Report generation method and device, equipment, storage medium and program product

Abstract

The embodiment of the invention provides a report generation method and device, equipment, a storage medium and a program product, and relates to the field of artificial intelligence and financial science and technology. The method comprises the steps of obtaining a name and a target time range of a target object of a to-be-generated report; screening a historical report and a structured index table of the target object based on the name of the target object; on the basis of the target time range, public opinion information of the target object in the target time range is obtained; and calling the large model to generate a target report corresponding to the target object by using the historical report, the structured index table and the public opinion information according to a pre-constructed cue word. According to the method, the report, the structured indexes and the public opinion information are automatically integrated, the manual processing time is shortened, the report generation efficiency is improved, the real-time performance of the report is enhanced, user-defined cue words are supported, system logic does not need to be modified, and the report generation flexibility is improved; and by utilizing the knowledge traceability of the large model, the reference content in the report can be quickly positioned.

Description

Technical Field

[0001] This application relates to the fields of artificial intelligence and financial technology, and in particular to a report generation method, apparatus, device, storage medium, and program product. Background Technology

[0002] In the approval process, approval reports are primarily generated through structured data indicators, manually analyzed by professional staff, who then write and generate evaluation reports. The report generation steps include: staff uploading historical report documents, verifying and storing the specific indicator values ​​within the documents in a structured database. Based on the structured data, staff calculate supplementary indicators such as annual growth rate and overall growth rate according to business processing logic, storing these in the database for business personnel to query and use, and then generating the reports through numerical analysis.

[0003] In the existing process, the approval report involves hundreds of specific indicators and values. The acquisition and analysis of public opinion information largely rely on manual processing and report writing by staff, which has pain points such as low manual efficiency and inconsistent professional quality. Summary of the Invention

[0004] This application provides a report generation method, apparatus, device, storage medium, and program product to solve the technical problem of low efficiency caused by reliance on manual report generation.

[0005] Firstly, this application provides a report generation method, including:

[0006] Obtain the name of the target object and the target time range for the report to be generated;

[0007] Based on the name of the target object, filter the target object's historical reports and structured indicator tables;

[0008] Based on the target time range, obtain public opinion information about the target object within the target time range;

[0009] Based on pre-built prompts, historical reports, structured indicator tables, and public opinion information are used to call a large model to generate target reports corresponding to the target objects.

[0010] Secondly, this application provides a report generation apparatus, comprising:

[0011] The data acquisition module is used to obtain the name of the target object and the target time range for the report to be generated;

[0012] The data filtering module is used to filter historical reports and structured indicator tables of a target object based on the object's name.

[0013] The public opinion acquisition module is used to acquire public opinion information about a target object within a target time range.

[0014] The report generation module is used to generate target reports corresponding to the target objects by calling a large model based on pre-built prompts, historical reports, structured indicator tables, and public opinion information.

[0015] Thirdly, this application provides an electronic device, including: a processor and a memory communicatively connected to the processor;

[0016] The memory stores the instructions that the computer executes;

[0017] The processor executes computer-executable instructions stored in memory to implement any of the methods of the first aspect.

[0018] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method of any one of the first aspects.

[0019] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the method of any one of the first aspects.

[0020] The report generation method, apparatus, equipment, storage medium, and program products provided in this application obtain historical reports and structured indicator tables of the target object by filtering based on the name of the target object. Based on the target object's target time range, they acquire public opinion information of the target object within that time range. By calling a large model and using prompts, they generate a target report using historical reports, structured indicator tables, and public opinion information. This achieves automated integration of financial reports, structured indicators, and public opinion information, greatly reducing manual processing time and improving report generation efficiency. Real-time acquisition of public opinion information enhances the report's timeliness. Support for user-defined prompts for analysis dimensions eliminates the need to modify system logic, increasing the flexibility of report generation. The knowledge tracing capabilities of the large model enable rapid location of referenced content within the generated report. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0022] Figure 1 A diagram illustrating the generation of reports in the financial sector;

[0023] Figure 2 A flowchart illustrating a report generation method provided in an embodiment of this application;

[0024] Figure 3 This is a schematic diagram of the structure of a report generation device provided in an embodiment of this application;

[0025] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0026] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0028] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0029] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0030] It should be noted that the report generation method, apparatus, equipment, storage medium, and program products provided in this application can be used in the fields of artificial intelligence and fintech, as well as in any other field. The application fields of the report generation method, apparatus, equipment, storage medium, and program products in this application are not limited.

[0031] In the financial sector, financial institutions need to conduct in-depth analysis of financial reports submitted by enterprises and, in conjunction with public opinion dynamics, comprehensively evaluate these enterprises. Therefore, high demands are placed on the real-time processing of data in financial reports, the professionalism of analytical results, and the efficiency of report generation. However, traditional manual processing methods suffer from low efficiency and significant quality fluctuations, necessitating intelligent solutions.

[0032] The specific application scenarios of this application include, but are not limited to, report generation applications in the financial field. Figure 1 A diagram generated for reports in the financial sector, such as Figure 1 As shown, the report primarily relies on a combination of manual processing and a structured data system. Staff manually upload historical reports from previous years, and the system extracts key indicator values, verifies them, and stores them in a structured database. The system generates derived indicators based on preset rules (such as the annual growth rate calculation formula) for subsequent analysis. Publicly available online public opinion information (such as news, announcements, and industry reports) is periodically acquired in batches. However, this information is typically T-1 days old, meaning it reflects the previous day's public opinion and cannot be updated in real-time. Business personnel then retrieve structured indicator tables from the structured database, combine them with the public opinion information, manually select key indicators, and manually write the report.

[0033] Existing report generation methods have limitations, including: data processing, public opinion analysis, and report writing all require manual completion, which is inefficient and prone to errors. Public opinion data relies on periodic batch acquisition and cannot reflect the latest corporate dynamics in real time. Manual writing is susceptible to the influence of personnel's professional level, resulting in poor consistency of report content, uneven report quality, and a significant workload required for report generation, leading to low efficiency.

[0034] The report generation method, apparatus, equipment, storage medium, and program products provided in this application generate reports based on prompts using a large model and utilizing historical reports, structured indicator tables, and public opinion information of the target object, aiming to solve the aforementioned technical problems of the prior art.

[0035] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0036] Figure 2 This is a flowchart illustrating a report generation method provided in an embodiment of this application, as shown below. Figure 2 As shown, the method includes:

[0037] S201. Obtain the name of the target object and the target time range for the report to be generated.

[0038] In one example, within the financial sector, the target object can include enterprises, individuals, etc., and this application does not impose any restrictions on the target object. Taking an enterprise as an example, the name of the enterprise that needs to generate the report and the target time range of the report are obtained to facilitate the target object's approval of the generated report.

[0039] S202. Based on the name of the target object, filter the historical reports and structured indicator tables of the target object.

[0040] In one example, when the report is in the financial field, the structured indicator table may include structured key indicators and the corresponding calculation logic for these key indicators. Key indicators may include, but are not limited to, key indicators in the financial field; the selection of key indicators is not restricted in this application. Historical reports may include the target object's historical reports from previous years, and these historical reports contain a description of the target object's report outline format.

[0041] S203. Based on the target time range, obtain public opinion information about the target object within the target time range.

[0042] In one example, dynamic public opinion information of enterprises within the target time range of the target report can be obtained through a large model; alternatively, dynamic public opinion information of enterprises within the target time range of the target report can be obtained through real-time online search. In this application, there are no restrictions on the method of obtaining public opinion information of the target object within the target time range.

[0043] S204. Based on pre-built prompts, utilize historical reports, structured indicator tables, and public opinion information to call the large model to generate target reports corresponding to the target objects.

[0044] In one example, a large model is invoked, which, based on pre-built prompts, integrates historical reports, structured metrics, and public opinion information to generate a target report corresponding to the target object. This target report includes reference links, allowing business users to locate historical reports through these links for verification, editing, and revision of the target report.

[0045] In one implementation scenario, taking a company as the target and a financial report as the target report, a financial analysis agent is constructed to generate a financial report and send it to the business user. Specifically, constructing the financial analysis agent may include: Step 1. Defining the input of the agent, namely the company name and the report time range. Step 2. Defining the output of the agent, namely the viewpoints and numerical references from different dimensions such as operational risk, operational advantages, and liquidity. Step 3. Debugging and constructing the large model prompts, including but not limited to key indicators from different dimensions, numerical calculations, and professional risk analysis methods (trend judgment, industry comparison, cross-analysis of indicators, etc.; no restrictions are placed on risk analysis methods in this application), report outline format instructions, etc.

[0046] For example, generating financial reports using the constructed financial analysis agent described above and sending them to business users may include: 1) Obtaining the business user's input of the name of the company for which a financial report needs to be generated and the target time range for the report. 2) The financial analysis agent automatically retrieves the company's historical financial report documents and structured indicator tables from a pre-built database. 3) Through the financial analysis agent, calling a large-scale model to search for dynamic public opinion information about the company within the target time range in real time. 4) Through the financial analysis agent, calling a large-scale model based on pre-built prompts, integrating the original documents of historical financial reports, structured indicator tables, and dynamic public opinion information, the financial analysis agent returns the generated target report to the business user. The business user can then quickly locate historical financial reports through the reference links in the target report, and perform verification, editing, revision, and follow-up dialogue interactions. This achieves intelligent analysis of financial reports and generation of risk approval reports through a deep-thinking large-scale model. By fully utilizing the output of the Chain of Thought (CoT) of the deep thinking big model, as well as its comprehensive advantages in numerical calculation, financial professional knowledge, real-time online search and text generation, it can achieve numerical analysis of professional financial reports, real-time public opinion analysis and risk identification, improve the intelligence level of complex business scenarios in financial analysis, and help businesses improve work efficiency.

[0047] The report generation method provided in this embodiment obtains historical reports and structured indicator tables for the target object by filtering based on the target object's name. It then acquires public opinion information for the target object within the target time range. By calling a large model based on prompts, it generates a target report using historical reports, structured indicator tables, and public opinion information. This achieves automated integration of reports, structured indicators, and public opinion information, significantly reducing manual processing time and improving report generation efficiency. Real-time acquisition of public opinion information enhances the report's timeliness. Support for user-defined prompts for analysis dimensions eliminates the need to modify system logic, increasing the flexibility of report generation. Furthermore, the knowledge tracing capabilities of the large model enable rapid location of referenced content within the generated report.

[0048] Optionally, the method may also include: calling the calculation process and reference path of the data in the target report generated by the large model, for use in verifying the target report.

[0049] In one example, while generating the target report using a large model, the thought process and data reference path of the large model are simultaneously output to make the numerical calculation logic and analysis methods transparent.

[0050] By using the thought chain of the large model to output the calculation process and reference path of the data, the numerical calculation logic and risk analysis methods are made transparent, supporting business personnel to verify and interact with the generated results, so as to realize the verification of the target report and further improve the accuracy of the target report.

[0051] Optionally, the reference path can support jumping to historical reports and public opinion information; the method also includes: in response to the operation on the reference path, jumping to and displaying historical reports and public opinion information.

[0052] In one example, the reference path supports navigation to historical reports and public opinion information. An interactive verification interface is designed to allow staff to click on the reference path to navigate to historical financial reports or public opinion information pages, and it also supports follow-up question functionality. Verification can be performed by responding to staff clicking on values ​​or conclusions in the target report to navigate to historical financial reports or public opinion information pages.

[0053] For example, the system also supports follow-up questions (such as "Please re-analyze the data from a certain year"), and the large model dynamically adjusts the generated target report based on the new follow-up questions. The large model can be trained and adjusted using a reinforcement learning framework, enabling it to dynamically adjust its prompting strategy in response to user follow-up questions.

[0054] By supporting links to the original reports and public opinion information, business personnel can click on the links to jump to the original report or public opinion information page, thereby further reducing the verification cost of reports and improving the efficiency and accuracy of report generation.

[0055] Optionally, the method also includes: obtaining prompt word feedback data, adjusting the prompt words using the prompt word feedback data through a reinforcement learning framework to obtain adjusted prompt words; and using historical reports, structured indicator tables, and public opinion information to call a large model to generate a target report corresponding to the target object based on the adjusted prompt words.

[0056] In one example, financial expertise (such as industry comparison methods) can be embedded into the prompt word engineering to guide the large model in generating target reports that conform to financial industry standards. Through a dynamic prompt word optimization mechanism, the prompt word template is adjusted based on historical feedback data provided by the user (such as revision records and follow-up questions); based on the adjusted prompt words, the large model is invoked to generate the target report corresponding to the target object.

[0057] By using a dynamic prompt word optimization mechanism, prompt words are adjusted based on historical user feedback data to improve the responsiveness of the large model to business needs. By dynamically adjusting prompt words, the number of interactions between users and the large model is reduced, further shortening the report generation time and improving the efficiency of report generation.

[0058] Optionally, the large model also includes a multimodal large model; the method also includes: calling the multimodal large model to parse the unstructured information in historical reports and public opinion information, and combining it with a structured indicator table to generate a target report corresponding to the target object.

[0059] In one example, a multimodal large model is introduced to parse unstructured content such as charts and footnotes from historical reports and public opinion information. This is then cross-validated using structured indicators to transform the data into structured historical reports and public opinion information. Using this structured historical reports and public opinion information, along with a structured indicator table, the large model is invoked to generate the target report corresponding to the target object.

[0060] By using a multimodal large model to analyze unstructured information in reports and cross-validating it with structured indicator tables, the scope of content identified by the large model in the reports is expanded, the ability to analyze unstructured data in financial reports is improved, the risk of omissions during manual verification is reduced, the comprehensiveness of the analysis is enhanced, and the efficiency of report generation is further improved.

[0061] Optionally, the method also includes: acquiring specialized models for different fields, which are used to characterize models for analyzing historical reports or public opinion information; and, based on the large model, calling the corresponding specialized model according to the requirements of the target report.

[0062] In one example, a multi-model collaborative architecture can be constructed, consisting of specialized models from different domains, such as financial indicator calculation models and public opinion sentiment analysis models, etc., without limitation in this application. The large model is used to call the specialized models corresponding to the requirements of the target report as needed.

[0063] By combining specialized models from different fields into a multi-model collaborative architecture, large models can call specialized models on demand. The reports generated by large models are more tailored to user needs, improving the user experience. The multi-model collaborative architecture also enhances the processing efficiency of complex analysis tasks, reduces the computational load of a single model, and strengthens system stability.

[0064] Optionally, based on the target time range, obtain public opinion information of the target object within the target time range, including: based on the target time range, perform streaming analysis on the public opinion information of the target object within the target time range through a pre-integrated real-time data stream processing framework to obtain real-time public opinion information.

[0065] In one example, a real-time data stream processing framework is additionally integrated into the system. The specific architecture of the real-time data stream processing framework is not limited in this application. The framework performs streaming analysis on public opinion information of the enterprise within a target time range to obtain real-time public opinion information.

[0066] By using a pre-integrated real-time data stream processing framework, public opinion information is analyzed in a streaming manner, and risk assessment results are dynamically updated. This enables streaming processing of public opinion information, thereby improving the timeliness of generated reports to meet the dynamic needs of users.

[0067] Figure 3 This is a schematic diagram of the structure of a report generation device provided in an embodiment of this application, as shown below. Figure 3 As shown, the report generation device 30 provided in this embodiment includes:

[0068] Data acquisition module 301 is used to acquire the name of the target object and the target time range of the report to be generated;

[0069] Data filtering module 302 is used to filter historical reports and structured indicator tables of target objects based on the name of the target object;

[0070] The public opinion acquisition module 303 is used to acquire public opinion information of a target object within a target time range based on the target time range.

[0071] The report generation module 304 is used to generate a target report corresponding to the target object by calling a large model based on pre-built prompts, historical reports, structured indicator tables, and public opinion information.

[0072] In one possible implementation, the report generation device is also specifically used to: invoke the calculation process and reference path of the data in the target report from the large model, so as to verify the target report.

[0073] In one possible implementation, the reference path supports jumping to historical reports and public opinion information; the report generation device is also specifically used to: in response to an operation on the reference path, jump to and display historical reports and public opinion information.

[0074] In one possible implementation, the report generation device is further configured to: acquire prompt word feedback data; adjust the prompt words using the prompt word feedback data through a reinforcement learning framework to obtain adjusted prompt words; and, based on the adjusted prompt words, use historical reports, structured indicator tables, and public opinion information to call a large model to generate a target report corresponding to the target object.

[0075] In one possible implementation, the large model also includes a multimodal large model; the report generation device is specifically used to: call the multimodal large model to parse the unstructured information in historical reports and public opinion information, and combine it with a structured indicator table to generate a target report corresponding to the target object.

[0076] In one possible implementation, the report generation device is also specifically used to: acquire specialized models for different fields, which are used to characterize models for analyzing historical reports or public opinion information; and, based on the large model, call the corresponding specialized model according to the requirements of the target report.

[0077] In one possible implementation, the public opinion acquisition module 303 is specifically used to: based on the target time range, perform streaming analysis on the public opinion information of the target object within the target time range through a pre-integrated real-time data stream processing framework to obtain real-time public opinion information.

[0078] The report generation device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0079] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 40 may include a memory 401 and a processor 402. Optionally, the electronic device may also include a transceiver 403, wherein the memory 401 and the processor 402 communicate with each other; for example, the memory 401, the processor 402 and the transceiver 403 may communicate via a communication bus 404, the memory 401 is used to store a computer program, and the processor 402 executes the computer program to implement the method of the above embodiments.

[0080] Optionally, the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps in the method embodiments disclosed in this application can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0081] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the methods in any of the above method embodiments.

[0082] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the methods in any of the above method embodiments.

[0083] All or part of the steps in the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable memory. When the program is executed, it performs the steps of the above method embodiments; and the aforementioned memory (storage medium) includes: read-only memory (ROM), RAM, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof.

[0084] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processing unit of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

[0088] In this application, the term "comprising" and its variations can refer to non-limiting inclusion; the term "or" and its variations can refer to "and / or". The terms "first", "second", etc., in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. In this application, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0089] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0090] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0091] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.

[0092] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.

[0093] When integrated units / modules are implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor can be any suitable hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC, etc. Unless otherwise specified, the storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.

[0094] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0095] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0096] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0097] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A report generation method, characterized in that, The method includes: Obtain the name of the target object and the target time range for the report to be generated; Based on the name of the target object, filter the historical reports and structured indicator tables of the target object; Based on the target time range, obtain public opinion information about the target object within the target time range; Based on pre-constructed prompts, the historical reports, structured indicator tables, and public opinion information are used to call a large model to generate a target report corresponding to the target object.

2. The method according to claim 1, characterized in that, The method further includes: The calculation process and reference path for generating the data in the target report by calling the large model are used to verify the target report.

3. The method according to claim 2, characterized in that, The reference path supports jumping to the historical reports and the public opinion information; the method further includes: In response to the operation on the referenced path, the system redirects to and displays the historical report and the public opinion information.

4. The method according to claim 1, characterized in that, The method further includes: Obtain prompt word feedback data, and adjust the prompt words using the prompt word feedback data through a reinforcement learning framework to obtain the adjusted prompt words; Based on the adjusted prompts, the historical reports, the structured indicator table, and the public opinion information are used to call the large model to generate the target report corresponding to the target object.

5. The method according to claim 1, characterized in that, The large model also includes a multimodal large model; the method further includes: The multimodal big model is invoked to parse the unstructured information in the historical reports and public opinion information, and combined with the structured indicator table, a target report corresponding to the target object is generated.

6. The method according to claim 1, characterized in that, The method further includes: Obtain specialized models for different fields, which are used to characterize the models for analyzing the historical reports or public opinion information; Based on the large model, the corresponding special model is invoked according to the requirements of the target report.

7. The method according to claim 1, characterized in that, The step of obtaining public opinion information about the target object within the target time range includes: Based on the target time range, a pre-integrated real-time data stream processing framework is used to perform streaming analysis on the public opinion information of the target object within the target time range to obtain real-time public opinion information.

8. A report generation device, characterized in that, The device includes: The data acquisition module is used to obtain the name of the target object and the target time range for the report to be generated; The data filtering module is used to filter historical reports and structured indicator tables of the target object based on the name of the target object; The public opinion acquisition module is used to acquire public opinion information of the target object within the target time range, based on the target time range. The report generation module is used to generate a target report corresponding to the target object by calling a large model based on pre-built prompts, the historical reports, the structured indicator table, and the public opinion information.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images