Information processing apparatus, control method of information processing apparatus, and program

Abstract

To provide an information processing apparatus which contributes to generating more accurate requirements definition using a language model.SOLUTION: An information processing apparatus includes acquiring means and first generating means. The acquiring means acquires first requirement information described in a natural language. The first generating means generates a prompt to be input into a first language model, to extract at least one of a component, a function of the component, and information related to performance from the acquired first requirement information. The information processing apparatus may further include verifying means that verifies validity of second requirement information acquired by inputting the generated prompt into the first language model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device, a control method for an information processing device, and a program. [Background technology]

[0002] There are techniques related to requirements (requirements definition) used in system design.

[0003] For example, Patent Document 1 describes a system for describing and verifying requirements specifications that enables verification of the necessity and sufficiency of lower-level requirements relative to higher-level requirements to suppress discrepancies in designer understanding and prevent requirement inconsistencies and oversights. The system in Patent Document 1 includes a requirements content input unit, a requirements structure input unit, and a requirements display unit. The requirements content input unit inputs requirements-related items representing requirements and tacit knowledge by symbolically representing each component of the requirements-related items and by expressing the relationships of the components with mathematical symbols. The requirements structure input unit sets and inputs the mutual relationships between multiple requirements-related items input from the requirements content input unit by breaking them down from higher to lower levels. The requirements display unit displays the relationship between the requirements-related items input to the requirements content input unit and the requirements-related items input to the requirements structure input unit. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-051234 Summary of the Invention [Problem to be solved by the invention]

[0005] Requirements definition is important in system development (system design) for large-scale networks, etc. In particular, requirements definition is important in developing systems that combine multiple components based on diverse system requirements.

[0006] In system development, engineers may create requirements definitions using natural language, and then use the created requirements definitions and language models (learning models obtained through machine learning) to organize the system's components, functions, performance, etc.

[0007] However, even if a requirements definition written in natural language is input to a language model (querying the language model), appropriate results are often not obtained. Specifically, the requirements definition generated by the language model may contain missing necessary information, added unnecessary information, or modifications not intended by the user. In particular, information (functions, performance) not included in the training data used to generate the language model may not be processed correctly, often leading to the missing or modified information.

[0008] A primary object of the present invention is to provide an information processing device, a control method for an information processing device, and a program that contribute to the generation of a more accurate requirements definition using a language model. [Means for solving the problem]

[0009] According to a first aspect of the present invention, there is provided an information processing device comprising: an acquisition means for acquiring first requirement information written in a natural language; and a first generation means for generating a prompt to be input into a first language model, the prompt being used to extract at least one of information relating to components, functions and performance of the components from the acquired first requirement information.

[0010] According to a second aspect of the present invention, there is provided a control method for an information processing device, comprising: an acquisition step of acquiring first requirement information written in a natural language; and a first generation step of generating a prompt to be input into a first language model, the prompt being used to extract at least one of information relating to components, functions and performance of the components from the acquired first requirement information.

[0011] According to a third aspect of the present invention, there is provided a program for causing a computer to execute an acquisition process for acquiring first requirement information written in a natural language, and a first generation process for generating prompts to be input into a first language model, the prompts being used to extract at least one of information relating to components, functions and performance of the components, from the acquired first requirement information. [Effects of the Invention]

[0012] According to each aspect of the present invention, an information processing device, a control method for an information processing device, and a program are provided that contribute to generating a more accurate requirements definition using a language model. Note that the effects of the present invention are not limited to those described above. The present invention may achieve other effects instead of or in addition to the effects described above. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram for explaining an outline of an embodiment. [Figure 2] FIG. 2 is a flowchart illustrating the operation of one embodiment. [Figure 3] FIG. 3 is a diagram illustrating an example of a schematic configuration of an information processing system according to an embodiment of the present disclosure. [Figure 4] FIG. 4 is a diagram illustrating an example of a processing configuration of the support device according to an embodiment of the present disclosure. [Figure 5] FIG. 5 is a diagram illustrating an example of template information according to an embodiment of the present disclosure. [Figure 6] FIG. 6 is a diagram illustrating an example of component information according to an embodiment of the present disclosure. [Figure 7] FIG. 7 is a flowchart illustrating an example of the operation of the system requirement control unit according to an embodiment of the present disclosure. [Figure 8] FIG. 8 is a diagram illustrating an example of a prompt according to an embodiment of the present disclosure. [Figure 9] FIG. 9 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 10]FIG. 10 is a diagram illustrating an example of a processing configuration of the support device according to an embodiment of the present disclosure. [Figure 11] FIG. 11 is a diagram illustrating an example of verification information according to an embodiment of the present disclosure. [Figure 12] FIG. 12 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 13] FIG. 13 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 14] FIG. 14 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 15] FIG. 15 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 16] FIG. 16 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 17] FIG. 17 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 18] FIG. 18 is a diagram illustrating an example of a processing configuration of the support device according to an embodiment of the present disclosure. [Figure 19] FIG. 19 is a diagram illustrating an example of example sentence information according to an embodiment of the present disclosure. [Figure 20] FIG. 20 is a diagram illustrating an example of instruction information according to an embodiment of the present disclosure. [Figure 21] FIG. 21 is a diagram illustrating an example of a prompt according to an embodiment of the present disclosure. [Figure 22] FIG. 22 is a diagram illustrating an example of a processing configuration of the support device according to an embodiment of the present disclosure. [Figure 23] FIG. 23 is a diagram illustrating an example of a display on a terminal according to an embodiment of the present disclosure. [Figure 24] FIG. 24 is a diagram illustrating an example of a hardware configuration of the support device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0014] First, an overview of one embodiment will be described. Note that the reference numerals in the drawings are added to each element for convenience as an example to facilitate understanding, and the description of this overview is not intended to be limiting in any way. Furthermore, unless otherwise specified, the blocks shown in each drawing represent functional units, not hardware units. Connection lines between blocks in each drawing include both bidirectional and unidirectional lines. Unidirectional arrows are used to schematically indicate the flow of main signals (data) and do not exclude bidirectionality. Note that in this specification and drawings, elements that can be similarly described may be assigned the same reference numerals to avoid redundant explanation.

[0015] An information processing device 100 according to an embodiment includes an acquiring means 101 and a first generating means 102 (see FIG. 1). The acquiring means 101 acquires first requirement information written in a natural language (step S1 in FIG. 2). The first generating means 102 generates a prompt to be input to a first language model, and to extract at least one of information on components, functions, and performance of the components from the acquired first requirement information (step S2).

[0016] The information processing device 100 generates a prompt including an instruction to extract at least one of components, functions, and performance of the components included in requirements information acquired from a user, and inputs the generated prompt into a language model. More specifically, the information processing device 100 generates a prompt including information on components, functions, and performance of the components that need to be extracted from the requirements information. The language model instructed by such a prompt can accurately extract functions and performance from the requirements information input by the user. As a result, the information processing device 100 can obtain a more accurate and appropriate requirements definition that is less likely to contain missing necessary information, added unnecessary information, or unintended modifications by the user.

[0017] Specific embodiments will be described in more detail below with reference to the drawings.

[0018] [First embodiment] The first embodiment will be described in more detail with reference to the drawings.

[0019] [System Configuration] As shown in FIG. 3, the information processing system according to the first embodiment includes a support device 10 and a terminal 20.

[0020] The support device 10 is an information processing device that supports a user in system design. For example, the support device 10 is a server installed on a network (on the cloud).

[0021] The support device 10 acquires requirements (information about requirements; requirement information) for a system that a user is designing. More specifically, the support device 10 acquires requirement information written in natural language (system requirements written in simple language).

[0022] The support device 10 extracts at least one of information on the system components, the functions of the components, and the performance of the components from the acquired requirements information. The support device 10 generates system requirements (requirements information) including information on the components, the functions of the components, and the performance of the components extracted from the requirements information input by the user.

[0023] More specifically, the support device 10 extracts components included in the system requirements obtained from the user and generates system requirements that itemize the functions and performance of each of the extracted components. The components that make up the system may include hardware and software.

[0024] For example, the support device 10 acquires requirement information such as "A live video distribution system that can be viewed by 10,000 people simultaneously. Recognizes human faces and displays names using AI application X1. Ensures communication quality of 10 Mbps or more."

[0025] The assistance device 10 generates requirement information such as "10,000 people using it simultaneously. Recognizes faces. Communication quality of 10 Mbps or more. Uses AI application X1." from the requirement information.

[0026] In the following description, the system requirements acquired by the assistance device 10 (requirements information input by the user) will be referred to as "input system requirements." Furthermore, the system requirements generated by extracting information about components, their functions, and performance from the input system requirements will be referred to as "generated system requirements." The assistance device 10 acquires the input system requirements and generates generated system requirements from the input system requirements.

[0027] The assistance device 10 uses a learning model when generating the generated system requirements. More specifically, the assistance device 10 outputs new system requirements generated from input system requirements input by a user using a language model such as a large language model (LLM).

[0028] A user operates the terminal 20 to access the assistance device 10. The user inputs various pieces of information to the assistance device 10 and obtains various pieces of information from the assistance device 10.

[0029] 3 is an example and is not intended to limit the configuration. For example, the information processing system may include multiple support devices 10. Load balancing and redundancy may be achieved by using multiple support devices 10.

[0030] [General operation] Next, the general operation of the support device 10 with respect to generating system requirements will be described.

[0031] As described above, the assistance device 10 acquires input system requirements (requirement information written in natural language) from the user's terminal 20. The assistance device 10 generates a prompt (query) to be input to a language model using the acquired input system requirements. The assistance device 10 generates a prompt to be input to the language model, which is a prompt for extracting at least one of information on components, functions, and performance of the components from the input system requirements.

[0032] At this time, the support device 10 generates a prompt using the "template information" and the "element information."

[0033] Template information is information that includes instructions for a language model. For example, a statement such as "Please implement requirements definition for system development" corresponds to template information. As a rule, fixed phrases are used for template information.

[0034] Component information consists of component names that can be used in system design, functional information about the configuration, and performance information. Component information defines the functions and performance of the components (hardware and software) that make up the system. For example, taking the component "camera," the configuration requirement information defines "taking photos and videos" as the camera's function, and "number of pixels, internal storage capacity" as the camera's performance indicators.

[0035] The component information is created by an engineer or the like who is familiar with the target system. The engineer envisions the components, functions, and performance required to generate the system requirements, and creates the component information.

[0036] When the assistance device 10 acquires input system requirements (requirement information) from a user, the assistance device 10 generates a prompt for input to a language model using template information and component information. The assistance device 10 generates a prompt composed of the template information, component information, and input system requirements.

[0037] The assistance device 10 acquires the generation system requirements by inputting the generated prompt into a language model, and presents the acquired generation system requirements to the user.

[0038] Next, the support device 10 according to the first embodiment will be described in detail.

[0039] 4 is a diagram illustrating an example of a processing configuration (processing modules) of the assistance device 10 according to an embodiment of the present disclosure. Referring to FIG. 4, the assistance device 10 includes a communication control unit 201, a language model management unit 202, a template management unit 203, a component management unit 204, a system requirement control unit 205, and a storage unit 206.

[0040] The communication control unit 201 is a means for controlling communication with other devices. For example, the communication control unit 201 receives data (packets) from the terminal 20. The communication control unit 201 also transmits data to the terminal 20. The communication control unit 201 passes data received from other devices to other processing modules. The communication control unit 201 transmits data acquired from other processing modules to other devices. In this way, other processing modules transmit and receive data to and from other devices via the communication control unit 201. The communication control unit 201 has a function as a receiving unit that receives data from other devices and a function as a transmitting unit that transmits data to other devices.

[0041] The language model management unit 202 is a means for controlling and managing the language model used to acquire generation system requirements.

[0042] As an example, the language model may be, but is not limited to, what is called a large-scale language model.

[0043] A language model is a machine learning model (also called a generative model) that takes language as input and outputs language. A language model learns the relationships between words in a sentence and generates related strings from a target string. By using a language model that has been trained on sentences and sentences from various contexts, it is possible to generate related strings with appropriate content that are related to the target string.

[0044] When a prompt generated based on an input system request or the like is input, the language model outputs an answer corresponding to the prompt.

[0045] The learning method of the language model is not particularly limited, but as an example, the language model may be learned to output at least one sentence including an input string. As a specific example, the language model is a GPT (Generative Pre-Transformer) that outputs a sentence including an input string by predicting a string that is likely to follow the input string.

[0046] Other examples of language models include T5 (Text-to-Text Transfer Transformer), BERT (Bidirectional Encoder Representations from Transformers), RoBERTa (Robustly optimized BERT approach), and ELECTRA (Efficiently Learning an Encoder that Classifies Token Replacements Accurately).

[0047] A system administrator or the like prepares a language model to be implemented in the support device 10 using a large amount of design materials, specifications, etc. related to system design.

[0048] The language model management unit 202 acquires a language model (a language model generated by learning using learning data) prepared by a system administrator or the like. For example, the language model management unit 202 acquires the language model using a GUI (Graphical User Interface) or the like. Alternatively, the language model management unit 202 may acquire the language model via a USB (Universal Serial Bus) memory or the like.

[0049] Alternatively, the language model management unit 202 may perform machine learning using training data to generate a language model.

[0050] The language model management unit 202 stores the acquired language model in the storage unit 206 .

[0051] The template management unit 203 is a means for controlling and managing template information. The template management unit 203 acquires template information used to generate prompts to be input to a language model from a system administrator or the like.

[0052] For example, the template management unit 203 acquires template information using a GUI, a USB memory, or the like. The template management unit 203 stores the acquired template information in the storage unit 206. For example, the template management unit 203 stores template information such as that shown in FIG. 5 in the storage unit 206.

[0053] The component management unit 204 is a means for controlling and managing component information. The component management unit 204 acquires, from a system administrator or the like, component information used to generate prompts to be input to a language model.

[0054] For example, the component management unit 204 acquires component information using a GUI, a USB memory, or the like. The component management unit 204 stores the acquired component information in the storage unit 206. For example, the template management unit 203 stores component information such as that shown in FIG. 6 in the storage unit 206.

[0055] The system requirement control unit 205 is a means for executing control for generating generated system requirements from input system requirements.

[0056] The system requirement control unit 205 has a function as an acquisition unit and a function as first generation means. As an acquisition unit, the system requirement control unit 205 acquires first requirement information written in a natural language. As first generation means, the system requirement control unit 205 generates a prompt to be input into a first language model, and a prompt to extract information about the function or performance of a component from the acquired first requirement information.

[0057] That is, the system requirement control unit 205 generates a prompt including an instruction to extract function information and performance information for each component element from the input information (system requirements input by the user to the support device 10).

[0058] 7 is a flowchart showing an example of the operation of the system requirement control unit 205 according to the embodiment of the present disclosure. The operation of the system requirement control unit 205 will be described with reference to FIG.

[0059] The system requirement control unit 205 acquires input system requirements from a user (user terminal 20) who has accessed the assistance device 10 (step S101). For example, the system requirement control unit 205 acquires input system requirements using a GUI or the like.

[0060] The system requirement control unit 205 generates a prompt for inputting to the language model using the acquired input system requirement, template information, and component information (step S102).

[0061] In this case, the system requirement control unit 205 generates a prompt having three areas. Specifically, the system requirement control unit 205 generates a prompt having a template information area, a component information area, and an input system requirement area. For example, the system requirement control unit 205 generates a prompt as shown in Fig. 8. In Fig. 8, the upper section corresponds to the template information area, the middle section corresponds to the component information area, and the lower section corresponds to the input system requirement area.

[0062] The system requirement control unit 205 copies the template information stored in the storage unit 206 into the template information area. The system requirement control unit 205 copies the component information stored in the storage unit 206 into the component information area. The system requirement control unit 205 copies the input system requirements acquired from the user into the input system requirement area.

[0063] The system requirement control unit 205 generates a prompt to be input to the language model by setting information corresponding to each of the three areas. The system requirement control unit 205 inputs the generated prompt to the language model and obtains an answer (generated system requirement) from the language model (step S103).

[0064] The system requirement control unit 205 provides the acquired generated system requirements to the user (step S104). For example, the system requirement control unit 205 displays a screen such as that shown in FIG.

[0065] In this way, the system requirement control unit 205 combines template information and component information to generate instructions (prompts) for the language model, and extracts components, their functions, and performance indicators from the requirement information entered by the user. For example, using the template information No. 3 shown in Figure 5 and a prompt including the component information shown in Figure 6, functional information (information related to functions) is extracted as a "functional requirement" from the system requirement entered by the user. Also, using the template information No. 4 shown in Figure 5 and a prompt including the component information shown in Figure 6, performance indicators (information related to performance) are extracted as a "non-functional requirement" from the system requirement entered by the user.

[0066] The storage unit 206 is a means for storing information necessary for the operation of the assistance device 10.

[0067] [Device] Detailed description of the terminal 20 will be omitted. Examples of the terminal 20 include mobile terminal devices such as smartphones, mobile phones, game consoles, and tablets, as well as computers (personal computers, laptop computers). The terminal 20 can be any equipment or device that can accept user operations and communicate with the assistance device 10.

[0068] As described above, the assistance device 10 according to the first embodiment holds information about components required for a requirements definition (component name, function information, performance information) as component information, and generates a prompt to be input to a language model using the component information. The assistance device 10 inputs the prompt generated using the requirement information and component information input by the user into the language model, extracting information about the components, functions, and performance included in the requirement information, and generates a new requirements definition. In other words, by generating an appropriate prompt using the component information, etc., the assistance device 10 can generate a more appropriate requirements definition (more useful requirements definition) that does not contain missing necessary information, added unnecessary information, or modifications unintended by the user.

[0069] Here, in order for the language model to output an appropriate response (result) to a user's request or inquiry, clear and specific instructions (prompts) are required. The assistance device 10 generates clear and specific instructions using template information and component information and inputs them to the language model, thereby obtaining a more accurate and precise requirements definition.

[0070] [Second embodiment] Next, the second embodiment will be described in detail with reference to the drawings.

[0071] The support device 10 according to the second embodiment holds information about the relationships between components and performance indicators. The support device 10 uses the information to verify the generated system requirements. More specifically, the support device 10 verifies the validity of the generated system requirements to detect defects (faults) in the generated system requirements that require correction or the like.

[0072] For example, the support device 10 detects inconsistencies in the generated system requirements, conflicts between components, conflicts between performance indicators, insufficient information, etc. The support device 10 requests the user to confirm the detected deficiencies and accepts the user's corrections to the generated system requirements.

[0073] The configuration of the information processing system according to the second embodiment can be the same as that of the first embodiment, and therefore a description corresponding to FIG. 3 will be omitted.

[0074] The following description will focus on the differences between the first and second embodiments.

[0075] 10 is a diagram illustrating an example of a processing configuration (processing module) of the support device 10 according to an embodiment of the present disclosure. Referring to FIG. 10, a verification information management unit 207 and a verification control unit 208 are added to the configuration of the support device 10 according to the first embodiment.

[0076] The verification information management unit 207 is a means for controlling and managing information (hereinafter referred to as verification information) used when verifying the validity of the generated system requirements.

[0077] A system administrator or the like registers verification information in the support device 10. The support device 10 uses the verification information to detect items that may be added to the generated system requirements, and to detect items that cause contradictions or conflicts in the generated system requirements, thereby verifying the validity of the generated system requirements.

[0078] Examples of items that can be added to the generated system requirements include performance indicators not explicitly stated in the generated system requirements, concrete components corresponding to abstract components included in the requirements, and components expected from components included in the requirements.

[0079] Examples of items that may cause contradictions or conflicts in the generated system requirements include combinations of components that cannot be used simultaneously, and combinations of performance indicators that are in a trade-off relationship.

[0080] The verification information includes items such as detection target items, validity verification items, and judgment conditions (see FIG. 11).

[0081] The detection target items are information about the components and performance indicators to be detected from the generated system requirements. For example, in Figure 11, components such as "video distribution system" and "virtual environment" and performance indicators such as "video quality" and "processing delay" are listed as detection target items.

[0082] Validity verification items are information about the viewpoints (perspectives) used when validating the detection target items. For example, Fig. 11 lists the validity verification items such as "no performance description," "abstract concrete," "expected configuration," "unavailable for simultaneous use," and "performance conflict."

[0083] The judgment condition is a condition for judging a generated system requirement including a detection target item as "valid" when the detection target item is included in the generated system requirement.

[0084] The validity verification item "No mention of performance" indicates that the validity of the generated system requirements is verified from the perspective of whether or not the performance indicators required for the verification target items listed in the generated system requirements are present. For example, in the example of the first line in Figure 11, if "video distribution system" is detected as a detection target item, the presence or absence of performance indicators such as "video quality" and "communication delay" is determined. If the generated system requirements for a video distribution system include a description of "video quality" and "communication delay," the generated system requirements are determined to be valid. In other words, if the generated system requirements for a video distribution system do not mention "video quality" or "communication delay," the generated system requirements are determined to be "invalid."

[0085] The validation item "Abstract Concrete" indicates that the validity of the generated system requirements is verified from the perspective of whether or not there are concrete components that should be included in the generated system requirements, corresponding to the abstract components included in the generated system requirements. For example, in the example in the second row of Figure 11, "Virtual Environment" is detected as a detection target item, and if the generated system requirements include components such as "Container" or "Virtual Machine," the generated system requirements are judged to be valid. In other words, a generated system requirement that lacks "Container" or "Virtual Machine," which are concrete means for realizing a "Virtual Environment," is judged to be invalid.

[0086] The validity verification item "expected configuration" indicates that the validity of the generated system requirements is verified from the perspective of whether or not other components expected from the components included in the generated system requirements exist. For example, in the example on the third line in Figure 11, if "virtual machine" is detected as a detection target item and "physical machine" is included in the generated system requirements, the generated system requirements are judged to be valid. In other words, a generated system requirement that lacks a "physical machine" that is essential for realizing a "virtual machine" is judged to be "invalid."

[0087] The validity verification item "Cannot be used simultaneously" indicates that the validity of the generated system requirements is verified from the perspective of whether there are any combinations of components that cannot be used simultaneously. For example, in the example on the fourth line in Figure 11, if "Software A" and "Software B" are not listed in the same generated system requirements, the generated system requirements are judged to be valid. In other words, if Software A and Software B are listed in the same generated system requirements, the generated system requirements are judged to be "invalid (or inappropriate)."

[0088] The validity verification item "performance conflict" indicates that the validity of the generation system requirements is verified from the perspective of whether or not there is a combination of performance indicators that are in a trade-off relationship. For example, in the example on the fifth line in Figure 11, it is determined that "video quality" and "processing delay" are in a trade-off relationship, and the conditions for achieving both of these performances are listed in the judgment conditions. Although not shown in Figure 11, a conflict (trade-off) can also occur between "cost" and "performance."

[0089] An engineer or the like generates verification information while anticipating detection target items and the like that are essential for system design, etc. A system administrator or the like inputs the generated verification information into the support device 10.

[0090] The verification information management unit 207 stores the acquired verification information in the storage unit 206. For example, the verification information management unit 207 stores table information such as that shown in FIG.

[0091] The verification control unit 208 is a means for controlling the verification of the generated system requirements generated from the input system requirements (the sentences input by the user). More specifically, the verification control unit 208 verifies the validity of the generated system requirements (requirement information) obtained by inputting a prompt generated based on the component information, etc., into a language model.

[0092] The verification control unit 208 acquires the generated system requirements from the system requirement control unit 205. The verification control unit 208 attempts to detect each detection target item described in the verification information from the generated system requirements. When the verification control unit 208 detects a detection target item from the generated system requirements, it determines the validity of the generated system requirements using the corresponding determination condition.

[0093] In the example of the first line in Figure 11, if "video distribution system" is detected from the generated system requirements, and if "video quality" or "communication delay" is described in the same generated system requirements, the generated system requirements are judged to be valid for the verification target item in the first line. On the other hand, if at least one of "video quality" and "communication delay" is not described in the generated system requirements, the generated system requirements for the verification target item "video distribution system" in the first line are judged to be invalid.

[0094] If the generated system requirements are judged to be "invalid," the verification control unit 208 requests the user to confirm this fact and make corrections as necessary. At this time, the verification control unit 208 requests the user to confirm and correct the above using an interface according to the verification target item judged to be invalid, the corresponding validity verification item (relationships between components, relationships between performance indicators), and the judgment conditions.

[0095] For example, if the generated system requirements are determined to be invalid due to "no description of performance" for the verification target item "video distribution system," the verification control unit 208 displays an interface screen such as that shown in Fig. 12 on the terminal 20. If the generated system requirements do not include communication delay, a necessary performance indicator, even though the "video distribution system" has been detected, the verification control unit 208 uses the interface such as that shown in Fig. 12 to ask the user to confirm this fact and make any necessary corrections. The verification control unit 208 acquires the performance indicators required for the video distribution system using information included in the judgment conditions.

[0096] If the generated system requirements are determined to be invalid due to the absence of performance information, the verification control unit 208 presents the user with performance parameter information for the components included in the generated system requirements. Furthermore, the verification control unit 208 prompts the user to input performance parameter information corresponding to the components included in the generated system requirements.

[0097] Alternatively, if the generated system requirements are determined to be invalid due to an "abstract concrete" related to the verification target item "virtual environment," the verification control unit 208 displays an interface screen such as that shown in FIG. 13 on the terminal 20. If the "virtual environment" is detected but the components necessary to instantiate the virtual environment are not included in the generated system requirements, the verification control unit 208 uses the interface shown in FIG. 13 to ask the user to confirm this fact and make any necessary corrections. The verification control unit 208 acquires components for realizing the abstract components from the judgment conditions and asks the user to add the acquired components. Note that if the factor that caused the invalidity to be determined is related to the "abstract concrete," the verification control unit 208 may present the user with an option that does not reveal the specific components, as shown in FIG. 13.

[0098] If the generated system requirements are determined to be invalid due to the abstract concrete, the verification control unit 208 presents the user with candidate components for embodying the components included in the generated system requirements. That is, the verification control unit 208 accepts input regarding the selection of components (parts) that embody the components included in the generated system requirements. The verification control unit 208 may also present the user with an option to indicate that additional information regarding the configuration requirements is not to be input.

[0099] Alternatively, if the generated system requirements are determined to be invalid due to the "expected configuration" related to the verification target item "virtual machine," the verification control unit 208 displays an interface screen such as that shown in FIG. 14 on the terminal 20. If the "virtual machine" is detected but another component required to run the virtual machine is not included in the generated system requirements, the verification control unit 208 uses the interface shown in FIG. 14 to ask the user to confirm this fact and make any necessary corrections. The verification control unit 208 obtains other components expected from one component from the judgment conditions and asks the user to add the obtained components. Note that even if the factor that caused the invalidity to be determined is related to the "expected configuration," the verification control unit 208 may also present the user with an option that does not explicitly state the specific components, as shown in FIG. 14.

[0100] As described above, if the system requirements are determined to be invalid because the relationships between components or performance indicators are "no performance description," "abstract concrete," or "expected configuration," the verification control unit 208 detects elements that can be added to the system requirements. That is, the verification control unit 208 verifies the validity of the generated system requirements information based on at least information about the relationships between components or performance indicators contained in the generated system requirements (second requirements information). If the generated system requirements are determined to be invalid, the verification control unit 208 detects elements (components, performance indicators) that can be added to the generated system requirements.

[0101] Furthermore, the verification control unit 208 provides an interface that allows the user to identify elements that can be added, and / or an interface that allows the user to input additional information about the elements that can be added (see FIGS. 12 to 14). Note that FIG. 12 and the like are diagrams illustrating a case where an interface that allows the user to identify elements that can be added and an interface that allows the user to input additional information about the elements that can be added are simultaneously provided. However, either one of these interfaces may be provided to the user.

[0102] If the generated system requirements are determined to be invalid due to "simultaneous use not permitted" for the verification target items "Software A" or "Software B," the verification control unit 208 displays an interface screen such as that shown in Fig. 15 on the terminal 20. If the generated system requirements are determined to be "invalid" due to the factor of "simultaneous use not permitted," the verification control unit 208 prompts the user to select one of the components that will be used simultaneously using the interface such as that shown in Fig. 15.

[0103] If the generated system requirements are determined to be invalid due to the inability to use them simultaneously, the verification control unit 208 presents the conflicting components to the user. Furthermore, the verification control unit 208 accepts an input regarding which of the conflicting components (parts) should be prioritized.

[0104] Alternatively, if the generated system requirements are determined to be inappropriate due to a "performance conflict" related to the verification target items "video quality" or "processing delay," the verification control unit 208 displays an interface screen such as that shown in Fig. 16 on the terminal 20. If the generated system requirements are determined to be "inappropriate" due to a performance conflict, the verification control unit 208 prompts the user to select a prioritized performance index using the interface such as that shown in Fig. 16.

[0105] If the generated system requirements are determined to be invalid due to performance conflicts, the verification control unit 208 presents the performance requirements in a trade-off relationship to the user. Furthermore, the verification control unit 208 accepts input regarding which performance among the performance requirements in the trade-off relationship should be prioritized for adoption.

[0106] As described above, when a system requirement is determined to be invalid because the relationship between components or performance indicators is "impossible to use simultaneously" or "performance conflict," the verification control unit 208 detects multiple conflicting components or multiple conflicting performance indicators. That is, the verification control unit 208 verifies the validity of the generated system requirement based at least on information regarding the relationship between components or performance indicators included in the generated system requirement (second requirement information). When a generated system requirement is determined to be invalid, the verification control unit 208 detects multiple conflicting components or multiple conflicting performance indicators in the generated system requirement.

[0107] The verification control unit 208 provides an interface that allows a user to identify multiple conflicting components or multiple conflicting performance indicators, and / or an interface that allows a user to input a component or performance indicator that the user prioritizes from the multiple conflicting components or performance indicators. Figures 15 and 16 show a case where an interface that allows a user to identify conflicting components and an interface that allows a user to input a component that the user prioritizes from the conflicting components are simultaneously provided. However, either one of these interfaces may be provided to the user.

[0108] The interface screens shown in FIGS. 12 to 16 are merely examples, and the interface used to prompt the user to confirm and correct the generated system requirements that have been determined to be invalid is not limited to the interface screen shown in FIG. 12 and the like.

[0109] For example, if the generated system requirements are determined to be invalid due to a "performance conflict" related to the verification target items "video quality" or "processing delay," the verification control unit 208 may display an interface screen such as that shown in Fig. 17 on the terminal 20. The verification control unit 208 may present to the user the components and performance indicators that caused the determination to be invalid so that they can be distinguished from other components. For example, in Fig. 17, the performance indicators (specific numerical values ​​of the performance indicators) that are causing the performance conflict are displayed in bold.

[0110] The method of indicating components by the verification control unit 208 is not limited to displaying the relevant components or performance indicators in bold. For example, the verification control unit 208 may indicate factors (components, performance indicators, etc.) that are determined to be invalid by changing the color, underlining, displaying line numbers, etc. Alternatively, the verification control unit 208 may display a pop-up or the like to allow the user to correct the generated system requirement that is determined to be invalid when the user places the cursor or the like over a factor (component, performance indicator, etc.) that is determined to be invalid.

[0111] In this way, the verification control unit 208 may display the factors that determined the generated system requirements to be invalid through verification, and present the user with an interface screen that clearly shows areas where information is lacking, areas where performance conflicts exist, etc.

[0112] As described above, the support device 10 according to the second embodiment holds information (verification information) on the relationships between components. Using the verification information, the support device 10 determines whether or not there are any contradictions, conflicts, etc. in the requirements information generated from the requirements information acquired from the user. If there are any deficiencies, such as contradictions or conflicts, in the generated requirements information, the support device 10 requests the user to confirm the deficiencies and accepts any necessary corrections. As a result, the support device 10 can provide the user with a more appropriate requirements definition (system requirements).

[0113] [Third embodiment] Next, the third embodiment will be described in detail with reference to the drawings.

[0114] In the first embodiment, a case has been described in which a prompt to be input to a language model is generated using input system requirements acquired from a user as they are. However, the input system requirements entered by the user may contain sentences with complex structures. When a prompt is generated using such input system requirements, it may be difficult to obtain more accurate generation system requirements.

[0115] Therefore, in the third embodiment, a case will be described in which the assistance device 10 converts input system requirements acquired from a user into a format suitable for a language model. More specifically, the assistance device 10 according to the third embodiment stores example sentence information that paraphrases (converts) complex sentences into simple sentences. The assistance device 10 generates prompts to be input to the language model using the input system requirements converted using the example sentence information.

[0116] The configuration of the information processing system according to the third embodiment can be the same as that of the first embodiment, and therefore a description corresponding to FIG. 3 will be omitted.

[0117] The following description will focus on the differences between the first to third embodiments.

[0118] 18 is a diagram illustrating an example of a processing configuration (processing module) of the assistance device 10 according to an embodiment of the present disclosure. Referring to FIG. 18, an example sentence information management unit 209 is added to the configuration of the assistance device 10 according to the first embodiment.

[0119] The example sentence information management unit 209 is a means for controlling and managing example sentence information for converting input system requirements.

[0120] A system administrator or the like generates example sentence information in order to obtain more accurate generation system requirements. For example, the system administrator or the like generates an example sentence that converts an input sentence into a predetermined format, an example sentence that converts an input sentence into a list of items, an example sentence that converts an input sentence from a complex structure into a simple structure, etc.

[0121] The system administrator or the like inputs the generated example sentence information into the assistance device 10.

[0122] For example, the sentence example information management unit 209 acquires sentence example information such as that shown in Fig. 19. The sentence example information management unit 209 stores the acquired sentence example information in the storage unit 206.

[0123] Here, the input system requirements are converted into input system requirements with more appropriate sentences using a language model, which is different from the language model for obtaining the generated system requirements from the input system requirements.

[0124] In the following description, a language model for converting input system requirements will be referred to as a "conversion language model," and a language model for obtaining generated system requirements will be referred to as a "requirements generation language model."

[0125] The language model management unit 202 according to the third embodiment acquires a conversion language model from a system administrator or the like, and stores it in the storage unit 206 .

[0126] In addition, the system administrator or the like generates instruction information that matches the generated example sentence information. The instruction information is used when generating prompts to be input to the conversion language model. The instruction information is information for giving instructions regarding the use of the example sentence information to the conversion language model.

[0127] For example, instruction information as shown in Fig. 20 is generated in conjunction with example sentence information as shown in Fig. 19. The example sentence information management unit 209 acquires the instruction information together with the example sentence information from the system administrator or the like, and stores them in the storage unit 206.

[0128] The system requirement control unit 205 according to the third embodiment has a function as a second generation unit. The system requirement control unit 205 as the second generation unit generates a prompt for inputting into a second language model (a conversion language model) and for converting requirement information acquired from a user based on pre-stored example sentence information.

[0129] For example, the system requirement control unit 205 may generate a prompt that divides a sentence with a complex structure described in the input system requirement into multiple sentences with simple structures. Alternatively, the system requirement control unit 205 may generate a prompt that complements a parameter included in the input system requirement.

[0130] When the system requirement control unit 205 acquires input system requirements from the user, it generates a prompt for inputting the acquired input system requirements, example sentence information, and instruction information into the conversion language model.

[0131] For example, the system requirement control unit 205 generates a prompt as shown in Fig. 21. Specifically, the system requirement control unit 205 copies the instruction information stored in the storage unit 206 to the instruction information area shown in Fig. 21. Similarly, the system requirement control unit 205 copies the example sentence information to the example sentence information area, and copies the input system requirements acquired from the user to the input system requirement area.

[0132] The system requirement control unit 205 converts the input system requirement acquired from the user by inputting the generated prompt into a conversion language model.

[0133] For example, in the example of Figure 21, the input system requirement (the input sentence entered by the user) is converted into a sentence such as "Five access points are combined to form a mesh network. The access points can connect up to 30 terminals."

[0134] The system requirement control unit 205 treats the converted input system requirements as the input system requirements described in the first embodiment and generates a prompt to be input to the requirements generation language model. The system requirement control unit 205 obtains the final generated system requirements using the generated prompt.

[0135] As described above, the support device 10 according to the third embodiment converts natural language requirements into a list of simple requirements using example sentences (example sentence information) that rephrase complex sentences into simple sentences. Since the sentences included in the system requirements are converted according to the simple example sentences defined for each component, the possibility of errors occurring in the extraction of configuration information and performance information using a language model (requirements generation language model) is reduced.

[0136] [Fourth embodiment] Next, the fourth embodiment will be described in detail with reference to the drawings.

[0137] The support device 10 according to the fourth embodiment outputs structured generated system requirements (generated system requirements converted into a predetermined format or generated system requirements expressed in a predetermined vocabulary). Furthermore, the support device 10 uses the structured generated system requirements to output a specific design proposal for the system. More specifically, the support device 10 utilizes automatic design technology to translate the system requirements input by the user into a specific design.

[0138] The configuration of the information processing system according to the fourth embodiment can be the same as that of the first embodiment, and therefore a description corresponding to FIG. 3 will be omitted.

[0139] The following description will focus on the differences between the first to fourth embodiments.

[0140] 22 is a diagram illustrating an example of a processing configuration (processing module) of the support device 10 according to the embodiment of the present disclosure. Referring to FIG. 22, an automatic design control unit 210 is added to the configuration of the support device 10 according to the first embodiment.

[0141] The system requirement control unit 205 of the support device 10 according to the fourth embodiment instructs the language model (language model for generating requirements) to output generated system requirements in a machine-processable format, rather than generated system requirements written in natural language.

[0142] For example, the system requirement control unit 205 adds an instruction such as "Please output the system requirements in JSON format" to the template information area described in the first embodiment. By receiving the prompt with this instruction added, the language model outputs the generated system requirements written in JSON (JavaScript Object Notation) format.

[0143] In other words, the system requirement control unit 205 can obtain system requirements in a machine-processable format such as JSON format (structured system requirements) by issuing appropriate instructions to the language model. The system requirement control unit 205 passes the obtained generated system requirements (structured system requirements) to the automatic design control unit 210.

[0144] Based on the acquired generated system requirements, the automatic design control unit 210 generates configuration information that embodies the system configuration defined by the system requirements. That is, the automatic design control unit 210 uses the structured generated system requirements to generate a specific design proposal for the system.

[0145] The automatic design control unit 210 generates the specific design proposal using information on specific parts, applications, etc. corresponding to components that may be included in the generated system requirements. For example, the automatic design control unit 210 generates specific design proposals for the physical server and camera described in the system requirements using information on various products for the component "physical server" and the component "camera."

[0146] The automatic design control unit 210 can generate a specific configuration plan from the generated system requirements using any automatic design technology. For example, the automatic design control unit 210 can obtain a specific design plan by utilizing the technology disclosed in Reference 1 below.

[0147] Reference 1: JP 2021-135625 A

[0148] Since the automatic design technology is different from the purpose of the present disclosure, detailed explanation will be omitted. Reference 1 discloses a technology for generating concrete configuration information by repeating the operation of gradually concretizing the abstract system configuration included in given system requirements.

[0149] The automatic design control unit 210 presents the system configuration plan obtained from the structured generated system requirements to the user. For example, the automatic design control unit 210 displays the system configuration plan shown in FIG. 23 on the terminal 20.

[0150] In this way, the automatic design control unit 210 generates specific design proposals for realizing the configuration of components included in the system requirements (e.g., physical servers, cameras, etc.). To enable the generation of the specific design proposals, the system requirement control unit 205 generates generated system requirements in a machine-processable format, rather than generated system requirements written in natural language.

[0151] The system requirement control unit 205 according to the fourth embodiment may generate a prompt to output the generated system requirements written in natural language and the structured generated system requirements. Furthermore, the system requirement control unit 205 may generate a prompt including an instruction to add and output the correspondence between the components included in the system requirements written in natural language and the structured system requirements. Furthermore, the automatic design control unit 210 may display on the terminal 20 the correspondence between the results of the automatic design (a design proposal that embodies the requirements definition) and the original input system requirements.

[0152] <Modification of the Fourth Embodiment> Here, when the automatic design control unit 210 tries to obtain detailed results when obtaining specific design proposals for system requirements, the processing takes a long time. In situations where requirements definitions are actually used, detailed information such as design proposals in the design process is often not necessary. Therefore, there is a demand for generating design proposals with the minimum necessary specificity in a short amount of time.

[0153] To meet such demands, the support device 10 may hold model information (model information to be referenced during automatic design) to which information regarding the realization level is added. Alternatively, the support device 10 may instruct the automatic design control unit 210 on the realization level desired in the automatic design.

[0154] For example, if the specified realization level is low, the type of OS to be used in the system is displayed and the generation of the design proposal is completed. On the other hand, if the specified realization level is high, not only the type of OS but also the specific product name is displayed and the generation of the design proposal is completed.

[0155] For example, when the support device 10 presents the generated system requirements to the user, it acquires information as to whether the user wishes to receive a design proposal for the system requirements. If the user wishes to receive a design proposal, the system requirement control unit 205 may acquire the desired concretization level of the user and notify the automatic design control unit 210 of the acquired concretization level.

[0156] Alternatively, the support device 10 may provide an interface that allows the specification of the concrete level for each component included in the generated system requirements.

[0157] In this way, the support device 10 may output, in a short time, a design proposal suited to the site and situation where the requirements definition is utilized.

[0158] As described above, the support device 10 according to the fourth embodiment generates a prompt including an instruction to generate structured second requirements information (generated system requirements). Based on the structured second requirements information, the support device 10 generates a specific design proposal for the structured second requirements information and presents the generated design proposal to the user. As a result, the burden on the user regarding system design work is reduced.

[0159] Next, the hardware of each device constituting the information processing system will be described. Fig. 24 is a diagram showing an example of the hardware configuration of the support device 10.

[0160] The support device 10 can be configured by an information processing device (so-called computer), and has the configuration exemplified in Fig. 24. For example, the support device 10 includes a processor 311, a memory 312, an input / output interface 313, and a communication interface 314. The components such as the processor 311 are connected by an internal bus or the like, and are configured to be able to communicate with each other.

[0161] However, the configuration shown in Fig. 24 is not intended to limit the hardware configuration of the support device 10. The support device 10 may include hardware not shown, and may not include the input / output interface 313 as necessary. Furthermore, the number of processors 311 and the like included in the support device 10 is not intended to be limited to the example shown in Fig. 24, and for example, the support device 10 may include multiple processors 311.

[0162] The processor 311 is a programmable device such as a central processing unit (CPU), a micro processing unit (MPU), a digital signal processor (DSP), a tensor processing unit (TPU), or a graphics processing unit (GPU). Alternatively, the processor 311 may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or other device. The processor 311 executes various programs including an operating system (OS).

[0163] The memory 312 is a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), a solid state drive (SSD), etc. The memory 312 stores an OS program, application programs, and various data.

[0164] The input / output interface 313 is an interface for a display device and an input device (not shown). The display device is, for example, a liquid crystal display. The input device is, for example, a device that accepts user operations such as a keyboard or a mouse.

[0165] The communication interface 314 is a circuit, module, etc. that communicates with other devices. For example, the communication interface 314 includes a network interface card (NIC).

[0166] The functions of the assistance device 10 are realized by various processing modules. The processing modules are realized, for example, by the processor 311 executing a program stored in the memory 312. The program can be recorded on a computer-readable storage medium. The storage medium can be a non-transitory medium such as a semiconductor memory, a hard disk, a magnetic recording medium, or an optical recording medium. That is, the present invention can also be embodied as a computer program product. The program can be downloaded via a network or updated using a storage medium storing the program. The processing modules can also be realized by a semiconductor chip.

[0167] The terminal 20 can also be configured by an information processing device, similar to the support device 10, and its basic hardware configuration is no different from that of the support device 10, so a description thereof will be omitted.

[0168] The support device 10, which is an information processing device, is equipped with a computer, and the functions of the support device 10 can be realized by causing the computer to execute a program. Furthermore, the support device 10 executes a control method for the support device 10 (information processing device) by the program.

[0169] [Variations] The configuration, operation, etc. of the information processing system described in the above embodiment are merely examples, and are not intended to limit the configuration, etc. of the system.

[0170] The assistance device 10 may acquire input system requirements (requirement information) from a user using voice recognition technology. Specifically, the system requirement control unit 205 acquires voice data via a microphone provided in the terminal 20. The system requirement control unit 205 acquires the input system requirements converted into text by applying voice recognition technology to the voice data. The system requirement control unit 205 uses the acquired input system requirements to generate a prompt to be input into a language model.

[0171] The assistance device 10 may acquire input system requirements using a technology related to generating minutes. For example, the system requirements control unit 205 may generate minutes from text data or voice data related to a meeting between a vendor company and a client company. The system requirements control unit 205 may treat the generated minutes as input system requirements and generate a prompt to input the minutes into a language model. Alternatively, the assistance device 10 may generate (acquire) input system requirements, as needed, by utilizing a learning model that generates a summary from the minutes.

[0172] The support device 10 may store system requirements previously acquired from users and system requirements previously generated, and detect differences (changes and additions) between a system requirement newly input by the user and the stored system requirements. The system requirement control unit 205 may generate a prompt including the detected differences, and extract function information, index information, etc. from the system requirements with the differences.

[0173] The assistance device 10 may hold template information and component information for each technical field (each type of system). For example, the assistance device 10 may store template information and component information for each system, such as an information distribution system, an online conference system, and a face authentication system. In this case, before providing a service to a user, the assistance device 10 may acquire the type of system that the user is designing, and use the acquired component information corresponding to the system to generate a prompt.

[0174] In the second embodiment, the assistance device 10 may use a learning model that, when a generation system requirement is input, outputs a determination result of validity and a portion that needs to be corrected.

[0175] In the third embodiment, the assistance device 10 may present to the user the system requirements (converted input sentences) converted by the system requirements control unit 205. In this case, the assistance device 10 may display the converted input sentences on the terminal 20 in a format that allows the user to compare the input sentences before and after conversion. By checking the converted input sentences, the user can learn how to generate more appropriate system requirements.

[0176] In the above embodiment, the support device 10 is described as a server in a server-client system. However, the support device 10 may be implemented as a user terminal 20 on which a predetermined application is installed.

[0177] In the above embodiment, an example has been described in which a language model is implemented in the assistance device 10. However, the language model may be implemented in a server or the like different from the assistance device 10. In this case, the assistance device 10 generates a prompt to be input to the language model and transmits the generated prompt to an external server (a server in which the language model is implemented). The assistance device 10 may obtain an output of the language model (for example, structured requirement information candidates, etc.) from the external server.

[0178] In the flow charts (flowcharts, sequence diagrams) used in the above explanation, multiple steps (processes) are described in order, but the execution order of the steps executed in the embodiments is not limited to the order described. In the embodiments, the order of the illustrated steps can be changed to the extent that the content is not affected, such as by executing each process in parallel.

[0179] The above-described embodiments have been described in detail to facilitate understanding of the present disclosure, and it is not intended that all of the above-described configurations are required. Furthermore, when multiple embodiments are described, each embodiment may be used alone or in combination. For example, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of one embodiment with another configuration.

[0180] From the above explanation, it is clear that the present invention has industrial applicability, and the present invention can be suitably applied to information processing systems that support system design by users.

[0181] Some or all of the above embodiments can be described as, but are not limited to, the following supplementary notes.

[0182] [Appendix 1] an acquisition means for acquiring first requirement information written in a natural language; a first generating means for generating a prompt to be input to a first language model, the prompt being for extracting at least one of information on components, functions and performances of the components from the acquired first requirements information; An information processing device comprising: [Appendix 2] 2. The information processing device according to claim 1, further comprising a verification means for verifying the validity of second requirement information obtained by inputting the generated prompt into the first language model. [Appendix 3] The verification means verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; 3. The information processing device according to claim 2, further comprising: detecting an element that can be added to the second requirement information when the second requirement information is determined to be invalid. [Appendix 4] 4. The information processing device of claim 3, further comprising a providing means for providing an interface that allows a user to identify the elements that can be added, and / or an interface that allows the user to input additional information regarding the elements that can be added. [Appendix 5] The verification means verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; 3. The information processing device according to claim 2, wherein, when the second requirement information is determined to be invalid, multiple conflicting components or multiple conflicting performance indicators are detected in the second requirement information. [Appendix 6] 6. The information processing device according to claim 5, further comprising a providing means for providing an interface that allows a user to identify the plurality of competing components or the plurality of competing performance indicators, and / or an interface that allows the user to input a component or performance indicator that the user prioritizes from the plurality of competing components or performance indicators. [Appendix 7] 7. The information processing device according to claim 1, further comprising: a second generating means for generating a prompt for inputting to a second language model, the prompt being for converting the first requirement information based on pre-stored example sentence information. [Appendix 8] the first generating means generates the prompt including an instruction to generate the second structured requirements information; 7. The information processing device according to any one of claims 2 to 6, further comprising a control means for generating a specific design proposal for the structured second requirements information based on the structured second requirements information, and presenting the generated design proposal to a user. [Appendix 9] an acquisition step of acquiring first requirement information written in a natural language; a first generation step of generating prompts to be input into a first language model, the prompts being for extracting at least one of information on components, functions and performances of the components from the acquired first requirements information; A control method for an information processing device comprising: [Appendix 10] The control method for an information processing device described in Appendix 9, further comprising a verification step of verifying the validity of second requirement information obtained by inputting the generated prompt into the first language model. [Appendix 11] The verification step includes: verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; The control method for an information processing device according to claim 10, further comprising detecting elements that can be added to the second requirement information if the second requirement information is determined to be invalid. [Appendix 12] The control method for an information processing device described in Appendix 11, further comprising a providing step of providing an interface that allows a user to identify the elements that can be added, and / or an interface that allows the user to input additional information regarding the elements that can be added. [Appendix 13] The verification step includes: verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; A control method for an information processing device according to claim 10, further comprising detecting multiple conflicting components or multiple conflicting performance indicators in the second requirement information when the second requirement information is determined to be invalid. [Appendix 14] The control method for an information processing device described in Appendix 13, further comprising a providing step of providing an interface that allows a user to identify the plurality of competing components or the plurality of competing performance indicators, and / or an interface that allows the user to input a component or performance indicator that the user prioritizes from the plurality of competing components or performance indicators. [Appendix 15] A control method for an information processing device according to any one of appendices 9 to 14, further comprising a second generation step of generating a prompt for inputting into a second language model, the prompt being for converting the first requirement information based on pre-stored example sentence information. [Appendix 16] the first generating step generates the prompt including an instruction to generate the structured second requirements information; 15. The control method for an information processing device according to any one of appendices 10 to 14, further comprising a control step of generating a specific design proposal for the structured second requirements information based on the structured second requirements information, and presenting the generated design proposal to a user. [Appendix 17] On the computer, an acquisition process for acquiring first requirement information written in a natural language; a first generation process for generating a prompt to be input to a first language model, the prompt being for extracting at least one of information on components, functions, and performance of the components from the acquired first requirements information; A program to execute. [Appendix 18] 18. The program of claim 17, further comprising a verification process for verifying the validity of second requirement information obtained by inputting the generated prompt into the first language model. [Appendix 19] The verification process includes: verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; 19. The program according to claim 18, further comprising: detecting elements that can be added to the second requirement information if the second requirement information is determined to be invalid. [Appendix 20] 20. The program of claim 19, further comprising: a providing process that provides an interface that allows a user to identify the elements that can be added, and / or an interface that allows the user to input additional information about the elements that can be added. [Appendix 21] The verification process includes: verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; 19. The program described in Appendix 18, which, if the second requirements information is determined to be invalid, detects multiple conflicting components or multiple conflicting performance indicators in the second requirements information. [Appendix 22] The program described in Appendix 21, further executing a providing process to provide an interface that allows a user to identify the plurality of competing components or the plurality of competing performance indicators, and / or an interface that allows the user to input a component or performance indicator that the user prioritizes from the plurality of competing components or performance indicators. [Appendix 23] 23. The program according to any one of appendices 17 to 22, further comprising a second generation process for generating a prompt for inputting to a second language model, the prompt for converting the first requirement information based on pre-stored example sentence information. [Appendix 24] the first generation process generates the prompt including an instruction to generate the structured second requirements information; 23. The program according to any one of appendices 18 to 22, further executing a control process to generate a specific design proposal for the structured second requirements information based on the structured second requirements information, and present the generated design proposal to a user.

[0183] Furthermore, some or all of the configurations described in Supplementary Notes 2 to 8, which are dependent on Supplementary Note 1, may also be dependent on Supplementary Notes 9 and 17 in the same dependent relationship as Supplementary Notes 2 to 8. Furthermore, not limited to Supplementary Notes 1, 9, and 17, but within the scope of each of the above-mentioned embodiments, some or all of the configurations described as Supplements may also be dependent on various hardware, software, various recording means for recording software, or systems.

[0184] The disclosures of the above-cited prior art documents are incorporated herein by reference. Although the embodiments of the present invention have been described above, the present invention is not limited to these embodiments. Those skilled in the art will understand that these embodiments are merely illustrative and that various modifications are possible without departing from the scope and spirit of the present invention. In other words, the present invention naturally includes various modifications and alterations that may be made by those skilled in the art in accordance with the entire disclosure, including the claims, and the technical concepts thereof. [Explanation of symbols]

[0185] 10 Support equipment 20 terminals 100 Information processing device 101 Acquisition method 102 first generating means 201 Communication control unit 202 Language Model Management Unit 203 Template Management Department 204 Component Management Department 205 System Requirements Control Section 206 Memory section 207 Verification Information Management Unit 208 Verification control section 209 Example sentence information management department 210 Automatic Design Control Unit 311 processor 312 memory 313 Input / Output Interface 314 Communication Interface

Claims

1. an acquisition means for acquiring first requirement information written in a natural language; a first generating means for generating a prompt to be input to a first language model, the prompt being for extracting at least one of information on components, functions and performances of the components from the acquired first requirements information; An information processing device comprising:

2. The information processing apparatus according to claim 1 , further comprising a verifying unit that verifies the validity of second requirement information obtained by inputting the generated prompt into the first language model.

3. The verification means verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; The information processing apparatus according to claim 2 , further comprising: detecting an element that can be added to the second requirement information when the second requirement information is determined to be invalid.

4. The information processing apparatus according to claim 3 , further comprising: a providing unit that provides an interface that allows a user to identify the element that can be added, and / or an interface that allows the user to input additional information related to the element that can be added.

5. The verification means verifying the validity of the second requirements information based on at least information on relationships between components or performance indicators included in the second requirements information; The information processing apparatus according to claim 2 , further comprising: detecting a plurality of conflicting components or a plurality of conflicting performance indicators in the second requirement information when the second requirement information is determined to be invalid.

6. 6. The information processing device according to claim 5, further comprising a providing means for providing an interface that allows a user to identify the plurality of competing components or the plurality of competing performance indicators, and / or an interface that allows the user to input a component or performance indicator that the user prioritizes from the plurality of competing components or performance indicators.

7. 7. The information processing apparatus according to claim 1, further comprising: second generating means configured to generate a prompt for inputting to a second language model, the prompt being for converting the first requirement information based on pre-stored example sentence information.

8. the first generating means generates the prompt including an instruction to generate the second structured requirements information; 7. The information processing device according to claim 2, further comprising a control means for generating a specific design proposal for the structured second requirements information based on the structured second requirements information, and presenting the generated design proposal to a user.

9. an acquisition step of acquiring first requirement information written in a natural language; a first generation step of generating prompts to be input into a first language model, the prompts being for extracting at least one of information on components, functions, and performance of the components from the acquired first requirements information; A control method for an information processing device comprising:

10. On the computer, an acquisition process for acquiring first requirement information written in a natural language; a first generation process for generating a prompt to be input to a first language model, the prompt being for extracting at least one of information on components, functions, and performance of the components from the acquired first requirements information; A program to execute.

Images