An industry analysis report generation method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application disclose an industry analysis report generation method and device, electronic equipment and a storage medium. The method comprises: determining one or more core element keywords based on core elements of user analysis requirements, and determining target analysis perspective information based on each core element keyword and an industry knowledge graph; obtaining report generation prompt words based on the target analysis perspective information, and calling a large model to generate an industry analysis report text based on the report generation prompt words; and performing perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtaining a target industry analysis report by performing perspective consistency verification and information integrity verification on the industry analysis report text. Embodiments of the present application can efficiently generate an industry analysis report that meets business core requirements.

Description

Technical Field

[0001] This invention relates to the field of large model technology, and in particular to a method, apparatus, electronic device and storage medium for generating industry analysis reports. Background Technology

[0002] The standard workflow for automatically generating industry analysis reports based on payment transaction data using existing technologies is as follows: 1. Input industry news events; 2. Extract analytical perspectives based on news, such as "assessing the flow of funds for technological upgrading in the manufacturing industry"; 3. Call the payment transaction data interface to obtain indicators such as industry capital flow and policy subsidies; 4. Iteratively generate the analysis report, specifically: initial draft → data supplementation → revised draft.

[0003] However, the aforementioned process in existing technologies is prone to perspective deviation during iteration. For example, the initial perspective might be the efficiency of manufacturing technology upgrade funds, followed by a draft analysis of equipment procurement cash flow, and then, after supplementing with consumer payment data, the final draft's perspective shifts to consumer behavior analysis. This can cause the analysis report's conclusions to deviate from core business needs, rendering the report ineffective and unable to support specific decisions; manual correction, on the other hand, leads to efficiency losses. Summary of the Invention

[0004] This invention provides a method for generating industry analysis reports, which can efficiently generate industry analysis reports that meet core business needs.

[0005] In a first aspect, embodiments of the present invention provide a method for generating an industry analysis report, comprising: Based on the core elements of user analysis needs, one or more core element keywords are determined, and target analysis perspective information is determined based on each core element keyword and industry knowledge graph. Based on the objective analysis perspective, information is used to obtain report generation prompts, and based on these prompts, a large model is invoked to generate the industry analysis report text; and Based on the target analysis perspective information, perform perspective consistency verification and information integrity verification on the industry analysis report text, and obtain the target industry analysis report by obtaining the industry analysis report text that passes the perspective consistency verification and information integrity verification.

[0006] Secondly, embodiments of the present invention provide an industry analysis report generation apparatus, comprising: The analysis perspective acquisition module is used to determine one or more core element keywords based on the core elements of the user's analysis needs, and to determine the target analysis perspective information based on each core element keyword and the industry knowledge graph. The report text generation module is used to obtain report generation prompts based on information from the target analysis perspective, and then use these prompts to call a large model to generate industry analysis report text; and The report text verification module is used to perform perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtain the target industry analysis report by obtaining the industry analysis report text that has passed the perspective consistency verification and information integrity verification.

[0007] Thirdly, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the industry analysis report generation method as described in any of the embodiments of the present invention.

[0008] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the industry analysis report generation method as described in any of the embodiments of the present invention.

[0009] This invention provides a method, apparatus, electronic device, and storage medium for generating industry analysis reports. It determines one or more core element keywords based on the core elements of user analysis needs, and determines target analysis perspective information based on these keywords and an industry knowledge graph. Further, it obtains report generation prompts based on the target analysis perspective information, calls a large model to generate the industry analysis report text based on these prompts, and performs perspective consistency and information integrity checks on the industry analysis report text based on the target analysis perspective information. Finally, it obtains the target industry analysis report by acquiring the industry analysis report text that passes both the perspective consistency and information integrity checks. This ensures that the process of generating the industry analysis report using the large model remains focused on core business needs, avoiding perspective deviation. Consequently, it efficiently generates industry analysis reports that meet business requirements without manual correction, ensuring the effectiveness of the report results and the scientific nature of the corresponding decisions. Attached Figure Description

[0010] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating the industry analysis report generation method provided in this embodiment of the invention; Figure 2 This is another flowchart illustrating the industry analysis report generation method provided in this embodiment of the invention; Figure 3 This is another flowchart illustrating the industry analysis report generation method provided in this embodiment of the invention; Figure 4 This is another flowchart illustrating the industry analysis report generation method provided in this embodiment of the invention; Figure 5 This is a schematic diagram of the industry analysis report generation device provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0012] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0013] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0014] Figure 1 This is a flowchart illustrating an industry analysis report generation method provided in an embodiment of the present invention. This method can be executed by an industry analysis report generation device provided in this embodiment, which can be implemented using software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a computer or server. The following embodiments will illustrate this using the integration of the device into an electronic device as an example. (Reference) Figure 1 The method may specifically include the following steps: Step 101 involves identifying one or more core element keywords based on the core elements of user analysis needs, and determining the target analysis perspective information based on each core element keyword and the industry knowledge graph. This step facilitates the acquisition of prompts for report generation constrained by the target analysis perspective information, as well as the generation of the target industry analysis report based on the large model constrained by the target analysis perspective information, thus avoiding perspective deviations at various stages of report generation.

[0015] Optional core elements include: time element, geographic element, industry element and / or analysis dimension element.

[0016] Specifically, the aforementioned analytical dimensions can include both the analytical object elements and the analytical target elements.

[0017] In a specific example, the content text of the user analysis requirement is "Analysis of the effectiveness of technological upgrading funds in the Yangtze River Delta manufacturing industry in 2025". The process of determining one or more core element keywords based on the core elements of the user analysis requirement includes: using a large model to extract the time element, regional element, industry element, analysis object element, and analysis target element of the user analysis requirement from the content text of the user analysis requirement, which corresponds to "time element keyword - 2025", "regional element keyword - Yangtze River Delta", "industry element keyword - manufacturing industry", "analysis object element keyword - technological upgrading funds", and "analysis target element keyword - fund effectiveness".

[0018] Specifically, the aforementioned industry knowledge graph can be understood as a knowledge network that uses entities (such as geographical distribution and potential report analysis dimensions) as nodes and the relationships between entities as edges. By extracting, integrating, cleaning, and modeling scattered unstructured and semi-structured information within the industry, it forms a knowledge network that can intuitively reflect the core elements, hierarchical structure, and internal logic of the industry. It can support accurate information retrieval and in-depth correlation analysis in scenarios such as industry analysis, decision support, intelligent question answering, and risk warning.

[0019] Specifically, the aforementioned industry knowledge graph can include core elements for analyzing different industries, and use the sub-dimensions of each core element as the corresponding core element's associated nodes.

[0020] In a specific example, the process of determining the target analysis perspective information based on the keywords of each core element and the industry knowledge graph includes: querying the industry knowledge graph to obtain the sub-dimensional elements of each core element, and using the sub-dimensional elements of each core element to determine the target analysis perspective information.

[0021] Specifically, the process of determining the target analysis perspective by utilizing the sub-dimensional elements of each core element can include: using the sub-dimensional elements to compose target analysis perspective information.

[0022] Optionally, after step 101, the industry analysis report generation method provided in this embodiment of the invention further includes: encrypting the target analysis perspective information to obtain an encrypted feature field; adding metadata, extracting hash digests and signatures to the encrypted feature field to obtain a signature field; and assembling the encrypted feature field, metadata and signature field to obtain an encrypted perspective token.

[0023] Specifically, the process of encrypting the target analysis perspective information to obtain the encrypted feature field may include: encrypting the target analysis perspective information using the AES-256 algorithm.

[0024] Specifically, other encryption algorithms, such as the SM4 algorithm, can also be used to encrypt the target's perspective information.

[0025] Specifically, the aforementioned metadata may include token identifier and token generation time, etc.

[0026] Specifically, the process of adding metadata to the encrypted feature field, extracting the hash digest, and signing to obtain the signature field may include: adding metadata to the encrypted feature field to obtain the encrypted data to be signed, using the SHA-256 algorithm to extract the hash digest of the encrypted data to be signed, and using the RSA-2048 private key to sign the hash digest to generate the signature field.

[0027] Specifically, other algorithms can also be used, such as SHA3-256, to extract the hash digest of the encrypted data to be signed.

[0028] Specifically, an ECCP-256 private key can be used to sign the hash digest to generate a signature field.

[0029] In a specific instance, the aforementioned cryptographic view token can be represented as: { "token_id": "token-20251001-001", "timestamp": "2025-10-01 10:00:30.123", "encrypted_features": "U2FsdGVkX1+9y09Zx78n...", / / Ciphertext of feature vectors after AES-256 encryption "signature": "MIICdgIBADANBgkqhkiG9w0BAQEFAASCAmAwggJcAgEAAoGBAO...", / / RSA-2048 signature value }

[0030] It is understandable that target analysis perspective information may be accidentally tampered with when it flows between modules or systems or is stored in media due to possible external attacks, network interception, server intrusion, unauthorized internal modifications, manual tampering by system operators without authorization, and data packet loss due to network fluctuations during transmission. Therefore, encrypting the target analysis perspective information and generating corresponding encrypted perspective tokens can prevent perspective deviations in subsequent report generation processes caused by accidental tampering of the target analysis perspective information.

[0031] Step 102 involves obtaining report generation prompts based on the target analysis perspective and then using these prompts to call the large model to generate the industry analysis report text. This step helps ensure that the report generation prompts are anchored to the target analysis perspective, thereby ensuring that the generated industry analysis report text does not deviate from core business needs.

[0032] Specifically, the process of obtaining report generation prompts based on target analysis perspective information may include: retrieving relevant industry information, relevant policy information, and relevant indicator data related to the user's analysis needs based on target analysis perspective information, and adding the relevant industry information, relevant policy information, and relevant indicator data related to the user's analysis needs, along with the target analysis perspective information, to the report generation prompts.

[0033] Specifically, the process of obtaining report generation prompts based on target analysis perspective information may also include: obtaining and adding URLs containing relevant industry information, relevant policy information, and relevant indicator data that meet user analysis needs to the report generation prompts, so that large models with network connectivity can access the corresponding URLs to extract relevant industry information, relevant policy information, and relevant indicator data when generating the report based on the report generation prompts.

[0034] Optionally, during step 102, target analysis perspective information is extracted from the encrypted perspective token in real time. Step 102 includes generating prompt words based on the target analysis perspective information extracted from the encrypted perspective token in real time.

[0035] Specifically, the process of extracting target analysis perspective information from the encrypted perspective token in real time may include: using the RSA-2048 public key corresponding to the aforementioned RSA-2048 private key to decrypt the signature field of the encrypted perspective token to obtain a decrypted hash digest; adding metadata from the encrypted perspective token to the encrypted feature field in the encrypted perspective token, and using the SHA-256 algorithm to extract a hash digest from the encrypted feature field after adding metadata to obtain a verification hash digest; comparing whether the decrypted hash digest and the verification hash digest are consistent, if they are consistent, it means that the signature is valid, and at this time, the encrypted feature field is decrypted using the AES-256 algorithm to obtain the target analysis perspective information.

[0036] Step 103 involves performing perspective consistency and information integrity checks on the industry analysis report text based on the target analysis perspective information, and obtaining the target industry analysis report text that has passed both checks. This step, by performing perspective consistency and information integrity checks on the industry analysis report text based on the target analysis perspective, combined with steps 101 and 102, can efficiently generate an industry analysis report that meets business needs without manual correction, thereby ensuring the validity of the report results and the scientific nature of the corresponding decisions.

[0037] Specifically, the process of verifying the perspective consistency of the industry analysis report text based on the target analysis perspective information may include: extracting the perspective information of the industry analysis report text, and verifying the perspective consistency of the industry analysis report text by comparing the perspective information of the industry analysis report text with the target analysis perspective information.

[0038] Specifically, it can also generate consistency verification prompts based on target analysis perspective information and industry analysis report text, and call the large model based on the consistency verification prompts to perform perspective consistency verification on the industry analysis report text.

[0039] Specifically, the process of verifying the information integrity of the industry analysis report text based on the target analysis perspective can include: verifying the information integrity of the industry analysis report text based on pre-determined information integrity verification items.

[0040] Optionally, when performing step 103, the industry analysis report generation method provided in this embodiment of the invention further includes: extracting target analysis perspective information from the encrypted perspective token in real time. Step 103 includes using the target analysis perspective information extracted from the encrypted perspective token in real time to perform perspective consistency verification and information integrity verification on the industry analysis report text.

[0041] The following further describes the industry analysis report generation method provided by the embodiments of the present invention, such as... Figure 2 As shown, that is Figure 1 Step 101 may include the following steps: Step 1011: Determine the analysis time range based on time element keywords.

[0042] In a specific example, the process of determining the analysis time range based on time element keywords includes: determining the analysis time range as "the whole year of 2025" based on the time element keyword "2025".

[0043] Step 1012: Determine the analysis area based on regional element keywords.

[0044] In a specific example, the process of determining the analysis area based on regional element keywords includes: determining the analysis area to include "Shanghai", "Jiangsu", "Zhejiang" and "Anhui" based on the regional element keyword "Yangtze River Delta".

[0045] Step 1013: Obtain sub-dimensional related words of industry element keywords from the industry knowledge graph, and determine the scope of the industry analysis based on the sub-dimensional related words of industry element keywords.

[0046] Specifically, the aforementioned industry knowledge graph may include different industry keywords and core analysis dimensions for analyzing different industries, and include sub-dimensional keywords of each industry keyword as related nodes of the corresponding industry keywords, and sub-dimensional keywords of each core analysis dimension as related nodes of the corresponding core analysis dimension.

[0047] In a specific example, the process of obtaining sub-dimensional related words of industry element keywords from the industry knowledge graph and determining the scope of the industry analysis based on the sub-dimensional related words of industry element keywords includes: obtaining the sub-dimensional keywords of "high-end equipment manufacturing", "new energy vehicles", "electronic information" and "biomedicine" of "manufacturing" from the industry knowledge graph based on the industry element keyword "manufacturing", and determining the scope of the industry analysis to include: "high-end equipment manufacturing", "new energy vehicles", "electronic information" and "biomedicine".

[0048] Step 1014: Obtain sub-dimension related words of the analysis dimension keywords from the industry knowledge graph, and determine the core analysis dimension based on the sub-dimension related words of the analysis dimension keywords.

[0049] In a specific example, the process of retrieving sub-dimension related terms of the analysis dimension keywords from the industry knowledge graph and determining the core analysis dimension based on these sub-dimension related terms can include: retrieving sub-dimension related terms of the analysis dimension keyword "technological upgrading funds" from the industry knowledge graph, such as "central government subsidies," "local matching funds," "enterprise's own funds," and "proportion and cost difference of loans from financial institutions"; and retrieving sub-dimension related terms of the analysis dimension keyword "fund efficiency" from the industry knowledge graph, such as "output growth driven by unit technological upgrading funds," "tax contribution," "number of new jobs," "equipment replacement rate driven by technological upgrading funds," "R&D investment intensity," "number of patents produced," "fund idle rate," "project failure rate," "compliance of subsidy fund use," and determining the core analysis dimension to include "central government subsidies," "local matching funds," "enterprise's own funds," and "proportion and cost difference of loans from financial institutions." “Increase in output value driven by unit technological upgrading funds”, “Tax contribution”, “Number of new jobs created”, “Equipment upgrade rate driven by technological upgrading funds”, “R&D investment intensity”, “Number of patents produced”, “Fund idle rate”, “Project failure rate”, “Compliance of subsidy funds”, “Fund idle rate”, “Project failure rate”, “Compliance of subsidy funds”.

[0050] Step 1015: Obtain target analysis perspective information, including the analysis time range, analysis geographical range, analysis industry range, and core analysis dimensions.

[0051] Specifically, the encoded features of the content text, including the analysis time range, analysis geographical range, analysis industry range, and core analysis dimensions, can be extracted and used as target analysis perspective information.

[0052] This invention determines the analysis time range and analysis geographical range based on time element keywords and geographical element keywords, and obtains the analysis industry range and core analysis dimensions from the industry knowledge graph based on industry element keywords and analysis dimension keywords. This enables the determination of a complete and refined analysis perspective that meets the user's analysis needs, which in turn facilitates anchoring a complete and refined analysis perspective in the subsequent report generation process, generating an industry analysis report that meets the user's needs.

[0053] The following further describes the industry analysis report generation method provided by embodiments of the present invention. For example... Figure 3 As shown, that is Figure 1 Step 102 may include the following steps: Step 1021: Based on the target analysis perspective, search for relevant industry information and relevant policy information related to the user's analysis needs through a search engine.

[0054] Specifically, when performing step 1021, target analysis perspective information can be extracted from the encrypted perspective token in real time. Step 1021 specifically includes using a search engine to search for relevant industry information and relevant policy information related to the user's analysis needs based on the target analysis perspective information extracted from the encrypted perspective token in real time.

[0055] Specifically, the process of obtaining relevant industry information and relevant policy information for user analysis needs through search engines based on target analysis perspective information may include: generating multiple search questions based on target analysis perspective information, and calling an intelligent search engine to search for and obtain the aforementioned relevant industry information and relevant policy information based on the search questions.

[0056] In a specific example, the above-mentioned multiple search questions could include: "2025 Yangtze River Delta Manufacturing Technology Upgrading Fiscal Subsidy Policy" and "2025 Anhui Biomedical Industry Technology Upgrading Subsidy Policy Interpretation".

[0057] Step 1022: Based on the target analysis perspective information, query the payment transaction data interface to obtain the relevant payment indicator data for the user's analysis needs.

[0058] Specifically, the aforementioned payment transaction data interface can be understood as an interface opened to the public by payment institutions, clearing platforms, or industry data service providers to provide standardized payment transaction statistics and analysis data to authorized entities. It is based on a unified data protocol and access control mechanism, and integrates multi-dimensional underlying data such as transaction details, merchant operations, and user consumption.

[0059] Specifically, when executing step 1022, target analysis perspective information can be extracted from the encrypted perspective token in real time. Step 1022 specifically includes querying and obtaining the relevant payment indicator data of the user's analysis needs through the payment transaction data interface based on the real-time extracted target analysis perspective information.

[0060] Specifically, the process of obtaining relevant payment indicator data for user analysis needs by querying payment transaction data interface based on target analysis perspective information may include: selecting relevant payment indicator data for user analysis needs from all payment transaction data indicators based on target analysis perspective information. For example, the indicator data corresponding to "Local matching fund payment scale for technological upgrading of high-end equipment manufacturing enterprises in Shanghai in 2025" and "Comparison of technological upgrading fund payment scale in the four major manufacturing sub-sectors of the Yangtze River Delta in 2025" can be selected.

[0061] Step 1023: Generate the prompt word skeleton of the model prompt words based on the target analysis perspective information. The prompt word skeleton includes: industry information field, policy information field and payment indicator data field.

[0062] Specifically, target analysis perspective information can be extracted from the encrypted perspective token in real time during step 1023. Step 1023 specifically includes generating the prompt word skeleton of the model prompt word based on the real-time extracted target analysis perspective information.

[0063] Specifically, the aforementioned prompt word skeleton may also include: analysis scope field and core analysis dimension field.

[0064] Specifically, the aforementioned prompt word skeleton may also include: analysis target field, constraint field, and output requirement field.

[0065] In a specific example, the analysis scope fields include: time range field, geographical range field, and industry range field; the core analysis dimension fields include: funding source dimension field and funding efficiency dimension field.

[0066] Specifically, the process of generating the prompt word skeleton of the model prompt words based on the target analysis perspective information can also include: generating the field content of the time range field, the geographical range field, the industry range field, and the core analysis dimension field based on the content text of the analysis time range, analysis geographical range, analysis industry range, and core analysis dimension fields mentioned above.

[0067] Specifically, the content of the fields required for the above output can be preset.

[0068] Specifically, the content of the target fields mentioned above can be obtained based on the text of the user's analysis requirements, as well as industry information fields, policy information fields, and payment indicator data fields.

[0069] Specifically, the content of the above constraint fields can be obtained based on the analysis time range, analysis geographical range, analysis industry range, and core analysis dimensions.

[0070] In a specific example, the prompt word skeleton is shown in Table 1: Step 1024: Based on the relevant industry information, relevant policy information, and relevant payment indicator data, generate the field content of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton, respectively, and obtain the report generation prompt words.

[0071] Specifically, the process of generating prompt words for the report based on the aforementioned industry information, policy information, and payment indicator data to generate prompt words may include: adding the content text of the industry information to the industry information field, adding the content text of the policy information to the policy information field, and adding the related payment indicator data to the payment indicator data field.

[0072] Specifically, different weights can be added to different information in related industry information, different weights to information in related policy information, and different weights to different data in related payment indicator data, and then written into the corresponding fields.

[0073] Optionally, before step 1024, the similarity between the content text and the prompt word skeleton based on the user's analysis needs is calculated, and a perspective consistency check is performed on the prompt word skeleton.

[0074] Specifically, the process of calculating the similarity between the content text of the user analysis needs and the prompt word skeleton, and performing perspective consistency verification on the prompt word skeleton, may include: using a text / multimodal embedding model to calculate the cosine similarity between the content text of the user analysis needs and the prompt word skeleton to obtain the first cosine similarity; determining whether the first cosine similarity is not less than the first similarity threshold; if it is not less than the threshold, the prompt word skeleton is determined to have passed the perspective consistency verification; otherwise, the prompt word skeleton is determined to have failed the perspective consistency verification.

[0075] Specifically, the aforementioned first similarity threshold can be determined based on empirical data or the results of multiple trials regarding the first similarity threshold, for example, it can be determined as any value within [0.7, 0.9].

[0076] In a specific example, the cosine similarity between the prompt word skeleton shown in Table 1 and the user analysis requirement, namely "efficiency of technological transformation funds for manufacturing industry in the Yangtze River Delta in 2025", is calculated to obtain the similarity between the content text of the user analysis requirement and the prompt word skeleton.

[0077] Optionally, step 1024 includes: when the prompt word skeleton passes the perspective consistency check, starting to execute the generation of the field content of the industry information field, policy information field and payment indicator data field in the prompt word skeleton based on the associated industry information, associated policy information and associated payment indicator data respectively.

[0078] Optionally, if the prompt word skeleton fails the perspective consistency check, restart the execution of searching for relevant industry information and relevant policy information based on the target analysis perspective information through the search engine until the prompt word skeleton passes the perspective consistency check.

[0079] It should be noted that existing business intelligence agents and text generation platforms often force models to fill in content from predetermined perspectives such as "contract risk" and "customer preference" by using predefined report templates, static prompts, or fixed paragraphs. These methods primarily rely on manually written template frameworks and prompt definitions. After system initialization, the template structure and prompt design are essentially fixed. Since the template content and prompts cannot be dynamically adjusted, existing technologies struggle to cover newly emerging perspective segmentation needs when data changes or analysis objectives are fine-tuned. This invention, by utilizing target analysis perspective information to generate report prompts, can dynamically adapt to changes in report data and adjustments to analysis objectives, better supporting multi-dimensional and refined industry analysis scenarios. Furthermore, this invention ensures that prompts are anchored to the original perspective by performing perspective consistency checks on the prompt skeleton. In addition, this invention, by using target analysis perspective information to constrain the data retrieval process, avoids acquiring payment transaction data unrelated to the core analysis objectives, thereby preventing deviations in the report's analysis direction.

[0080] The following further illustrates the industry analysis report generation method provided by the embodiments of the present invention, such as... Figure 4 As shown, that is Figure 1 Step 103 may include the following steps: Step 1031: Calculate and verify the consistency of perspective between the content text and the industry analysis report text based on the target analysis perspective information.

[0081] Specifically, the process of calculating and verifying the perspective consistency of the industry analysis report text based on the similarity between the content text of the target analysis perspective information and the industry analysis report text may include: using a text / multimodal embedding model to calculate the cosine similarity between the content text of the target analysis perspective information and the industry analysis report text to obtain a second cosine similarity; determining whether the second cosine similarity is not less than the second similarity threshold; if it is not less than the threshold, the industry analysis report text is determined to have passed the perspective consistency verification; otherwise, the industry analysis report text is determined to have failed the perspective consistency verification.

[0082] Specifically, the aforementioned second similarity threshold can be determined based on empirical data or the results of multiple trials regarding the second similarity threshold, for example, it can be determined as any value within [0.6, 0.8].

[0083] Step 1032: Obtain the industry analysis report text that has passed the perspective consistency check to obtain the initial draft of the industry analysis report, and generate integrity check prompts based on the target analysis perspective information and the initial draft of the industry analysis report.

[0084] Optionally, the process of obtaining the initial draft of the industry analysis report by acquiring the industry analysis report text that has passed the perspective consistency check may include: when the industry analysis report text passes the perspective consistency check, using the industry analysis report text as the initial draft of the industry analysis report; when the industry analysis report text fails the perspective consistency check, restarting the execution of the report generation prompt words and subsequent steps based on the target analysis perspective information until the industry analysis report text passes the perspective consistency check.

[0085] Specifically, the integrity verification prompts mentioned above may include: verification target field, target analysis perspective information field, industry analysis report draft field, and output format field.

[0086] Specifically, the content of the target field and the output format field can be preset. The process of generating integrity verification prompts based on the target analysis perspective information and the draft industry analysis report can include: adding the target analysis perspective information to the target analysis perspective information field and adding it to the draft industry analysis report field.

[0087] In a specific example, the integrity verification prompts are shown in Table 2: In Table 2, “need_more_info”: boolean indicates whether more information is needed. If yes, it outputs “True”; otherwise, it outputs “False”.

[0088] Step 1033: Based on the integrity verification prompt, call the large model to perform information integrity verification on the draft of the industry analysis report.

[0089] Specifically, the system can call a large model based on the integrity verification prompt to perform information integrity verification on the initial draft of the industry analysis report. If the output of the large model is "False", the initial draft of the industry analysis report is considered to have passed the information integrity verification; otherwise, if the output of the large model is "True", the initial draft of the industry analysis report is considered to have failed the information integrity verification.

[0090] Step 1034: Obtain the initial draft of the industry analysis report that has passed the information integrity verification to obtain the target industry analysis report.

[0091] Specifically, the process of obtaining the target industry analysis report by acquiring the initial draft of the industry analysis report that has passed the information integrity verification may include: when the initial draft of the industry analysis report passes the information integrity verification, acquiring the initial draft of the industry analysis report as the target industry analysis report; when the initial draft of the industry analysis report fails the information integrity verification, restarting the process of acquiring the report generation prompt based on the target analysis perspective information until the initial draft of the industry analysis report passes the information integrity verification.

[0092] This invention, through perspective consistency and information integrity checks on the report text generated by the text generation model, can further ensure that the final generated target industry analysis report deviates from the required perspective and ensure the information integrity of the report, thereby ensuring the validity of the report results and the scientific nature of the corresponding decisions.

[0093] Figure 5 This is a structural diagram of an industry analysis report generation apparatus provided in an embodiment of the present invention. This apparatus is suitable for executing the industry analysis report generation method provided in an embodiment of the present invention. Figure 5 As shown, the device may specifically include: The analysis perspective acquisition module 501 is used to determine one or more core element keywords based on the core elements of the user's analysis needs, and to determine target analysis perspective information based on each core element keyword and the industry knowledge graph. This module can facilitate the generation of industry analysis report text by acquiring report generation prompts based on the target analysis perspective information, and to perform perspective consistency verification and information integrity verification on the industry analysis report text.

[0094] Optionally, core element keywords include: time element keywords, geographical element keywords, industry element keywords, and analysis dimension keywords.

[0095] Optionally, the aforementioned analysis perspective acquisition module 501 can be specifically used to: determine the analysis time range based on time element keywords; determine the analysis geographical range based on geographical element keywords; query the industry knowledge graph to obtain sub-dimension related words of industry element keywords, and determine the analysis industry range based on the sub-dimension related words of industry element keywords; query the industry knowledge graph to obtain sub-dimension related words of analysis dimension keywords, and determine the core analysis dimension based on the sub-dimension related words of analysis dimension keywords; and acquire target analysis perspective information including analysis time range, analysis geographical range, analysis industry range, and core analysis dimension.

[0096] The report text generation module 502 is used to obtain report generation prompts based on the target analysis perspective information, and then use these prompts to call a large model to generate the industry analysis report text. This module helps ensure that the report generation prompts are anchored to the target analysis perspective, thereby ensuring that the generated industry analysis report text does not deviate from core business needs.

[0097] Optionally, the aforementioned report text generation module 502 can be specifically used to: search for relevant industry information and relevant policy information related to the user's analysis needs through a search engine based on the target analysis perspective information; query relevant payment indicator data related to the user's analysis needs through a payment transaction data interface based on the target analysis perspective information; generate a prompt word skeleton for model prompt words based on the target analysis perspective information, the prompt word skeleton including: industry information field, policy information field, and payment indicator data field; and generate the field content of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton based on the relevant industry information, relevant policy information, and relevant payment indicator data respectively, to obtain the report generation prompt words.

[0098] The report text verification module 503 is used to perform perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtain the target industry analysis report by acquiring the industry analysis report text that has passed the perspective consistency verification and information integrity verification. This module, together with modules 101 and 102, can efficiently generate industry analysis reports that meet business needs without manual correction, thereby ensuring the validity of the report results and the scientific nature of the corresponding decisions.

[0099] Optionally, the aforementioned report text verification module 503 can be specifically used to: calculate and verify the similarity between the content text based on the target analysis perspective information and the industry analysis report text; obtain the industry analysis report text that passes the perspective consistency verification to obtain a draft of the industry analysis report; generate integrity verification prompts based on the target analysis perspective information and the draft of the industry analysis report; call the large model based on the integrity verification prompts to perform information integrity verification on the draft of the industry analysis report; and obtain the target industry analysis report by obtaining the draft of the industry analysis report that passes the information integrity verification.

[0100] Optionally, the industry analysis report generation device provided in this embodiment of the invention further includes an encrypted token generation module, which is used to encrypt the target analysis perspective information to obtain an encrypted feature field after determining the target analysis perspective information based on the keywords of each core element and the industry knowledge graph; add metadata, extract hash digest and sign to the encrypted feature field to obtain a signature field; and assemble the encrypted feature field, metadata and signature field to obtain an encrypted perspective token.

[0101] Optionally, the industry analysis report generation device provided in this embodiment of the invention further includes: a token information extraction module, which extracts the target analysis perspective information from the encrypted perspective token in real time when obtaining report generation prompt words based on the target analysis perspective information; and extracts the target analysis perspective information from the encrypted perspective token in real time when performing perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information.

[0102] Optionally, the industry analysis report generation device provided in this embodiment of the invention further includes a prompt word skeleton verification module, which is used to calculate and verify the perspective consistency of the prompt word skeleton based on the similarity between the content text of the user's analysis needs and the prompt word skeleton before generating the prompt word based on the corresponding field content of the industry information field, policy information field and payment indicator data field in the prompt word skeleton based on the related industry information, related policy information and related payment indicator data.

[0103] Optionally, the aforementioned report text generation module 502 can be specifically used to, when the prompt word skeleton passes the perspective consistency check, initiate the execution of generating the field content of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton based on the associated industry information, associated policy information, and associated payment indicator data, respectively.

[0104] Optionally, the aforementioned report text generation module 502 can also be specifically used to restart the execution of searching for relevant industry information and relevant policy information based on the target analysis perspective information through a search engine when the prompt word skeleton fails the perspective consistency check, until the prompt word skeleton passes the perspective consistency check.

[0105] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is merely an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the functional modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0106] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the industry analysis report generation method provided in any of the above embodiments.

[0107] This invention also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the industry analysis report generation method provided in any of the above embodiments.

[0108] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the industry analysis report generation method as described in any of the embodiments of this invention.

[0109] The following is for reference. Figure 6It shows a schematic diagram of the structure of a computer system 600 suitable for implementing an electronic device according to embodiments of the present invention. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0110] like Figure 6 As shown, the computer system 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 602 or programs loaded from storage section 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the system 600. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0111] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed into storage section 608 as needed.

[0112] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by central processing unit (CPU) 601, it performs the functions defined above in the system of this invention.

[0113] It should be noted that the computer-readable medium shown in this invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0114] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0115] The modules and / or units described in the embodiments of the present invention can be implemented in software or hardware. The described modules and / or units can also be housed in a processor; for example, a processor may be described as including an analysis perspective acquisition module, a report text generation module, and a report text verification module. The names of these modules do not necessarily constitute a limitation on the module itself.

[0116] In another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist independently and not assembled into the device. The computer-readable medium carries one or more programs that, when executed by the device, cause the device to: determine one or more core element keywords based on the core elements of the user's analysis needs, and determine target analysis perspective information based on each core element keyword and an industry knowledge graph; obtain report generation prompts based on the target analysis perspective information, and call a large model based on the report generation prompts to generate an industry analysis report text; and perform perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtain the target industry analysis report by obtaining the industry analysis report text that has passed the perspective consistency verification and information integrity verification.

[0117] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating an industry analysis report, characterized in that, include: Based on the core elements of user analysis needs, one or more core element keywords are determined, and target analysis perspective information is determined based on each core element keyword and industry knowledge graph. Based on the target analysis perspective, information is obtained to generate report prompts, and based on the report generation prompts, a large model is called to generate industry analysis report text; as well as Based on the target analysis perspective information, perform perspective consistency verification and information integrity verification on the industry analysis report text, and obtain the target industry analysis report by obtaining the industry analysis report text that passes the perspective consistency verification and information integrity verification.

2. The method for generating an industry analysis report according to claim 1, characterized in that, The core element keywords include: time element keywords, geographical element keywords, industry element keywords, and analysis dimension keywords; The information used to determine the target analysis perspective based on keywords of each core element and industry knowledge graph includes: Determine the analysis time range based on time element keywords; The analysis area is determined based on regional element keywords; The sub-dimensional related words of industry element keywords are obtained by querying the industry knowledge graph, and the scope of the industry analysis is determined based on the sub-dimensional related words of industry element keywords. The core analytical dimensions are determined based on the sub-dimensional related terms of the analytical dimension keywords obtained from the industry knowledge graph query; and... Obtain the target analysis perspective information, including the analysis time range, the analysis geographical range, the analysis industry range, and the core analysis dimensions.

3. The method for generating an industry analysis report according to claim 2, characterized in that, After determining the target analysis perspective information based on keywords of each core element and industry knowledge graph, the method further includes: The target analysis perspective information is encrypted to obtain an encrypted feature field; The signature field is obtained by adding metadata to the encrypted feature field, extracting the hash digest, and signing; and The encrypted feature field, the metadata, and the signature field are assembled to obtain an encrypted perspective token; The method further includes: When generating prompts based on the target analysis perspective information, the target analysis perspective information is extracted from the encrypted perspective token in real time; and When performing perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, the target analysis perspective information is extracted from the encrypted perspective token in real time.

4. The method for generating an industry analysis report according to claim 1, characterized in that, The prompt words generated in the report based on target analysis perspective information acquisition include: Based on the perspective of target analysis, relevant industry information and policy information related to user analysis needs are obtained through search engines; Based on the target analysis perspective, the relevant payment indicator data for user analysis needs is obtained by querying the payment transaction data interface; The prompt word skeleton is generated based on the target analysis perspective information generation model. The prompt word skeleton includes: industry information fields, policy information fields, and payment indicator data fields; and Based on the related industry information, the related policy information, and the related payment indicator data, the field contents of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton are generated respectively, thus obtaining the report generation prompt word.

5. The method for generating an industry analysis report according to claim 4, characterized in that, Before generating the report generation prompt by generating the field contents of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton based on the related industry information, the related policy information, and the related payment indicator data, respectively, the method further includes: Calculate and verify the perspective consistency of the prompt word skeleton based on the similarity between the content text of the user analysis needs and the prompt word skeleton; The process involves generating the report generation prompts by generating the field contents of the industry information field, policy information field, and payment indicator data field in the prompt keyword skeleton based on the relevant industry information, relevant policy information, and relevant payment indicator data, respectively. When the prompt word skeleton passes the perspective consistency check, the process of generating the field contents of the industry information field, policy information field, and payment indicator data field in the prompt word skeleton based on the associated industry information, the associated policy information, and the associated payment indicator data is initiated.

6. The method for generating an industry analysis report according to claim 5, characterized in that, Also includes: If the prompt word skeleton fails the perspective consistency check, the process of searching for relevant industry information and relevant policy information based on the target analysis perspective information through the search engine to obtain the user's analysis needs will be restarted until the prompt word skeleton passes the perspective consistency check.

7. The method for generating an industry analysis report according to claim 1, characterized in that, The process of performing perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtaining the target industry analysis report text that has passed the perspective consistency verification and information integrity verification, includes: Calculate and verify the consistency of perspective between the content text and the industry analysis report text based on the target analysis perspective information; Obtain the initial draft of the industry analysis report by acquiring the text of the industry analysis report that has passed the perspective consistency check, and generate integrity check prompts based on the target analysis perspective information and the initial draft of the industry analysis report; Based on integrity verification prompts, a large model is invoked to perform information integrity verification on the initial draft of the industry analysis report; and The target industry analysis report is obtained by obtaining the initial draft of the industry analysis report that has passed the information integrity verification.

8. An industry analysis report generation device, characterized in that, include: The analysis perspective acquisition module is used to determine one or more core element keywords based on the core elements of the user's analysis needs, and to determine the target analysis perspective information based on each core element keyword and the industry knowledge graph. The report text generation module is used to obtain report generation prompts based on target analysis perspective information, and to call the large model based on the report generation prompts to generate industry analysis report text; as well as The report text verification module is used to perform perspective consistency verification and information integrity verification on the industry analysis report text based on the target analysis perspective information, and obtain the target industry analysis report by obtaining the industry analysis report text that has passed the perspective consistency verification and information integrity verification.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the industry analysis report generation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the industry analysis report generation method as described in any one of claims 1 to 7.

Images