 2026204855   23 Jun 2026 creation of a drawing according to user operation. When activation of the CAD software is requested, the CPU 30 activates a CAD program stored in a predetermined area of the ROM 32, and executes processing according to the program. A designer can activate the CAD software to create a sectional detail, a detail plan, and other architectural drawings.

[0051] Next, a data structure in the storage device 42 will be described. Fig. 2 is a table illustrating the structure of finish schedule data.

[0052] As illustrated in Fig. 2, the storage device 42 stores the finish schedule data regarding a finish schedule. The finish schedule is a table summarizing the finish of each location of a building, and includes an external finish schedule showing the finish of a roof, an outer wall, etc. and an internal finish schedule showing the finish of a wall, a floor, a ceiling, etc. in each room. In the finish schedule data, one line is registered for each finish element. In the third line in Fig. 2, “TILE CARPET (A), FINISH WITH STRIPE PATTERN” as floor finish, “VINYL WALLPAPER (A)” as wall finish, and “VINYL WALLPAPER (A)” as ceiling finish are registered for a finish element “INNER CORRIDOR.” This indicates that the floor of the inner corridor is finished with “TILE CARPET (A), FINISH WITH STRIPE PATTERN,” and the walls and ceiling of the inner corridor are finished with “VINYL WALLPAPER (A).” In addition, in the ninth line in Fig. 2, “2450” is registered as a ceiling height for a finish element “AIRLOCK ROOM.” This indicates that the ceiling height of the airlock room is finished with 2450 [mm].

[0053] Fig. 3 is a view illustrating the structure of CAD data on a sectional detail. As illustrated in Fig. 3, the storage device 42 stores CAD data and BIM data (hereinafter, these types of data will be collectively referred to as “CAD data”) on the sectional 2026204855   23 Jun 2026 detail.

[0054] The sectional detail is a drawing illustrating a cross section of a building in detail. The CAD data on the sectional detail is created using the CAD software by the designer. In the CAD software, the designer creates the sectional detail by adding, editing, or deleting an editable editing element on the sectional detail. In the example of Fig. 3, the editing elements of a floor, a wall, and a ceiling in a region displayed as “INNER CORRIDOR” are arranged respectively, and the editing elements of a floor, a wall, and a ceiling in a region displayed as “AIRLOCK ROOM” are arranged respectively. These elements correspond to the finish elements in the third and ninth lines in the finish schedule data of Fig. 2. The finish element and the editing element corresponding to each other may be associated with each other by setting the same ID in the finish schedule data and the CAD data.

[0055] Fig. 4 is a view illustrating the structure of CAD data on a detail plan. Fig. 5 is a wall type table.

[0056] As illustrated in Fig. 4, the storage device 42 stores the CAD data on the detail plan. The detail plan is a drawing illustrating the plane of a building in detail. The CAD data on the detail plan is created using the CAD software by the designer. In the CAD software, the designer creates the detail plan by adding, editing, or deleting an editable editing element on the detail plan. In the detail plan, a tag is arranged for an editing element for a wall so that a wall type can be identified. The tag is three or four digits of alphanumeric characters surrounded by a rectangle or an ellipse, and the designer sets a value with reference to the wall type table of Fig. 5. In the example of Fig. 4, a tag “L4DA” is arranged in a region displayed as “LD,” and a tag “L4JA” is arranged in a region displayed as “BATHROOM.” These tags correspond to the finish elements in the thirteenth and seventeenth 2026204855   23 Jun 2026 lines in the finish schedule data of Fig. 2. The finish element and the editing element corresponding to each other may be associated with each other by setting the same ID in the finish schedule data and the CAD data.

[0057] The storage device 42 stores CAD data on other architectural drawings and data on a learned model. The learned model is obtained by learning based on information regarding the finish element described in the finish schedule and information regarding a matching element matching the specifications of the finish element and described in the architectural drawing. Since the finish schedule data and the CAD data on the sectional detail, the detail plan, and other architectural drawings are used for AI learning, the storage device 42 stores a large number of pieces of data created in the past. Such CAD data is data in which each editing element in the architectural drawing matches the specifications of the finish schedule data, and serves as teacher data in AI learning.

[0058] First, operation of the present embodiment will be described. Fig. 6 is a flowchart illustrating learned model generation processing.

[0059] Fig. 7 is a block diagram illustrating a process of generating and using the learned model. The learned model generation processing is processing executed to generate a learned model, and when executed in the CPU 30, the processing proceeds to Step S100 to execute finish schedule data analysis processing as illustrated in Fig. 6. In the finish schedule data analysis processing, the finish schedule data is read from the storage device 42, and the information on the finish element is extracted from the read finish schedule data.

[0060] Next, the processing proceeds to Step S102. In Step S102, CAD data analysis 2026204855   23 Jun 2026 processing is executed. In the CAD data analysis processing, the CAD data on the sectional detail, the detail plan, and other architectural drawings is read from the storage device 42, and the information regarding the editing element is extracted from the read CAD data.

[0061] Next, the processing proceeds to Step S104. In Step S104, as illustrated in Fig. 7, a learning data set is generated based on the information extracted in Steps S100, S102, and the processing proceeds to Step S106. The generated learning data set is input to a learning program to generate the learned model by the learning program. The learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to updates the pre-learning parameter. Then, the learned model is output as a learning result.

[0062] The learned model includes a learned parameter updated from the pre-learning parameter by learning and an inference program. The inference program receives the input finish schedule data and editing element, and based on the learned parameter, estimates a matching element (editing element matching the specifications of the finish element) corresponding to the finish element of the finish schedule data and the editing element from the input finish schedule data and editing element and outputs the estimated matching element. The input to the inference program is not limited to the finish schedule data itself, and may be made by extracting a finish element of the finish schedule data corresponding to an editing element to be input as long as a correspondence relationship can be specified by, e.g., an ID. Note that since a relationship between the input editing element and the matching element is determined by AI learning, such a relationship shows a tendency similar to the contents of the past finish schedule data and CAD data, but is ambiguous that the relationship is not precisely coincident with these contents. However, such ambiguity can be reduced by the amount of learning data and learning accuracy. 2026204855   23 Jun 2026

[0063] Fig. 8 is a flowchart illustrating matching element estimation processing. The matching element estimation processing is processing executed in response to a request from a user, and when executed in the CPU 30, the processing first proceeds to Step S150 as illustrated in Fig. 8. In Step S150, in the CAD software, it is determined whether or not any of the editable editing elements in the architectural drawing has been instructed by user operation, and in a case where it is determined that an editing element has been instructed (YES), the processing proceeds to Step S152.

[0064] In Step S152, finish schedule data corresponding to the CAD data currently being edited is read from the storage device 42, and the editing element instructed in Step S150 is acquired. The processing proceeds to Step S154, and the matching element is estimated using the learned model in the storage device 42 from the read finish schedule data and the acquired editing element. Estimation is performed by inputting the finish schedule data and the editing element to the learned model and acquiring a matching element output from the learned model.

[0065] Then, the processing proceeds to Step S156, and it is determined whether or not the editing element instructed in Step S150 and the matching element estimated in Step S154 do not match in terms of identification and contents. In a case where it is determined that these elements do not match (YES), the processing proceeds to Step S158, a message indicating that the editing element has a defect is displayed on the display device 44, and a series of processing ends.

[0066] On the other hand, in both a case where it is determined in Step S156 that the editing element and the matching element do not mismatch (NO) and a case where it is determined in 2026204855   23 Jun 2026 Step S150 that the editing element is not instructed (NO), a series of processing ends.

[0067] [Creation Example of Sectional Detail] Next, an operation of changing an editing element will be described with a case of creating a sectional detail as an example. Note that it is assumed that finish schedule data corresponding to CAD data on the sectional detail currently being edited has contents illustrated in Fig. 2.

[0068] Fig. 9 is a sectional detail before the editing element is changed. Fig. 10 is a sectional detail after the editing element has been changed.

[0069] In the CAD software, the designer instructs an editing element 400 of a floor in a region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9. At this time, while the editing element 400 is set to “VINYL FLOOR TILE (C),” the corresponding specification element (third line) in the finish schedule data of Fig. 2 is “TILE CARPET (A), FINISH WITH STRIPE PATTERN.” Thus, when the finish schedule data of Fig. 2 and the editing element 400 are input to the learned model, the matching element “TILE CARPET (A), FINISH WITH STRIPE PATTERN” is estimated, and therefore, a message 402 indicating that the editing element 400 has a defect is displayed. As illustrated in Fig. 10, the designer changes the editing element 400 to “TILE CARPET (A), FINISH WITH STRIPE PATTERN” according to the message 402.

[0070] Subsequently, in the CAD software, the designer instructs an editing element 404 of a wall in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9. At this time, while the editing element 404 is set to “VINYL WALLPAPER (B),” the corresponding specification element (third line) in the finish schedule data of Fig. 2 is 2026204855   23 Jun 2026 “VINYL WALLPAPER (A).” Thus, when the finish schedule data of Fig. 2 and the editing element 404 are input to the learned model, the matching element “VINYL WALLPAPER (A)” is estimated, and therefore, a message 406 indicating that the editing element 404 has a defect is displayed. As illustrated in Fig. 10, the designer changes the editing element 404 to “VINYL WALLPAPER (A)” according to the message 406.

[0071] Subsequently, in the CAD software, the designer instructs an editing element 408 of a ceiling in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9. At this time, while the editing element 408 is set to “VINYL WALLPAPER (B),” the corresponding specification element (third line) in the finish schedule data of Fig. 2 is “VINYL WALLPAPER (A).” Thus, when the finish schedule data of Fig. 2 and the editing element 408 are input to the learned model, the matching element “VINYL WALLPAPER (A)” is estimated, and therefore, a message 406 indicating that the editing element 408 has a defect is displayed. As illustrated in Fig. 10, the designer changes the editing element 408 to “VINYL WALLPAPER (A)” according to the message 406.

[0072] Similarly, in the CAD software, the designer instructs the ceiling height (editing element) 410 of a ceiling in a region displayed as “AIRLOCK ROOM” in the sectional detail of Fig. 9. Since the ceiling height 410 does not match the specifications of the finish schedule data of Fig. 2 either, a message 412 indicating that the ceiling height 410 has a defect is displayed. As illustrated in Fig. 10, the designer changes the ceiling height 410 to 2450 [mm] according to the message 412.

[0073] Note that In the CAD software, the designer instructs a usable dimension 414 in a region displayed as “COMMON CORRIDOR” in the sectional detail of Fig. 9. Since the usable dimension 414 fails to conform to construction-related laws and regulations, a message 2026204855   23 Jun 2026 416 indicating that the usable dimension 414 has a defect is displayed. Determination on whether to conform to the construction-related laws and regulations can be made by the technique of Patent Literature 1, for example. A method for changing the usable dimension 414 may be performed by adjusting a right wall position or adjusting a left wall position. However, it is difficult for AI to determine which one should be prioritized. Thus, such an editing element 414 is taken as an editing element requiring person’s determination for change, and is not automatically changed by AI, but the designer changes the right or left wall position depending on designer's determination.

[0074] [Creation Example of Detail Plan] Next, an operation of changing the editing element will be described with a case of creating a detail plan as an example. Note that it is assumed that finish schedule data corresponding to CAD data on the detail plan currently being edited has contents illustrated in Fig. 2.

[0075] Fig. 11 is a detail plan before an editing element is changed. Fig. 12 is a detail plan after the editing element has been changed.

[0076] In the CAD software, the designer instructs a tag (editing element) 420 in a region displayed as “BATHROOM” in the detail plan of Fig. 11. At this time, while the tag 420 is set to “L4JB,” the corresponding specification element (seventeenth line) in the finish schedule data of Fig. 2 is “L4,” “GBt12.5,” and “VINYL WALLPAPER (C)” (= “L4JA” in the wall type table of Fig. 5). Thus, when the finish schedule data of Fig. 2 and the tag 420 are input to the learned model, the matching element “L4JA” is estimated, and therefore, a message 422 indicating that the tag 420 has a defect is displayed. As illustrated in Fig. 12, the designer changes the tag 420 to “L4JA” according to the message 422. 2026204855   23 Jun 2026 Subsequently, in the CAD software, the designer instructs a tag 424 of a wall in a region displayed as “LD” in the detail plan of Fig. 11. At this time, while the tag 424 is set to “L4LA,” the corresponding specification element (thirteenth line) in the finish schedule data of Fig. 2 is “L4,” “GBt12.5*SURFACE BASE IS WATERPROOF GB,” and “VINYL WALLPAPER (C)” (= “L4DA” in the wall type table of Fig. 5). Thus, when the finish schedule data of Fig. 2 and the tag 424 are input to the learned model, the matching element “L4DA” is estimated, and therefore, a message 426 indicating that the tag 424 has a defect is displayed. As illustrated in Fig. 12, the designer changes the tag 424 to “L4DA” according to the message 426.

[0078] Next, advantageous effects of the present embodiment will be described. In the present embodiment, the finish schedule data and the editing element are acquired, the matching element is estimated from the acquired finish schedule data and editing element using the learned model obtained by learning based on the information regarding the finish element described in the finish schedule and the information regarding the matching element matching the specifications of the finish element and described in the architectural drawing, and in a case where the acquired editing element and the estimated matching element do not match, the message indicating that the editing element has a defect is displayed.

[0079] With this configuration, it can be expected that the editing element matches the specifications of the finish schedule. In the present embodiment, Step S152 corresponds to the element information acquisition means of the twelfth or thirteenth example of the invention, Step S154 corresponds to the estimation means of the twelfth or thirteenth example of the invention, and Step S158 corresponds to the output means of the thirteenth example of the invention. 2026204855   23 Jun 2026

[0080] [Second Embodiment] Next, a second embodiment of the present invention will be described. Figs. 13 and 14 are charts illustrating the present embodiment.

[0081] The present embodiment is different from the first embodiment in that other editing elements to be changed with a change of an editing element are displayed. In a case where a certain editing element is changed, other editing elements may no longer match the specifications of a finish schedule due to the change. For example, in a case where the ceiling height 410 in Fig. 9 is changed, it is necessary to change, e.g., the height of an attic space. An object of the present embodiment is to display other editing elements to be changed with a change of an editing element, thereby supporting a designer to also change the other editing elements to be changed. Hereinafter, only contents different from those of the first embodiment will be described, and description of overlapping contents will be omitted.

[0082] First, the data structure in the storage device 42 will be described. The storage device 42 stores CAD data on a sectional detail, a detail plan, and other architectural drawings, the CAD data including CAD data (hereinafter referred to as “prechanged CAD data”) before an editing element is changed and CAD data (hereinafter referred to as “changed CAD data”) after the editing element has been changed.

[0083] The storage device 42 stores data on a learned model. The learned model is obtained by learning based on information regarding an editable editing element in an architectural drawing and information regarding other editing elements changed with a change of the editing element. Since the pre-changed CAD data and the changed CAD data are used for AI learning, the storage device 42 stores a large number of pieces of data created in the 2026204855   23 Jun 2026 past.

[0084] First, operation of the present embodiment will be described. Fig. 13 is a block diagram illustrating a process of generating and using the learned model.

[0085] The learned model generation processing is processing executed to generate a learned model, and when executed in the CPU 30, the pre-changed CAD data and the changed CAD data are read from the storage device 42, and the information regarding the editing element is extracted from the read CAD data.

[0086] As illustrated in Fig. 13, a learning data set is generated based on the extracted information, and the generated learning data set is input to a learning program to generate the learned model by a learning program. An inference program of the learned model receives the input of the changed editing element among the editable editing elements in the architectural drawing, estimates other editing elements to be changed with a change of the editing element from the input editing element based on a learned parameter, and outputs the estimated editing element as a relevant element.

[0087] Fig. 14 is a flowchart illustrating relevant element estimation processing. The relevant element estimation processing is processing executed in response to a request from a user, and when executed in the CPU 30, the processing first proceeds to Step S200 as illustrated in Fig. 14. In Step S200, in the CAD software, it is determined whether or not any of the editable editing elements in the architectural drawing has been changed, and in a case where it is determined that an editing element has been changed (YES), the processing proceeds to Step S202. 2026204855   23 Jun 2026

[0088] In Step S202, the changed editing element is acquired, and the processing proceeds to Step S204 to estimate a relevant element from the acquired editing element using the learned model in the storage device 42. Estimation is performed by inputting the editing element to the learned model and acquiring the relevant element output from the learned model.

[0089] Then, the processing proceeds to Step S206, other editing elements to be changed are displayed in a specific form (e.g., highlight) based on the estimated relevant element, and a series of processing ends.

[0090] On the other hand, in a case where it is determined in Step S200 that the editing element is not changed (NO), a series of processing ends.

[0091] Next, advantageous effects of the present embodiment will be described. In the present embodiment, the editing element having been changed is acquired, the relevant element is estimated from the acquired editing element using the learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing and the information regarding other editing elements having been changed with a change of the editing element, and the other editing elements to be changed are displayed in the specific form based on the estimated relevant element.

[0092] With this configuration, the other editing elements to be changed can also be changed. In the present embodiment, Step S202 corresponds to the element information acquisition means of the fifteenth example of the invention, Step S204 corresponds to the estimation means of the fifteenth or sixteenth example of the invention, and Step S206 corresponds to the output means of the sixteenth example of the invention. 2026204855   23 Jun 2026

[0093] [Third Embodiment] Next, a third embodiment of the present invention will be described. Figs. 15 and 16 are charts illustrating the present embodiment.

[0094] The present embodiment is different from the second embodiment in that an editing element requiring person’s determination for change is estimated. In a case where a certain editing element is changed, it is desirable to automatically change other editing elements according to the change, but it is necessary to exclude an editing element requiring person’s determination for the change from the target of the automatic change. For example, in a case where the usable dimension 414 in FIG. 9 is changed, the designer determines which one of the right or left wall position is changed. An object of the present embodiment is to automatically change other editing elements to be changed with a change of an editing element and exclude an editing element requiring person’s determination for the change from the target of the automatic change, thereby supporting a person to grasp whether to change the other editing elements to be changed and the contents of the other editing elements to be changed. Hereinafter, only contents different from those of the first and second embodiments will be described, and description of overlapping contents will be omitted.

[0095] First, the data structure in the storage device 42 will be described. The storage device 42 stores determination necessity data in which an editing element is registered in association with information on whether to require person’s determination for change. The editing element requiring person’s determination for the change may be set by person’s determination. In addition, an editing element having two or more change patterns before and after the change may be extracted based on the pre-changed CAD data and the changed CAD data, and the extracted editing element may be set as an editing element 2026204855   23 Jun 2026 requiring person’s determination for the change.

[0096] The storage device 42 stores data on a learned model. The learned model is obtained by learning based on information regarding an editable editing element in an architectural drawing, information regarding other editing elements changed with the change of the editing element and the contents of the other editing elements, and determination necessity information regarding whether to require person’s determination for the change in the other editing elements. Since the pre-changed CAD data, the changed CAD data, and the determination necessity data are used for AI learning, the storage device 42 stores a large number of pieces of data created in the past.

[0097] First, operation of the present embodiment will be described. Fig. 15 is a block diagram illustrating a process of generating and using the learned model.

[0098] The learned model generation processing is processing executed to generate a learned model, and when executed in the CPU 30, the pre-changed CAD data, the changed CAD data, and the determination necessity data are read from the storage device 42, and the information regarding the editing element and the determination necessity information are extracted from the read CAD data and determination necessity data.

[0099] As illustrated in Fig. 15, a learning data set is generated based on the extracted information, and the generated learning data set is input to a learning program to generate the learned model by a learning program. An inference program of the learned model receives the input of the changed editing element among the editable editing elements in the architectural drawing, estimates other editing elements to be changed with a change of the 2026204855   23 Jun 2026 editing element and the determination necessity information from the input editing element based on a learned parameter, and outputs the estimated editing element as a relevant element.

[0100] Fig. 16 is a flowchart illustrating relevant element estimation processing. The relevant element estimation processing is processing executed in response to a request from a user, and when executed in the CPU 30, the processing first proceeds to Step S250 as illustrated in Fig. 16. In Step S250, in the CAD software, it is determined whether or not any of the editable editing elements in the architectural drawing has been changed, and in a case where it is determined that an editing element has been changed (YES), the processing proceeds to Step S252.

[0101] In Step S252, the changed editing element is acquired, and the processing proceeds to Step S254 to estimate a relevant element and determination necessity information from the acquired editing element using the learned model in the storage device 42. Estimation is performed by inputting the editing element to the learned model and acquiring the relevant element and determination necessity information output from the learned model.

[0102] Then, the processing proceeds to Step S256, and based on the estimated determination necessity information, it is determined whether or not person's determination is required to change the estimated relevant element. In a case where it is determined that person's determination is not required to change the relevant element (NO), the processing proceeds to Step S258 to change the other editing elements based on the estimated relevant element such that the contents of the other editing elements to be changed are coincident with the contents of the relevant element, and a series of processing ends.

[0103] On the other hand, in a case where it is determined in Step S256 that person’s 2026204855   23 Jun 2026 determination is required to change the relevant element (YES), the processing proceeds to Step S260, other editing elements to be changed are displayed in a specific form (e.g., highlight) based on the estimated relevant element, and a series of processing ends.

[0104] On the other hand, in a case where it is determined in Step S250 that the editing element is not changed (NO), a series of processing ends.

[0105] Next, advantageous effects of the present embodiment will be described. In the present embodiment, the editing element having been changed among the editable editing elements in the architectural drawing is acquired, the relevant element and the determination necessity information are estimated from the acquired editing element using the learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing, the information regarding the other editing elements having been changed with the change of the editing element and the contents of the other editing elements, and the determination necessity information regarding whether to require person's determination for the change in the other editing elements. The other editing elements are displayed in the specific form based on the estimated relevant element in a case where it is determined based on the estimated determination necessity information that person's determination is required to change the relevant element, and are changed based on the estimated relevant element in a case where it is determined that person's determination is not required to change the relevant element.

[0106] With this configuration, it is possible to grasp whether or not a person should change the other editing elements to be changed and the contents of the other editing elements to be changed. 2026204855   23 Jun 2026 In the present embodiment, Step S252 corresponds to the element information acquisition means of the eighteenth or twentieth example of the invention, Step S254 corresponds to the estimation means of the eighteenth to twenty-second example of the invention, Step 258 corresponds to the change means of the twenty-second example of the invention, and Step S260 corresponds to the output means of the nineteenth or twenty-first example of the invention.

[0108] [Modifications] Note that in the first embodiment, the defect is displayed for the editing element instructed by user operation, but the present invention is not limited to the configuration. A configuration may be employed, in which all the editing elements in the drawing are scanned and defects are collectively displayed for the defective editing elements.

[0109] In the first embodiment and the modifications of the first embodiment, the message indicating that the editing element has the defect is displayed, but the present invention is not limited to the configuration. A configuration may also be employed, in which the instructed editing element is changed based on the estimated matching element. In this case, as the learned model, one obtained by learning based on the information regarding the finish element described in the finish schedule and the information regarding the matching element matching the specifications of the finish element and described in the architectural drawing and the contents of the matching element is used.

[0110] In the second embodiment and the modifications of the second embodiment, the other editing elements to be changed are displayed in the specific form based on the estimated relevant element, but the present invention is not limited to the configuration. A configuration may also be employed, in which the other editing elements to be changed are 2026204855   23 Jun 2026 changed based on the estimated relevant element. In this case, as the learned model, one obtained by learning based on the information regarding the editable editing element in the architectural drawing and the information regarding other editing elements changed with the change of the editing element and the contents of the changed other editing elements is used.

[0111] In the third embodiment and the modifications of the third embodiment, the other editing elements to be changed are changed or displayed in the specific form based on the estimated relevant element, but the present invention is not limited to the configuration. In a configuration in which the change is not made, a learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing, the information regarding the other editing elements changed with the change of the editing element, and the determination necessity information regarding whether to require person’s determination for the change in the other editing elements may also be used. In a configuration in which the other editing elements are not displayed in the specific form either (e.g., configuration in which only the necessity of person’s determination is estimated), a learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing and the determination necessity information regarding whether to require person’s determination for the change in other editing elements with the change of the editing element may also be used.

[0112] In the second and third embodiments and the modifications of the second and third embodiments, the learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing and the information regarding the other editing elements changed with the change of the editing element is used, but the present invention is not limited to the configuration. A learned model obtained by learning based on the information regarding the editable editing element in the architectural drawing and the 2026204855   23 Jun 2026 information regarding the other editing elements to be changed with the change of the editing element may also be used. That is, instead of the pre-changed CAD data and the changed CAD data, correspondence information indicating that an editing element B is also changed when an editing element A is changed may be used for AI learning.

[0113] In the first to third embodiments and the modifications of the first to third embodiments, the learned model is generated and the generated learned model is used, but the present invention is not limited to the configuration. A configuration using a large language model may also be employed. Specifically, for example, the following configuration may be employed.

[0114] The drawing creation support device 100 is connected to a large language model server having a large language model via a network. The large language model server has a hardware configuration similar to that of the drawing creation support device 100. The large language model is a deep learning model for learning, in advance from an enormous amount of data, a so-called language model for modeling words spoken by a human with their probabilities of occurrence. When receiving a request, the large language model server statistically estimates the probability of generation of a next word from a sentence included in the received request using the large language model, and transmits an estimation result to a request source. As the large language model, for example, one of techniques described in Internet sites “https: / / chatgpt-lab.com / n / n418d3aa56f0b” and “https: / / agirobots.com / chatgpt-mechanism-and-problem / ” may be employed.

[0115] The following examples A1 to A3 or B1 to B8 of the invention are applicable to the drawing creation support device 100.

[0116] 2026204855   23 Jun 2026 [Example A1 of Invention] A drawing creation support device includes request input means that inputs, to a large language model, a request including finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing and including a request for generating matching element information regarding a matching element matching specifications of the finish element and described or to be described in the architectural drawing, and element information acquisition means that acquires the matching element information output from the large language model in response to the request.

[0117] [Example A2 of Invention] In the example A1 of the invention, the drawing creation support device further includes output means that outputs information regarding the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means.

[0118] An embodiment of the examples A1, A2 of the invention will be described with Fig. 9 as an example. This embodiment is a modification of the first embodiment, and the processing in Step S154 is replaced with processing of acquiring the matching element information from the large language model server.

[0119] When the designer instructs the editing element 400 of the floor in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9 in the CAD software, in the drawing creation support device 100, a request is transmitted to the large language model server. The request includes the finish schedule data of Fig. 2 and the editing element 400, and includes a request for generating a matching element of the editing element 400. In 2026204855   23 Jun 2026 response to this request, the matching element information is output from the large language model server. When the matching element information is received in the drawing creation support device 100, a message 402 indicating that the editing element 400 has a defect is displayed based on the received matching element information.

[0120] [Example A3 of Invention] In the exampleA1 or A2 of the invention, the request includes finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing, and includes a request for generating matching element information regarding a matching element matching the specifications of the finish element and described or to be described in the architectural drawing and the contents of the matching element, and the drawing creation support device further includes change means that changes the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means.

[0121] An embodiment of the examples A1, A3 of the invention will be described with Fig. 9 as an example. This embodiment is a modification of the first embodiment.

[0122] When the designer instructs the editing element 400 of the floor in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9 in the CAD software, in the drawing creation support device 100, a request is transmitted to the large language model server. The request includes the finish schedule data of Fig. 2 and the editing element 400, and includes a request for generating a matching element of the editing element 400. In response to this request, the matching element information is output from the large language model server. When the matching element information is received in the drawing creation 2026204855   23 Jun 2026 support device 100, the editing element 400 is changed based on the received matching element information.

[0123] [Example B1 of Invention] A drawing creation support device includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element, and element information acquisition means that acquires the relevant element information output from the large language model in response to the request.

[0124] [Example B2 of Invention] In the example B1 of the invention, the drawing creation support device further includes output means that outputs information regarding the other editing elements related to relevant element information acquired by the element information acquisition means based on the relevant element information.

[0125] An embodiment of the examples B1, B2 of the invention will be described with Fig. 9 as an example. This embodiment is a modification of the second embodiment, and the processing in Step S202 is replaced with processing of acquiring relevant element information from the large language model server.

[0126] When the designer changes the editing element 400 of the floor in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9, in the drawing creation support device 100, a request is transmitted to the large language model server. The request 2026204855   23 Jun 2026 includes the editing element 400, and includes a request for generating a relevant element of the editing element 400. In response to this request, the relevant element information is output from the large language model server. In the drawing creation support device 100, when the relevant element information is received, other editing elements to be changed with a change of the editing element 400 are displayed in a specific form based on the received relevant element information.

[0127] [Example B3 of Invention] In the example B1 or B2 of the invention, the request includes change element information regarding an editing element having been changed among editable editing elements in an architectural drawing, and includes a request for generating relevant element information regarding other editing elements to be changed with the change of the editing element and the contents of the other editing elements, and the drawing creation support device further includes change means that changes the other editing elements related to the relevant element information based on the relevant element information acquired by the element information acquisition means.

[0128] An embodiment of the examples B1, B3 of the invention will be described with Fig. 9 as an example. This embodiment is a modification of the second embodiment.

[0129] When the designer changes the editing element 400 of the floor in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9, in the drawing creation support device 100, a request is transmitted to the large language model server. The request includes the editing element 400, and includes a request for generating a relevant element of the editing element 400. In response to this request, the relevant element information is output from the large language model server. In the drawing creation support device 100, 2026204855   23 Jun 2026 when the relevant element information is received, other editing elements to be changed with a change of the editing element 400 are changed based on the received relevant element information.

[0130] [Example B4 of Invention] A drawing creation support device includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating determination necessity information regarding whether to require person’s determination for a change in other editing elements to be changed with the change of the editing element, and determination necessity information acquisition means that acquires the determination necessity information output from the large language model in response to the request.

[0131] [Example B5 of Invention] In the example B4 of the invention, the drawing creation support device further includes output means that outputs information regarding whether to require person’s determination for the change in the other editing elements related to the determination necessity information acquired by the determination necessity information acquisition means based on the determination necessity information.

[0132] [Example B6 of Invention] A drawing creation support device includes request input means that inputs, to a large language model, a request including change element information regarding an editing element having been changed among editable editing 2026204855   23 Jun 2026 elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with the change of the editing element and the contents of the other editing element to be changed and determination necessity information regarding whether to require person’s determination for the change in the other editing elements, and element information acquisition means that acquires the relevant element information and determination necessity information output from the large language model in response to the request.

[0133] [Example B7 of Invention] In the example B6 of the invention, the drawing creation support device further includes output means that outputs information regarding the other editing elements based on the relevant element information acquired by the element information acquisition means in a case where it is determined that person's determination is required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means.

[0134] [Example B8 of Invention] In the example B6 or B7 of the invention, the drawing creation support device further includes change means that changes the other editing elements based on the relevant element information acquired by the element information acquisition means in a case where it is determined that person's determination is not required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means. 2026204855   23 Jun 2026 An embodiment of the examples B4 to B8 of the invention will be described with Fig. 9 as an example. This embodiment is a modification of the third embodiment, and the processing in Step S254 is replaced with processing of acquiring relevant element information and determination necessity information from the large language model server.

[0136] When the designer changes the editing element 400 of the floor in the region displayed as “INNER CORRIDOR” in the sectional detail of Fig. 9, in the drawing creation support device 100, a request is transmitted to the large language model server. The request includes the editing element 400, and includes a request for generating a relevant element of the editing element 400 and the determination necessity information regarding whether to require person’s determination for a change of the relevant element. In response to this request, the relevant element information and the determination necessity information are output from the large language model server. In the drawing creation support device 100, when the relevant element information and the determination necessity information are received, whether to require person’s determination for the change of the relevant element is determined based on the received relevant element information. As a result, when it is determined that person’s determination is required, the relevant element is displayed in a specific form based on the received relevant element information. On the other hand, when it is determined that person’s determination is not required, the relevant element is changed based on the received relevant element information.

[0137] In the first to third embodiments and the modifications of the first to third embodiments, the drawing creation support device is implemented as a single device, but the present invention is not limited to the configuration. The drawing creation support device may be implemented as a network system. As an example of the network system, some or all of the functions of the drawing creation support device 100 may be configured as a virtual 2026204855   23 Jun 2026 server on a server that provides a cloud computing service.

[0138] In the first to third embodiments and the modifications of the first to third embodiments, the drawing creation support device 100 is configured to use the storage device 42, but the present invention is not limited to the configuration. The drawing creation support device 100 may be configured to use an external storage device such as a database server.

[0139] In the first to third embodiments and the modifications of the first to third embodiments, the case where the program saved in advance in the ROM 32 is executed for executing the processing illustrated in the flowcharts of Figs. 6, 8, 14, and 16 has been described, but the present invention is not limited to the configuration. A program indicating steps of such processing may be read from a storage medium into the RAM 34, and may be executed thereafter.

[0140] Here, the storage medium is a semiconductor storage medium such as a RAM or a ROM, a magnetic storage type storage medium such as an FD or an HD, an optical reading type storage medium such as a CD, a CDV, an LD or a DVD, or a magnetic storage / optical reading type storage medium such as an MO, and includes any storage medium regardless of, e.g., an electronic, magnetic, or optical reading method as long as the storage medium is a computer-readable storage medium.

[0141] The first to third embodiments and the modifications of the first to third embodiments are applicable to each other. The present invention is not limited to the first to third embodiments and the modifications of the first to third embodiments, and is also applicable to other cases without 2026204855   23 Jun 2026 departing from the gist of the present invention.

[0142] Unless the context requires otherwise, where the terms “comprise”, “comprises”, “comprised” or “comprising” are used in this specification (including the claims) they are to be interpreted as specifying the presence of the stated features, integers, steps or components, but not precluding the presence of one or more other features, integers, steps or components, or group thereof. LIST OF REFERENCE SIGNS

[0143] 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, 404, 408, 410       Editing Element 402, 406, 412, 416, 422, 426      Message Indicating Defect 414 Usable Dimension 420, 424Tag

Claims

1. An architectural drawing creation support system for generating a large languagemodel for an architectural drawing by learning based on information regarding finish element described in the finish schedule and information regarding matching element matching specifications of the finish element and described in the architectural drawing and / or other architectural drawings, comprising:request input means that inputs, to the large language model, a request including finish element information regarding a finish element described or to be described in a finish schedule and editing element information regarding an editable editing element in an architectural drawing and including a request for generating matching element information regarding a matching element matching specifications of the finish element and described or to be described in the architectural drawing; andelement information acquisition means that acquires the matching element information output from the large language model in response to the request,wherein the large language model is generated by a data set of the finish schedule being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter;wherein the large language model includes a learned parameter updated from the prelearning parameter by learning and an inference program; andwherein the inference program receives the input finish schedule data and editing element, and based on the learned parameter, estimates the matching element;a storage device configured to store the matching element information on the large language model.

2. The architectural drawing creation support system according to claim 1, further2026204855   23 Jun 2026comprising:output means that outputs information regarding the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means.

3. The architectural drawing creation support system according to any one of claims 1and 2, whereinthe request includes the finish element information regarding the finish element described or to be described in the finish schedule and the editing element information regarding the editable editing element in the architectural drawing, and includes a request for generating matching element information regarding the matching element matching the specifications of the finish element and described or to be described in the architectural drawing and contents of the matching element; andthe architectural drawing creation support system further comprises change means that changes the editing element related to the editing element information based on the matching element information acquired by the element information acquisition means.

4. An architectural drawing creation support system for updating a large languagemodel, comprising:request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element, andelement information acquisition means that acquires the relevant element information output from the large language model in response to the request,2026204855   23 Jun 2026wherein the large language model is generated by a learning data set being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter;wherein the large language includes a learned parameter updated from the prelearning parameter by learning and an inference program; andwherein the inference program receives the input of the changed editing element among the editable editing elements in the architectural drawing, and based on the learned parameter, estimates the other editing elements to be changed with the change of the editing element;a storage device configured to store the relevant element information on the large language model.

5. The architectural drawing creation support system according to claim 4, furthercomprising:output means that outputs information regarding the other editing elements related to the relevant element information acquired by the element information acquisition means based on the relevant element information.

6. The architectural drawing creation support system according to any one of claims 4and 5, whereinthe request includes the change element information regarding the editing element having been changed among the editable editing elements in the architectural drawing, and includes a request for generating the relevant element information regarding the other editing elements to be changed with the change of the editing element and contents of the other editing elements, and2026204855   23 Jun 2026the architectural drawing creation support system further comprises change means that changes the other editing elements related to the relevant element information acquired by the element information acquisition means based on the relevant element information.

7. An architectural drawing creation support system for updating a large languagemodel, comprising:request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating determination necessity information regarding whether to require person’s determination for a change in other editing elements to be changed with a change of the editing element; anddetermination necessity information acquisition means that acquires the determination necessity information output from the large language model in response to the request,wherein the large language model is generated by a learning data set being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter;wherein the large language model includes a learned parameter updated from the prelearning parameter by learning and an inference program; andwherein the inference program receives the input of the changed editing element among the editable editing elements in the architectural drawing, estimates the determination necessity information from the input editing element based on the learned parameter;a storage device configured to store the determination necessity information on the large language model.2026204855   23 Jun 20268.      The architectural drawing creation support system according to claim 7, furthercomprising:output means that outputs information regarding whether to require the person’s determination for the change in the other editing elements related to the determination necessity information acquired by the determination necessity information acquisition means based on the determination necessity information.

9. An architectural drawing creation support system for updating a large languagemodel, comprising:request input means that inputs, to the large language model, a request including change element information regarding an editing element having been changed among editable editing elements in an architectural drawing and including a request for generating relevant element information regarding other editing elements to be changed with a change of the editing element and contents of the other editing elements and determination necessity information regarding whether to require person’s determination for a change in the other editing elements; andelement information acquisition means that acquires the relevant element information and the determination necessity information output from the large language learned model in response to the request,wherein the large language is generated by a learning data set being input to a learning program, the learning program includes a pre-learning parameter and a hyperparameter, and performs learning based on the input learning data set and hyperparameter to update the pre-learning parameter;wherein the large language model includes a learned parameter updated from the prelearning parameter by learning and an inference program; andwherein the inference program receives the input of the changed editing element2026204855   23 Jun 2026among the editable editing elements in the architectural drawing, estimates other editing elements to be changed with the change of the editing element and the determination necessity information from the input editing element based on the learned parameter;a storage device configured to store the change element information and the determination necessity information on the large language model.

10. The architectural drawing creation support system according to claim 9, furthercomprising:output means that outputs information regarding the other editing elements based on the relevant element information acquired by the element information acquisition means in a case where it is determined that the person's determination is required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means.

11. The architectural drawing creation support system according to any one of claims 9and 10, further comprising:change means that changes the other editing elements based on the relevant element information acquired by the element information acquisition means in a case where it is determined that the person's determination is not required to change the other editing elements related to the relevant element information based on the determination necessity information acquired by the element information acquisition means.

Images