Demand analysis method and device based on artificial intelligence
By employing an AI-based requirements analysis method and utilizing large language models and mind mapping tools, the requirements documents are processed automatically, solving the problems of low efficiency and insufficient accuracy in traditional software requirements analysis and achieving highly efficient requirements analysis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-30
- Publication Date
- 2026-04-10
AI Technical Summary
Traditional software requirements analysis relies on manual processing, which is inefficient, makes it difficult to accurately extract core requirements and analyze the relationships between requirements, resulting in information omissions and high complexity.
An AI-based requirements analysis approach is adopted, which displays a list of requirements information through a pre-set interactive interface, receives user analysis requests, calls large language models and mind mapping tools, automatically analyzes requirements documents, and generates requirements analysis results.
It improved the efficiency and accuracy of requirements analysis, reduced manual workload, and realized an automated requirements analysis process.
Smart Images
Figure CN121835618A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of artificial intelligence and demand management, and particularly relates to a demand analysis method and device based on artificial intelligence. BACKGROUND
[0002] In the process of software development and project management, demand analysis is a crucial link. In traditional practice, software demand analysis often relies on manual analysis, but manual analysis may have some problems. For example, the format of demand documents is various, and manual processing is inefficient. For another example, it is difficult to extract core demand content, which may cause information omission. For another example, the correlation analysis between demands is complex and relies on manual experience. Therefore, how to improve the efficiency and accuracy of demand analysis has become a problem to be solved in the industry. SUMMARY
[0003] In view of this, the embodiments of the present application provide a demand analysis method and device based on artificial intelligence to eliminate or improve one or more defects in the prior art.
[0004] According to a first aspect, a demand analysis method based on artificial intelligence is provided, comprising: displaying a demand information list through a preset interactive interface, wherein each piece of demand information in the demand information list comprises a demand document identifier; receiving a demand analysis request sent by a user based on the demand information list, wherein the demand analysis request comprises a target demand document identifier and a target analysis type, different analysis types correspond to different analysis tool sets, and each analysis type corresponds to an analysis tool set comprising a large language model; determining at least one demand document to be analyzed from a preset demand document library based on the target demand document identifier and the target analysis type, wherein the at least one demand document to be analyzed comprises a target demand document corresponding to the target demand document identifier; calling an analysis tool set corresponding to the target analysis type to analyze the at least one demand document to be analyzed, and obtaining a demand analysis result.
[0005] According to a second aspect, there is provided an artificial intelligence-based requirement analysis apparatus, comprising: a display unit configured to display a requirement information list via a preset interactive interface, wherein each piece of requirement information in the requirement information list comprises a requirement document identifier; a receiving unit configured to receive a requirement analysis request sent by a user based on the requirement information list, wherein the requirement analysis request comprises a target requirement document identifier and a target analysis type, different analysis types correspond to different analysis tool sets, and each analysis type corresponds to an analysis tool set comprising a large language model; a determining unit configured to determine at least one requirement document to be analyzed from a preset requirement document library based on the target requirement document identifier and the target analysis type, wherein the at least one requirement document to be analyzed comprises a target requirement document corresponding to the target requirement document identifier; and an analysis unit configured to call an analysis tool set corresponding to the target analysis type, and analyze the at least one requirement document to be analyzed to obtain a requirement analysis result.
[0006] According to a third aspect, there is provided a computer-readable storage medium having stored thereon computer programs / instructions, wherein the computer programs / instructions, when executed by a processor, implement the steps of the method according to any one of the first aspect.
[0007] According to a fourth aspect, there is provided a computing device comprising a processor, a memory, and computer programs / instructions stored on the memory, wherein the processor is configured to execute the computer programs / instructions, and the device implements the steps of the method according to any one of the first aspect when the computer programs / instructions are executed.
[0008] The artificial intelligence-based requirement analysis method and apparatus provided by the embodiments of the present specification first display a requirement information list via a preset interactive interface, wherein each piece of requirement information in the requirement information list can comprise a requirement document identifier. Then, a requirement analysis request sent by a user based on the requirement information list is received, wherein the requirement analysis request can comprise a target requirement document identifier and a target analysis type, different analysis types correspond to different analysis tool sets, and each analysis type corresponds to an analysis tool set comprising a large language model. Then, at least one requirement document to be analyzed is determined from a preset requirement document library based on the target requirement document identifier and the target analysis type, wherein the at least one requirement document to be analyzed can comprise a target requirement document corresponding to the target requirement document identifier. Finally, an analysis tool set corresponding to the target analysis type is called, and the at least one requirement document to be analyzed is analyzed to obtain a requirement analysis result. Thus, artificial intelligence-based automated software requirement analysis is achieved, the workload of manual work is reduced, and the efficiency and accuracy of requirement analysis are improved.
[0009] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.
[0010] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description
[0011] The accompanying drawings, which are provided to further illustrate the invention and form part of this application, are not intended to limit the scope of the invention.
[0012] Figure 1 This diagram illustrates an example of creating a requirements document library; Figure 2 A flowchart of an AI-based requirements analysis method according to one embodiment is shown; Figure 3 A schematic block diagram of an artificial intelligence-based demand analysis apparatus according to one embodiment is shown. Detailed Implementation
[0013] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.
[0014] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.
[0015] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.
[0016] It is understood that the ordinal numbers such as "first" and "second" mentioned in this specification are only used to distinguish multiple objects of the same or different categories (such as components, steps, parameters, etc.), and do not indicate the priority, importance or order relationship between objects, nor do they constitute a limitation on the technical features.
[0017] As mentioned earlier, improving the efficiency and accuracy of requirements analysis has become an urgent problem for the industry.
[0018] Therefore, embodiments of this specification provide an artificial intelligence-based requirements analysis method that can automatically analyze requirements documents in a requirements document library based on user requests.
[0019] In the embodiments of this specification, a requirements document library can be pre-built. This library can include requirements documents and their corresponding information, such as requirements document identifiers, requirements document summaries, and entry dates. During software development, the Software Requirements Document (SRD) is the core document, clearly defining the software's goals, functions, performance, constraints, and other relevant conditions. It serves as the formal agreement basis for communication between clients, project managers, designers, developers, and testers. Depending on actual needs, the requirements document may include project introductions, overall descriptions, functional requirements specifications, and other information.
[0020] In some examples, the requirements document library can store requirements documents along with their corresponding requirements document identifiers, software product line identifiers, and tags, etc.
[0021] Please see Figure 1 , Figure 1 This diagram illustrates an example of creating a requirements document library. Figure 1 As shown, a requirements document library can be built through the following steps 1) to 3), specifically: Step 1), receive the request to store the required document.
[0022] In this example, a user uploading a requirement document can do so through a pre-defined document upload interface. They upload the requirement document to be stored and enter its requirement document identifier and the software product line identifier to which it belongs. The system then receives a requirement document storage request from the user, which can include the requirement document to be stored, its corresponding requirement document identifier, and the software product line identifier. In this example, the requirement document identifier (ID) can be used to uniquely identify a requirement document; for example, it could be the requirement document name, number, etc.
[0023] Step 2), generate tags for the document to be stored.
[0024] In this example, the requirements document refers to a structured and formalized representation of business objectives and user needs at the system level, serving as the benchmark for subsequent design, development, testing, and acceptance. Tags for the requirements document to be stored can include business modules, functions, etc. Business modules can refer to different operational stages of a business process, such as the authentication process. Functions can refer to specific functions within the business, such as facial recognition. In this example, tags for the requirements document to be stored can be generated in various ways. For example, algorithms such as TF-IDF (Term Frequency–Inverse Document Frequency) and TextRank can be used to extract keywords and phrases from the requirements document as tags.
[0025] In some examples, step 2) above may specifically include the following: calling the large language model based on the requirement document to be stored, and generating tags for the requirement document to be stored from the large language model.
[0026] In this example, the content of the document to be stored and a preset prompt can be input into the large language model. The prompt can be used to guide the large language model to generate tags for the input document content. Thus, the large language model can generate tags for the document to be stored.
[0027] Step 3) Associate and store the requirement documents to be stored, along with their corresponding requirement document identifiers, software product line identifiers, and tags, into the requirement document library.
[0028] In this example, the requirement document to be stored, along with its corresponding requirement document identifier, software product line identifier, and tags, can be associated and stored in the requirement document library. This enables the creation of a requirement document library.
[0029] based on Figure 1 The example shown creates a requirements document library that can store multiple requirements documents. Based on this, AI-based requirements analysis methods can be further implemented.
[0030] Please continue reading Figure 2 , Figure 2 A flowchart illustrating an AI-based requirements analysis method according to one embodiment is shown. It will be understood that this method can be executed by any device, equipment, platform, or cluster of devices with computing and processing capabilities. Figure 2 As shown, this AI-based demand analysis method may include the following steps 201 to 204, specifically: Step 201: Display the list of requirements information through a preset interactive interface.
[0031] In this embodiment, a list of requirements information can be displayed through a preset interactive interface. Each requirement in the list may include a requirement document identifier. Here, the requirement information displayed in the list may be related information about requirement documents in a requirement document library, such as requirement document identifier, software product line identifier, tags, etc.
[0032] Step 202: Receive the requirement analysis request sent by the user based on the requirement information list.
[0033] In this embodiment, in addition to displaying a list of requirement information, the interactive interface can also display multiple analysis types for the user to choose from. Different analysis types can correspond to different analysis toolsets, and each analysis type's corresponding analysis toolset can include a large language model. Here, the analysis tools in the analysis toolset can be used to analyze requirement documents. Thus, the user can view the list of requirement information displayed on the interactive interface, select the target requirement information and target analysis type through the interface, and send a requirement analysis request. This requirement analysis request can include the target requirement document identifier and the target analysis type.
[0034] In some examples, the types of analysis performed on requirements documents may include ontology requirements analysis, related requirements analysis, and historical requirements analysis. Ontology requirements analysis can be used to analyze the target requirements document. Related requirements analysis can be used to analyze the target requirements document and its related requirements documents; here, the related requirements documents corresponding to the target requirements document can be requirements documents retrieved from a requirements document library that are related to the target requirements document. Historical requirements analysis can be used to analyze the target requirements document and its historical requirements documents within the software product line to which it belongs.
[0035] Step 203: Based on the target requirement document identifier and target analysis type, determine at least one requirement document to be analyzed from the preset requirement document library.
[0036] In this embodiment, the requirement documents to be analyzed are different for different analysis types. Here, at least one requirement document can be determined from the requirement document library as the requirement document to be analyzed based on the target requirement document identifier and the target analysis type.
[0037] Step 204: Invoke the analysis toolset corresponding to the target analysis type to analyze at least one requirement document to be analyzed and obtain the requirement analysis results.
[0038] In this embodiment, after identifying at least one requirement document to be analyzed, the analysis toolset corresponding to the target analysis type can be invoked to analyze the requirement document and obtain the requirement analysis results. It can be understood that different analysis types can correspond to different requirement analysis results.
[0039] In some examples, the analytics toolset corresponding to ontology requirements analysis may include a primary language model and a primary mind map generation tool. Here, the primary language model can be various large language models, including but not limited to the GPT series (Generative Pre-trained Transformer), Gemini series, Wenxin Yiyan, Hunyuan, Doubao, etc. The primary mind map generation tool can be various mind map generation tools, including but not limited to MindManager, MindNode, iMindMap, MindMeister, etc.
[0040] When the target analysis type is ontology requirement analysis, step 203 above, which involves determining at least one requirement document to be analyzed from a pre-set requirement document library based on the target requirement document identifier and the target analysis type, can specifically include the following steps A1): Step A1) Based on the target requirement document identifier and ontology requirement analysis, identify the target requirement document as the requirement document to be analyzed from the requirement document library.
[0041] Furthermore, step 204 above, which involves calling the analysis toolset corresponding to the target analysis type to analyze at least one requirement document to obtain the requirement analysis results, can specifically include the following steps A2) and A3). Specifically: Step A2) Construct the first prompt word based on the information in the target requirement document, input the first prompt word into the first large language model, and output the first analysis result from the first large language model.
[0042] Step A3), call the first mind map generation tool to generate a mind map corresponding to the target requirement document.
[0043] In this example, when the target analysis type is ontology requirement analysis, the target requirement document corresponding to the target requirement document identifier can be retrieved from the requirement document library as the requirement document to be analyzed. Then, the target requirement document can be analyzed based on a large language model. Specifically, a first prompt word can be constructed based on the information of the target requirement document (e.g., the content of the target requirement document). This first prompt word can be used to guide the large language model to analyze the target requirement document, for example, guiding the large language model to output information such as the core content, keywords, importance, feasibility, and scope of impact of the target requirement document. For example, a first prompt word template can be pre-constructed (e.g., manually written), and placeholders can be used to fill in the information in the first prompt word template. In this way, the placeholders in the prompt word template can be replaced with the information of the target requirement document, thus obtaining the first prompt word. After constructing the first prompt word, it can be input into the large language model, which will output a first analysis result. The first analysis result can include at least one of the following: core content, keywords, importance, feasibility, scope of impact, etc.
[0044] Furthermore, the first mind map generation tool can be used to generate a mind map corresponding to the target requirement document. This allows the first analysis results output by the first language model to be combined with the mind map generated by the first mind map generation tool. Figure 1 The results of the requirements analysis, presented as part of the target requirements document, are displayed to the user for their review.
[0045] In some examples, the analytical toolset corresponding to the requirements analysis may include a second major language model and a second mind map generation tool. Here, the second major language model may be the same as or different from the first major language model. Similarly, the second mind map generation tool may be the same as or different from the first mind map generation tool.
[0046] When the target analysis type is related requirement analysis, step 203 above, which involves determining at least one requirement document to be analyzed from a pre-set requirement document library based on the target requirement document identifier and the target analysis type, can specifically include the following steps B1) and B2): Step B1) Based on the target requirement document identifier, obtain the target tag corresponding to the target requirement document from the requirement document library.
[0047] Step B2) Based on the target tag, identify at least one associated requirement document across product lines that is related to the target requirement document from the requirement document library, and use the target requirement document and at least one associated requirement document as the requirement documents to be analyzed.
[0048] Furthermore, step 204 above, which involves calling the analysis toolset corresponding to the target analysis type to analyze at least one requirement document to obtain the requirement analysis results, can specifically include the following steps B3) and B4), specifically: Step B3): Construct a second prompt word based on the information in the target requirement document and at least one related requirement document. Input the second prompt word into the second language model, which then outputs a second analysis result. The second analysis result may include the analysis results corresponding to the target requirement document and each related requirement document, as well as the relationship information between each related requirement document and the target requirement document.
[0049] Step B4) Use the second mind map generation tool to generate mind maps for the target requirement document and each related requirement document.
[0050] In this example, when the target analysis type is related requirement analysis, firstly, based on the target requirement document identifier, the target requirement document corresponding to the target requirement document identifier can be retrieved from the requirement document library, and the tags corresponding to the target requirement document can be obtained from the requirement document library as target tags. Then, based on the target tags, at least one related requirement document across product lines associated with the target requirement document can be identified from the requirement document library. For example, the target tags can be matched with the tags of each requirement document in the requirement document library (e.g., similarity calculation), and the requirement document corresponding to the tag that matches the target tag (e.g., similarity exceeds a preset threshold) can be identified as a related requirement document of the target requirement document. Here, the target requirement document and the related requirement document can be requirement documents from different product lines; that is, the software product line identifier of the related requirement document can be the same as or different from the software product line identifier of the target requirement document. In this example, the target requirement document and at least one related requirement document can be used as the requirement documents to be analyzed. Then, the requirement documents to be analyzed can be performed based on a large language model. Specifically, a second prompt word can be constructed based on information from the target requirement document (e.g., its content) and information from at least one related requirement document (e.g., its content). This second prompt word can guide a second language model to analyze the target requirement document and at least one related requirement document. For example, it can guide the second language model to output information such as the core content, keywords, importance, feasibility, and scope of impact of the target requirement document and each related requirement document. It can also guide the second language model to output information about the relationship between the target requirement document and each related requirement document. For example, the relationship information can be in numerical form, with higher values indicating a closer relationship. For instance, a second prompt word template can be pre-constructed (e.g., manually written), where placeholders can be used to fill in the information. Then, the placeholders in the prompt word template can be replaced with information from the target requirement document and each related requirement document, thus obtaining the second prompt word. After constructing the second prompt words, they can be input into the second language model. The second language model will then output the second analysis results, which may include the analysis results corresponding to the target requirement document and each related requirement document. Each analysis result may include at least one of the following: core content, keywords, importance, feasibility, scope of impact, etc. Furthermore, the second analysis results may also include information on the relationship between each related requirement document and the target requirement document.
[0051] Furthermore, a second mind map generation tool can be used to generate mind maps corresponding to the target requirement document and each related requirement document. This allows the second analysis results output by the second language model to be combined with the mind maps generated by the second mind map generation tool. Figure 1The results of the requirements analysis, including the target requirements document and related requirements documents, are presented to the user for viewing.
[0052] In some examples, the analytical toolset corresponding to historical requirements analysis may include a third major language model and a third mind map generation tool. Here, the third major language model may be the same as or different from the first major language model. Similarly, the third mind map generation tool may be the same as or different from the first mind map generation tool.
[0053] When the target analysis type is historical requirements analysis, step 203 above, which involves determining at least one requirement document to be analyzed from a pre-set requirement document library based on the target requirement document identifier and the target analysis type, can specifically include the following steps C1) and C2): Step C1) Based on the target requirement document identifier, obtain the target tag and target software product line identifier corresponding to the target requirement document from the requirement document library.
[0054] Step C2), based on the target tag and the target software product line identifier, identify at least one historical requirement document from the requirement document library that is associated with the target requirement document and is part of the same product line, and use the target requirement document and at least one historical requirement document as the requirement documents to be analyzed.
[0055] Furthermore, step 204 above, which involves calling the analysis toolset corresponding to the target analysis type to analyze at least one requirement document to obtain the requirement analysis results, may specifically include the following steps C3) and C4): Step C3) Constructs a third prompt word based on the information in the target requirement document and at least one historical requirement document. This third prompt word is then input into a third language model, which outputs a third analysis result. This third analysis result may include the analysis results corresponding to the target requirement document and each historical requirement document, as well as the changes made to the target requirement document relative to each historical requirement document.
[0056] Step C4) Call the third mind mapping tool to generate mind maps corresponding to the target requirement document and each historical requirement document.
[0057] In this example, when the target analysis type is historical requirements analysis, firstly, based on the target requirements document identifier, the target requirements document corresponding to the target requirements document identifier can be retrieved from the requirements document library. The tags corresponding to the target requirements document are then obtained from the requirements document library as target tags, and the software product line identifier corresponding to the target requirements document is obtained as the target software product line identifier. Next, based on the target tags and the target software product line identifier, at least one historical requirements document associated with the target requirements document within the same product line can be identified from the requirements document library. For example, based on the product line identifier, multiple requirements documents within the same product line as the target requirements document can be identified from the requirements document library. The target tags are then matched with the tags of each of the multiple requirements documents (e.g., similarity calculation). The requirements document corresponding to the tag that matches the target tag (e.g., similarity exceeds a preset threshold) is identified as the historical requirements document associated with the target requirements document. Here, the target requirements document and the historical requirements document can be requirements documents within the same product line; that is, the software product line identifier of the historical requirements document can be the same as the software product line identifier of the target requirements document. In this example, the target requirements document and at least one historical requirements document can be used as the requirements documents to be analyzed. Then, the large language model can be used to analyze the requirement document to be analyzed. Specifically, a third cue word can be constructed based on information from the target requirement document (e.g., its content) and information from at least one historical requirement document (e.g., its content). This third cue word can guide the large language model to analyze the target requirement document and at least one historical requirement document. For example, it can guide the large language model to output information such as the core content, keywords, importance, feasibility, and scope of impact of the target requirement document and each historical requirement document, and also guide the large language model to output the changes in the target requirement document relative to each historical requirement document. For example, a third cue word template can be pre-constructed (e.g., manually written), and placeholders can be used for the information to be filled in the third cue word template. In this way, the placeholders in the cue word template can be replaced with the information from the target requirement document and each historical requirement document, thereby obtaining the third cue word. After constructing the third prompt words, they can be input into a third language model, which will output third analysis results. These third analysis results can include analysis results corresponding to the target requirement document and various historical requirement documents. Each analysis result can include at least one of the following: core content, keywords, importance, feasibility, scope of impact, etc. Furthermore, the third analysis results can also include changes made to the target requirement document compared to historical requirement documents.
[0058] Furthermore, a third-party mind mapping tool can be used to generate mind maps corresponding to the target requirement document and each historical requirement document. This allows the third analysis results output by the third language model to be combined with the mind maps generated by the third-party mind mapping tool. Figure 1 The results of the requirements analysis, including the target requirements document and various historical requirements documents, are presented to the user for viewing.
[0059] As an example, in some implementations, the AI-based requirements analysis method of this embodiment can be executed through an AI (Artificial Intelligence) requirements analysis platform. This platform may include a front-end, back-end, database system, AI service module, etc. The front-end can be based on various frameworks, such as React.js, Vue.js, Next.js, etc., and can provide a user-friendly interface. The back-end can be based on various frameworks, such as Java Spring Boot, Micronaut, Dropwizard, etc., and can handle business logic and data storage. The database system can use databases such as MySQL, PostgreSQL, MariaDB, etc., to store user information, requirements documents, and analysis results. The database system can store multiple tables, such as a user table (users) to store basic user information and permissions; a project table (projects) to manage project information; a requirements table (requirements) to store requirements documents and basic information; a requirements analysis results table (requirement_analyses) to store AI analysis results; and a requirements relation table (requirement_relations) to manage the relationships between requirements. The historical requirements analysis table (historical_requirement_analyses) can be used to store historical analysis data. The document template table (document_templates) can be used to manage document templates.
[0060] The requirements analysis platform offers a variety of functionalities, including user management (registration and login, role-based access control, and user information management), requirements document management (uploading multiple document formats such as Word, PDF, and Markdown, displaying and searching requirements lists, categorizing and tagging requirements, version control, summarizing project requirements, and managing requirement relationships), AI analysis capabilities (extracting core requirements content, intelligently generating requirement mind maps, extracting keywords, assessing requirement importance, analyzing requirement relationships, intelligently interpreting historical requirements, analyzing requirement feasibility, identifying change points, and intelligently assessing the scope of impact), and multiple AI model selection and configuration. Specifically, it supports selecting different AI models based on requirement type and analysis objectives, including general language models and domain-specific models. Furthermore, it enables historical requirements document analysis, automatically comparing and analyzing the differences between the target requirements document and historical requirements documents. AI analytics can also support feasibility analysis and decision support. Specifically, based on historical requirement execution data and technical implementation experience, it intelligently assesses the technical feasibility, implementation difficulty, and resource requirements of new requirements. AI analytics can also perform impact point and change point analysis. Specifically, it automatically identifies the scope of impact of new requirements on the existing system, the functional modules that need to be modified, and the data structures. The requirements analysis platform also provides requirements extraction and editing functions, such as intelligent extraction of key requirements content, collaborative editing support, intelligent generation of modification plans, and automatic generation of change impact analysis reports. The requirements analysis platform also provides document output functions, such as support for multi-format (e.g., Word, PDF, Markdown, HTML, and Xmind) document output, custom document templates, and automatic generation of analysis reports.
[0061] In summary, the requirements analysis platform can perform the following functions: text preprocessing (converting formats, cleaning, and standardizing uploaded requirements documents); core content extraction (using natural language processing to identify core functionalities, constraints, and acceptance criteria); keyword extraction (using algorithms such as TF-IDF and TextRank to extract keywords and phrases from requirements documents); mind map generation (building hierarchical mind map structures based on extracted core content and keywords); requirements correlation analysis (calculating similarity between requirements and establishing relationships); historical requirements comparison (comparing target requirements with historical requirements to identify change points and scope of impact); multi-model selection and configuration (providing a model selection interface where users can choose appropriate AI models based on requirement type, analysis objectives, and accuracy requirements; the system supports custom configuration of model parameters); and historical requirements knowledge base construction (structuring and indexing historical requirements documents, establishing a mapping relationship between requirements, functions, and modules, recording the implementation status, problems, and solutions of historical requirements). The feasibility analysis decision-making process involves retrieving the implementation status of similar requirements from a historical knowledge base, analyzing technological dependencies and constraints, assessing resource requirements and implementation risks, and generating feasibility analysis reports and recommendations. Impact point and change point analysis identifies the system modules and functionalities involved in the requirements, analyzes potential changes to data structures and interfaces, assesses the impact of changes on existing functionality, and generates detailed modification recommendations and risk warnings.
[0062] According to another embodiment, an artificial intelligence-based demand analysis device is provided. This artificial intelligence-based demand analysis device can be deployed in any device, platform, or device cluster with computing and processing capabilities.
[0063] Figure 3 A schematic block diagram of an artificial intelligence-based demand analysis apparatus according to one embodiment is shown. Figure 3 As shown, the AI-based demand analysis device 300 includes: The display unit 301 is used to display a list of requirements information through a preset interactive interface, wherein each requirement information in the list of requirements information includes a requirement document identifier. The receiving unit 302 is used to receive a requirement analysis request sent by the user based on the above requirement information list. The requirement analysis request includes a target requirement document identifier and a target analysis type. Different analysis types correspond to different analysis toolsets. The analysis toolsets corresponding to each analysis type include a large language model. The determining unit 303 is used to determine at least one requirement document to be analyzed from a preset requirement document library based on the target requirement document identifier and the target analysis type, wherein the at least one requirement document to be analyzed includes the target requirement document corresponding to the target requirement document identifier. Analysis unit 304 is used to call the analysis toolset corresponding to the above target analysis type to analyze at least one of the above requirements documents to be analyzed and obtain the requirements analysis results.
[0064] According to another embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed in a computer, causes the computer to perform... Figure 2 The method described.
[0065] According to another embodiment, a computing device is also provided, including a memory and a processor, characterized in that the memory stores executable code, and when the processor executes the executable code, it implements... Figure 2 The method described.
[0066] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.
[0067] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.
[0068] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.
[0069] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A demand analysis method based on artificial intelligence, comprising: A list of requirements information is displayed through a preset interactive interface, wherein each requirement information in the list includes a requirement document identifier. Receive a requirement analysis request sent by a user based on the requirement information list, wherein the requirement analysis request includes a target requirement document identifier and a target analysis type, different analysis types correspond to different analysis toolsets, and the analysis toolsets corresponding to each analysis type include a large language model; Based on the target requirement document identifier and the target analysis type, at least one requirement document to be analyzed is determined from a preset requirement document library, wherein the at least one requirement document to be analyzed includes the target requirement document corresponding to the target requirement document identifier; The analysis toolset corresponding to the target analysis type is invoked to analyze the at least one requirement document to be analyzed, and the requirement analysis results are obtained.
2. The method according to claim 1, characterized in that, The analysis types include ontology requirement analysis, related requirement analysis, and historical requirement analysis. Ontology requirement analysis is used to analyze the target requirement document, related requirement analysis is used to analyze the target requirement document and its related requirement documents, and historical requirement analysis is used to analyze the target requirement document and its historical requirement documents for the software product line.
3. The method according to claim 2, characterized in that, The target analysis type is ontology requirement analysis, and the corresponding analysis toolset includes a first large language model and a first mind map generation tool; and, based on the target requirement document identifier and the target analysis type, determining at least one requirement document to be analyzed from a preset requirement document library includes: Based on the target requirement document identifier and ontology requirement analysis, the target requirement document is identified as the requirement document to be analyzed from the requirement document library; And, the step of invoking the analysis toolset corresponding to the target analysis type to analyze the at least one requirement document to be analyzed and obtain the requirement analysis results includes: Based on the information in the target requirement document, a first prompt word is constructed, and the first prompt word is input into the first large language model. The first large language model outputs a first analysis result, wherein the first analysis result includes at least one of the following: core content, keywords, importance, feasibility, and scope of influence. The first mind map generation tool is invoked to generate a mind map corresponding to the target requirement document.
4. The method according to claim 2, characterized in that, The requirement document library stores requirement documents and their corresponding requirement document identifiers, software product line identifiers, and tags.
5. The method according to claim 4, characterized in that, The target analysis type is related requirement analysis, and the analysis toolset corresponding to related requirement analysis includes a second large language model and a second mind map generation tool; and, based on the target requirement document identifier and the target analysis type, determining at least one requirement document to be analyzed from a preset requirement document library includes: Based on the target requirement document identifier, obtain the target tag corresponding to the target requirement document from the requirement document library; Based on the target tag, at least one associated requirement document across product lines and related to the target requirement document is identified from the requirement document library, and the target requirement document and the at least one associated requirement document are used as requirement documents to be analyzed. And, the step of invoking the analysis toolset corresponding to the target analysis type to analyze the at least one requirement document to be analyzed and obtain the requirement analysis results includes: Based on the information of the target requirement document and the information of at least one related requirement document, a second prompt word is constructed, and the second prompt word is input into the second large language model. The second large language model outputs a second analysis result, wherein the second analysis result includes the analysis results corresponding to the target requirement document and each related requirement document, as well as the association relationship information between each related requirement document and the target requirement document. The second mind map generation tool is invoked to generate mind maps corresponding to the target requirement document and each related requirement document.
6. The method according to claim 4, characterized in that, The target analysis type is historical requirements analysis, and the corresponding analysis toolset for historical requirements analysis includes the third major language model and the third mind map generation tool; and, based on the target requirements document identifier and the target analysis type, determining at least one requirements document to be analyzed from a preset requirements document library includes: Based on the target requirement document identifier, obtain the target tag and target software product line identifier corresponding to the target requirement document from the requirement document library; Based on the target tag and the target software product line identifier, at least one historical requirement document associated with the target requirement document from the requirement document library is identified, and the target requirement document and the at least one historical requirement document are used as requirement documents to be analyzed. And, the step of invoking the analysis toolset corresponding to the target analysis type to analyze the at least one requirement document to be analyzed and obtain the requirement analysis results includes: Based on the information of the target requirement document and the information of at least one historical requirement document, a third prompt word is constructed, and the third prompt word is input into the third language model. The third language model outputs a third analysis result, wherein the third analysis result includes the analysis results corresponding to the target requirement document and each historical requirement document, as well as the changes made to the target requirement document relative to each historical requirement document. The third mind map generation tool is invoked to generate mind maps corresponding to the target requirement document and each historical requirement document.
7. The method according to claim 1, characterized in that, The requirements document library was constructed in the following way: Receive a requirement document storage request, wherein the requirement document storage request includes the requirement document to be stored and its corresponding requirement document identifier and software product line identifier; Generate tags for the document to be stored; The required documents to be stored, along with their corresponding required document identifiers, software product line identifiers, and tags, are associated and stored in the required document library.
8. The method according to claim 7, characterized in that, The tags used to generate the document to be stored include: Based on the document to be stored, the large language model is invoked, and the large language model generates tags for the document to be stored.
9. A computing device, comprising a processor, a memory, and computer programs / instructions stored in the memory, characterized in that, The processor is configured to execute the computer program / instructions, and when the computer program / instructions are executed, the device implements the steps of the method as described in any one of claims 1 to 8.
10. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method as described in any one of claims 1 to 8.