System, method and device for realizing intelligent generation processing of industry research report based on large model and causal atlas, processor and storage medium

Abstract

The invention relates to a system for intelligently generating and processing industry research reports based on a large model and a causal atlas, which comprises a data extraction module, a data arrangement module and a research report generation module, and is characterized in that the data extraction module obtains viewpoints in research report fragments, event data and causal relationships among events; the data arrangement module is used for further cleaning and structuring the viewpoint, event and causal relationship data obtained from the data extraction module; and the research report generation module converts the processed data into a structured research report. By adopting the system, the method, the device, the processor and the computer readable storage medium for realizing the intelligent generation processing of the industry research report based on the large model and the causal map, the causal relationship map of the industry event is constructed, the map is used for enhancing the model generation content, and the traceability and the interpretability of the model generation content are enhanced; the time for industry research personnel to output industry research reports is shortened, and the working efficiency of related personnel is improved to a great extent.

Description

Technical Field

[0001] The present invention relates to the field of big data and artificial intelligence technology, and in particular to the field of data processing and analysis, and specifically refers to a system, method, device, processor and computer-readable storage medium thereof for intelligent generation and processing of industry research reports based on a large model and a causal graph. Background Art

[0002] In today's information age, with the rapid development of big data and artificial intelligence technology, data processing and analysis have become particularly critical. Especially in the financial and investment fields, the processing and generation of research reports (research reports) are important tools for obtaining market insights and guiding investment decisions. This technical background mainly discusses the application of large language models, data extraction technology and causal graphs in the field of research report processing and generation.

[0003] Large language models have been widely used in the field of natural language processing. These models can understand and generate complex text content through deep learning technology. In the application of research report generation, large language models can automatically write report drafts, provide market trend analysis, data interpretation and other content, greatly improving the efficiency and quality of report production. However, the text directly generated by the large language model has problems such as hallucinations and untraceability, and it is necessary to enhance the quality of the generated content through data extraction, graph retrieval and other methods.

[0004] Data extraction technology is used to identify valuable information from a large amount of unstructured and semi-structured data and convert it into a format that is easier to analyze and report. In research report processing, data extraction technology can be used to extract key data from a variety of financial documents and online resources, such as stock price changes, company financial report data, market trading dynamics, etc. After extraction and processing, these data can be directly used in the data analysis part of the research report to ensure the accuracy and timeliness of the information.

[0005] Causal event graph technology helps analyze the influence and interaction between events by building a causal event relationship model. In the generation of research reports, by building a causal graph, the causal relationship between various economic indicators, policy changes and market performance can be analyzed more deeply. This in-depth analysis helps research report writers provide more insightful market forecasts and suggestions, and provides investors with more scientific decision-making support.

[0006] At present, the technology of using large language models to assist in the generation of research reports is not yet mature and has poor usability, making it difficult for investment researchers to trust the results generated by the model. In addition, the financial sector is experiencing large industry changes and the causal relationships between events are complex, which can also cause problems such as poor accuracy and low value in the research reports generated by the model. This has hindered the implementation of the application of large language models to assist in the generation of industry research reports. Summary of the invention

[0007] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and to provide a system, method, device, processor and computer-readable storage medium thereof for intelligent generation and processing of industry research reports based on large models and causal graphs, which have high traceability, high interpretability and a wide range of applications.

[0008] In order to achieve the above objectives, the system, method, device, processor and computer-readable storage medium thereof for intelligent generation and processing of industry research reports based on a large model and a causal graph of the present invention are as follows:

[0009] The system for intelligently generating and processing industry research reports based on large models and causal graphs has the following main features: the system includes a data extraction module, a data sorting module and a research report generation module; the data sorting module is connected to the data extraction module, and the research report generation module is connected to the data sorting module; the data extraction module is used to obtain viewpoints, event data and causal relationships between events in the research report segment; the data sorting module is used to further clean and structure the viewpoints, events and causal relationship data obtained from the data extraction module; and the research report generation module is used to convert the processed data into a structured research report.

[0010] Preferably, the data extraction module includes an industry opinion extraction unit, an industry event extraction unit and an industry event causal relationship extraction unit, the industry event extraction unit is connected to the industry opinion extraction unit, and the industry event causal relationship extraction unit is connected to the industry opinion extraction unit and the industry event extraction unit, the industry opinion extraction unit is used to divide the complete research report into segments of fixed length, annotate the industry opinion in small batch data segments through the opinion extraction model, and infer the industry opinion extraction results; the industry event extraction unit is used to annotate the industry events contained in small batch opinion data through the event extraction model, and infer the industry event extraction results; the industry event causal relationship extraction unit is used to combine the research report segments and the events extracted from the segments, and use a large language model to extract the causal relationship between events; and obtain the opinions, event data and causal relationships between events in the research report segments.

[0011] Preferably, the data sorting module includes a data verification and cleaning unit, a data normalization unit and a data association unit, the data normalization unit is connected to the data verification and cleaning unit, and the data association unit is connected to the data normalization unit. The data verification and cleaning unit is used to confirm the accuracy and completeness of the data, automatically detect and correct errors and inconsistencies in the data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information; the data normalization unit is used to normalize the cleaned data and standardize the data; the data association unit is used to record the views and events extracted from the original text segment, and at the same time find the views and original text segments containing the event according to a specific event; and link related events and views in different data segments.

[0012] Preferably, the research report generation module selects the industry and time range for generating the industry research report, searches the database for events that match the selected industry within the time period, and counts several events with the highest frequency of occurrence in the selected time and selected industry as core events to generate the research report.

[0013] The method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph using the above system is mainly characterized in that the method comprises the following steps:

[0014] (1) Obtain the opinions, event data, and causal relationships between events in the research report fragments;

[0015] (2) Further cleaning and structuring of the opinions, events, and causal relationship data obtained from the data extraction module;

[0016] (3) Convert the processed data into a structured research report.

[0017] Preferably, the step (1) specifically comprises the following steps:

[0018] (1.1) Divide the complete research report into segments of fixed length, annotate the industry opinions in small batches of data segments through the opinion extraction model, and infer the industry opinion extraction results;

[0019] (1.2) Annotate the industry events contained in the small batch opinion data through the event extraction model and infer the industry event extraction results;

[0020] (1.3) Combine the research report fragments and the events extracted from the fragments, and use the large language model to extract the causal relationship between the events;

[0021] (1.4) Obtain the opinions, event data and causal relationships between events in the research report fragment.

[0022] Preferably, the step (2) specifically comprises the following steps:

[0023] (2.1) Perform data verification and cleaning;

[0024] (2.2) Normalize the cleaned data and perform standardization on the data;

[0025] (2.3) Associating data;

[0026] (2.4) Record the opinion event data and causal relationship data into the database.

[0027] Preferably, the step (2.1) is specifically as follows:

[0028] Automatically detect and correct errors and inconsistencies in data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information.

[0029] Preferably, the step (2.3) specifically comprises the following steps:

[0030] (2.3.1) Record the opinions and events extracted from the original text, and find the opinions and original text that contain a specific event;

[0031] (2.3.2) Link related events and ideas in different data fragments.

[0032] Preferably, the step (3) is specifically as follows:

[0033] Select the industry and time range for generating an industry research report, search the database for events that match the selected industry within the time period, and count the most frequently occurring events in the selected time and industry as core events to generate a research report.

[0034] The main feature of the device for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph is that the device comprises:

[0035] a processor configured to execute computer executable instructions;

[0036] A memory stores one or more computer executable instructions. When the computer executable instructions are executed by the processor, the various steps of the method for intelligently generating and processing industry research reports based on a large model and a causal graph are implemented.

[0037] The processor for realizing intelligent generation and processing of industry research reports based on large models and causal graphs has the main feature that the processor is configured to execute computer-executable instructions. When the computer-executable instructions are executed by the processor, the various steps of the above-mentioned method for realizing intelligent generation and processing of industry research reports based on large models and causal graphs are implemented.

[0038] The main feature of this computer-readable storage medium is that a computer program is stored thereon, and the computer program can be executed by a processor to implement the various steps of the above-mentioned method for intelligently generating and processing industry research reports based on large models and causal graphs.

[0039] The system, method, device, processor and computer-readable storage medium thereof for intelligent generation and processing of industry research reports based on large models and causal graphs of the present invention are adopted, and a paradigm for large language models to assist in the generation of industry research reports is proposed, which solves the pain points that the use of large language models to assist in the generation of research reports has poor technical usability and investment researchers have difficulty trusting the results generated by the models. A causal graph of industry events is constructed, and the graph is used to enhance the content generated by the model, thereby enhancing the traceability and explainability of the content generated by the model. The solution provided by the present invention shortens the time it takes for industry investment researchers to produce industry research reports, allowing practitioners to obtain the key points of relevant research reports within a period of time in a short period of time, greatly improving the work efficiency of relevant personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a schematic diagram of the workflow of the system for intelligently generating and processing industry research reports based on a large model and a cause-effect graph. DETAILED DESCRIPTION

[0041] In order to more clearly describe the technical content of the present invention, further description is given below in conjunction with specific embodiments.

[0042] The system of the present invention realizes intelligent generation and processing of industry research reports based on a large model and a causal graph, which includes a data extraction module, a data sorting module and a research report generation module. The data sorting module is connected to the data extraction module, and the research report generation module is connected to the data sorting module. The data extraction module is used to obtain the viewpoints, event data and the causal relationship between events in the research report segment; the data sorting module is used to further clean and structure the viewpoints, events and causal relationship data obtained from the data extraction module; the research report generation module is used to convert the processed data into a structured research report.

[0043] As a preferred embodiment of the present invention, the data extraction module includes an industry opinion extraction unit, an industry event extraction unit and an industry event causal relationship extraction unit. The industry event extraction unit is connected to the industry opinion extraction unit, and the industry event causal relationship extraction unit is connected to the industry opinion extraction unit and the industry event extraction unit. The industry opinion extraction unit is used to divide the complete research report into segments of fixed length, annotate the industry opinion in small batch data segments through the opinion extraction model, and infer the industry opinion extraction results; the industry event extraction unit is used to annotate the industry events contained in the small batch opinion data through the event extraction model, and infer the industry event extraction results; the industry event causal relationship extraction unit is used to combine the research report segments and the events extracted from the segments, and use a large language model to extract the causal relationship between events; and obtain the opinions, event data and causal relationships between events in the research report segments.

[0044] As a preferred embodiment of the present invention, the data sorting module includes a data verification and cleaning unit, a data normalization unit and a data association unit. The data normalization unit is connected to the data verification and cleaning unit, and the data association unit is connected to the data normalization unit. The data verification and cleaning unit is used to confirm the accuracy and completeness of the data, automatically detect and correct errors and inconsistencies in the data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information; the data normalization unit is used to normalize the cleaned data and standardize the data; the data association unit is used to record the views and events extracted from the original text segment, and at the same time find the views and original text segments containing the event according to a specific event; link the related events and views in different data segments.

[0045] As a preferred embodiment of the present invention, the research report generation module selects the industry and time range for generating the industry research report, searches the database for events that match the selected industry within the time period, and counts several events with the highest frequency of occurrence under the selected time and selected industry as core events to generate the research report.

[0046] The method of the present invention using the above system to realize intelligent generation and processing of industry research reports based on a large model and a causal graph includes the following steps:

[0047] (1) Obtain the opinions, event data, and causal relationships between events in the research report fragments;

[0048] (2) Further cleaning and structuring of the opinions, events, and causal relationship data obtained from the data extraction module;

[0049] (3) Convert the processed data into a structured research report.

[0050] As a preferred embodiment of the present invention, the step (1) specifically comprises the following steps:

[0051] (1.1) Divide the complete research report into segments of fixed length, annotate the industry opinions in small batches of data segments through the opinion extraction model, and infer the industry opinion extraction results;

[0052] (1.2) Annotate the industry events contained in the small batch opinion data through the event extraction model and infer the industry event extraction results;

[0053] (1.3) Combine the research report fragments and the events extracted from the fragments, and use the large language model to extract the causal relationship between the events;

[0054] (1.4) Obtain the opinions, event data and causal relationships between events in the research report fragment.

[0055] As a preferred embodiment of the present invention, the step (2) specifically comprises the following steps:

[0056] (2.1) Perform data verification and cleaning;

[0057] (2.2) Normalize the cleaned data and perform standardization on the data;

[0058] (2.3) Associating data;

[0059] (2.4) Record the opinion event data and causal relationship data into the database.

[0060] As a preferred embodiment of the present invention, the step (2.1) is specifically as follows:

[0061] Automatically detect and correct errors and inconsistencies in data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information.

[0062] As a preferred embodiment of the present invention, the step (2.3) specifically includes the following steps:

[0063] (2.3.1) Record the opinions and events extracted from the original text, and find the opinions and original text that contain a specific event;

[0064] (2.3.2) Link related events and ideas in different data fragments.

[0065] As a preferred embodiment of the present invention, the step (3) is specifically as follows:

[0066] Select the industry and time range for generating an industry research report, search the database for events that match the selected industry within the time period, and count the most frequently occurring events in the selected time and industry as core events to generate a research report.

[0067] The device of the present invention is used to realize intelligent generation and processing of industry research reports based on a large model and a cause-effect graph, wherein the device comprises:

[0068] a processor configured to execute computer executable instructions;

[0069] A memory stores one or more computer executable instructions. When the computer executable instructions are executed by the processor, the various steps of the method for intelligently generating and processing industry research reports based on a large model and a causal graph are implemented.

[0070] The processor of the present invention is used to realize the intelligent generation and processing of industry research reports based on large models and cause-effect graphs, wherein the processor is configured to execute computer-executable instructions, and when the computer-executable instructions are executed by the processor, the various steps of the above-mentioned method for realizing the intelligent generation and processing of industry research reports based on large models and cause-effect graphs are realized.

[0071] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program can be executed by a processor to implement the various steps of the above-mentioned method for intelligently generating and processing industry research reports based on a large model and a cause-and-effect graph.

[0072] In order to support the large language model to assist in extracting industry data from massive research reports and generate industry research reports, it is necessary to train data extraction models, causal relationship extraction models, and industry research report generation models to form a paradigm for large language models to assist in the generation of research reports for specific industries, and produce high-quality, traceable, and causal industry research reports. The present invention provides a complete solution for generating specific industry research reports from massive research reports, which mainly includes three modules: data extraction module, data sorting module, and research report generation module.

[0073] In the specific implementation of the present invention, a complete solution for generating a specific industry research report from a large amount of research reports is provided. The solution mainly includes three modules: data extraction module, data sorting module and research report generation module. The overall flow chart of the solution is as follows: Figure 1 shown.

[0074] The data extraction module divides the complete research report into segments of fixed length. The powerful large language model is used to annotate the industry opinions in the small batch data segments, and the fine-tuned open source model is used for industry opinion extraction. The large language model is then used to annotate the industry events contained in the small batch opinion data, and the fine-tuned open source model is used for industry event extraction. Combining the research report segments and the events extracted from the segments, the language understanding ability of the large language model can be used to extract the causal relationship between events. Through the above extraction work, we can eventually obtain the opinions in the research report segments, event data, and the causal relationship between events.

[0075] It should be pointed out that the industries marked in the module are selected through the analysis of the industrial chain of a specific industry in the financial field, and the common industries in the industry are selected for the extraction of opinion events with industry as the label. For the expansion of subdivided industries, the trained extraction model can be used for unsupervised extraction. The experimental results are better than using a large language model directly on the original research report for unsupervised extraction.

[0076] The main purpose of the data organization module is to further clean and structure the opinions, events and causal data obtained from the data extraction module.

[0077] The first is data verification and cleaning, which is mainly used to confirm the accuracy and completeness of the data. The system will automatically detect and correct errors and inconsistencies in the data, such as wrong dates, duplicate data, or incorrect formats. In addition, for semantically unclear or incomplete information, the system will use natural language processing technology to correct or supplement it.

[0078] After cleaning, the data will be normalized to ensure consistency and comparability. At this stage, the data will be standardized, including unified measurement units, formatted timestamps, and standardized terminology.

[0079] At the end of the data sorting module, the system will associate the data, which includes two aspects. The first aspect is the association of original text fragments, opinions, and events, recording the opinions and events extracted from the original text fragments, and at the same time, it can find the opinions and original text fragments containing the event based on a specific event. The second aspect is based on the extracted causal relationship. This module links the related events and opinions in different data fragments to build a more complete and systematic industry view. After completing the data association, the opinion event data and causal relationship data will be recorded in the database.

[0080] The research report generation module is the output interface of the entire system, which converts the processed data into a structured research report. The analyst selects the industry and time range for which the industry research report is to be generated. The system searches the database for events that match the selected industry within the time period and counts the events with the highest frequency in the selected time and industry as the core events.

[0081] Given the associated results in the data sorting module, the opinion set associated with the core event is screened out from the database. Based on the inductive sorting ability of the large language model, the opinion texts under the selected industry within the time range can be summarized to form core opinions, and combined with the upstream and downstream causal information of the core event in the causal relationship map, logically coherent and fluent texts such as market trend analysis, industry event discussion, and future forecasts can be generated and visualized.

[0082] This paper proposes an innovative data processing solution that deeply integrates the advanced technology of the big language model with the actual needs of financial research report analysis. By building an efficient opinion extraction and event extraction model, this method can automatically and accurately capture key information points from massive research report data, including market trends, company dynamics, expert opinions, etc. Utilizing the deep understanding and context analysis capabilities of the big language model, a deep interpretation and refinement of the research report content is achieved, providing investors, analysts and other users with valuable decision support.

[0083] This invention innovatively proposes a technical solution for automatically extracting structured events from the original research report and constructing a causal relationship map. This technology uses a complex natural language processing algorithm to accurately identify the events described in the research report and the causal relationship between them, and then stores this information in a structured form in the map. The causal relationship map not only intuitively displays the logical chain between events, but also supports multi-dimensional query and analysis, providing users with unprecedented deep insight capabilities. This innovative method greatly improves the efficiency and accuracy of research report analysis, and provides strong support for decision-making in the fields of finance and economics.

[0084] The present invention aims to significantly improve the interpretability and traceability of large language models in the process of generating research reports by introducing causal relationship graphs. This method uses causal relationship graphs as an important input and auxiliary tool for large language models, guiding the model to pay more attention to the logical connections and causal relationships between events when generating research reports. The generated research reports are not only rich in content and clear in viewpoints, but each part can be traced back to specific causal chains and original data, thus achieving full chain interpretability and traceability from data to viewpoints, and from viewpoints to conclusions. This innovative method not only enhances the credibility and persuasiveness of research reports, but also provides users with a more transparent and reliable basis for decision-making.

[0085] The technical solution of the present invention focuses on integrating large-scale machine learning models and causal graph technology to analyze and generate industry research reports. The technical solution of the present invention utilizes more comprehensive data processing and analysis methods, such as in-depth analysis that combines structured and unstructured data. The technical solution of the present invention includes innovative applications of big data and pre-trained models in the automatic generation of financial industry research reports, especially how to use these technologies to optimize data understanding, processing, and report generation. The technical solution of the present invention proposes a new method to dynamically integrate and analyze information from multiple data sources, and effectively generate in-depth industry analysis reports without human intervention, reflecting significant technological progress and advancement.

[0086] The specific implementation scheme of this embodiment can refer to the relevant description in the above embodiment, which will not be repeated here.

[0087] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0088] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0089] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0090] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0091] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the corresponding program may be stored in a computer-readable storage medium, which, when executed, includes one of the steps of the method embodiment or a combination thereof.

[0092] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0093] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0094] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0095] The system, method, device, processor and computer-readable storage medium thereof for intelligent generation and processing of industry research reports based on large models and causal graphs of the present invention are adopted, and a paradigm for large language models to assist in the generation of industry research reports is proposed, which solves the pain points that the use of large language models to assist in the generation of research reports has poor technical usability and investment researchers have difficulty trusting the results generated by the models. A causal graph of industry events is constructed, and the graph is used to enhance the content generated by the model, thereby enhancing the traceability and explainability of the content generated by the model. The solution provided by the present invention shortens the time it takes for industry investment researchers to produce industry research reports, allowing practitioners to obtain the key points of relevant research reports within a period of time in a short period of time, greatly improving the work efficiency of relevant personnel.

[0096] In this specification, the present invention has been described with reference to specific embodiments thereof. However, it is apparent that various modifications and variations may be made without departing from the spirit and scope of the present invention. Therefore, the specification and drawings should be regarded as illustrative rather than restrictive.

Claims

1. A system for intelligently generating and processing industry research reports based on a large model and causal graph, characterized in that: The system includes a data extraction module, a data sorting module and a research report generation module. The data sorting module is connected to the data extraction module, and the research report generation module is connected to the data sorting module. The data extraction module is used to obtain the viewpoints, event data and causal relationships between events in the research report segment; the data sorting module is used to further clean and structure the viewpoints, events and causal relationship data obtained from the data extraction module; The research report generation module is used to convert the processed data into a structured research report.

2. According to claim 1, the system for realizing intelligent generation and processing of industry research reports based on large models and causal graphs is characterized in that: The data extraction module includes an industry viewpoint extraction unit, an industry event extraction unit and an industry event causal relationship extraction unit. The industry event extraction unit is connected to the industry viewpoint extraction unit, and the industry event causal relationship extraction unit is connected to the industry viewpoint extraction unit and the industry event extraction unit. The industry viewpoint extraction unit is used to divide the complete research report into segments of fixed length, annotate the industry viewpoints in small batch data segments through the viewpoint extraction model, and infer the industry viewpoint extraction results; the industry event extraction unit is used to annotate the industry events contained in the small batch viewpoint data through the event extraction model, and infer the industry event extraction results; the industry event causal relationship extraction unit is used to combine the research report segments and the events extracted from the segments, and use the large language model to extract the causal relationship between events; And obtain the opinions, event data and causal relationships between events in the research report fragments.

3. The system for realizing intelligent generation and processing of industry research reports based on large models and causal graphs according to claim 1 is characterized in that: The data sorting module includes a data verification and cleaning unit, a data normalization unit and a data association unit. The data normalization unit is connected to the data verification and cleaning unit, and the data association unit is connected to the data normalization unit. The data verification and cleaning unit is used to confirm the accuracy and completeness of the data, automatically detect and correct errors and inconsistencies in the data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information; the data normalization unit is used to normalize the cleaned data and perform standardized processing on the data; The data association unit is used to record the viewpoints and events extracted from the original text segment, and find the viewpoints and original text segment containing the event according to a specific event; and link the related events and viewpoints in different data segments.

4. The system for realizing intelligent generation and processing of industry research reports based on large models and causal graphs according to claim 1 is characterized in that: The research report generation module selects the industry and time range for generating the industry research report, searches the database for events that match the selected industry within the time period, and counts several events with the highest frequency in the selected time and selected industry as core events to generate the research report.

5. A method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph using the system of claim 1, characterized in that: The method comprises the following steps: (1) Obtain the opinions, event data, and causal relationships between events in the research report fragments; (2) Further cleaning and structuring of the opinions, events, and causal relationship data obtained from the data extraction module; (3) Convert the processed data into a structured research report.

6. The method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph according to claim 5 is characterized in that: The step (1) specifically comprises the following steps: (1.1) Divide the complete research report into segments of fixed length, annotate the industry opinions in small batches of data segments through the opinion extraction model, and infer the industry opinion extraction results; (1.2) Annotate the industry events contained in the small batch opinion data through the event extraction model and infer the industry event extraction results; (1.3) Combine the research report fragments and the events extracted from the fragments, and use the large language model to extract the causal relationship between the events; (1.4) Obtain the opinions, event data and causal relationships between events in the research report fragment.

7. The method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph according to claim 5 is characterized in that: The step (2) specifically comprises the following steps: (2.1) Perform data verification and cleaning; (2.2) Normalize the cleaned data and perform standardization on the data; (2.3) Associating data; (2.4) Record the opinion event data and causal relationship data into the database.

8. The method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph according to claim 7 is characterized in that: The step (2.1) is specifically: Automatically detect and correct errors and inconsistencies in data, and use natural language processing technology to correct or supplement semantically unclear or incomplete information.

9. The method for realizing intelligent generation and processing of industry research reports based on a large model and a cause-effect graph according to claim 7 is characterized in that: The step (2.3) specifically comprises the following steps: (2.3.1) Record the opinions and events extracted from the original text, and find the opinions and original text that contain a specific event; (2.3.2) Link related events and ideas in different data fragments.

10. The method for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph according to claim 5 is characterized in that: The step (3) is specifically as follows: Select the industry and time range for generating an industry research report, search the database for events that match the selected industry within the time period, and count the most frequently occurring events in the selected time and industry as core events to generate a research report.

11. A device for realizing intelligent generation and processing of industry research reports based on a large model and a causal graph, characterized in that: The device comprises: a processor configured to execute computer executable instructions; A memory storing one or more computer executable instructions, which, when executed by the processor, implements the steps of the method for intelligently generating and processing industry research reports based on a large model and a causal graph as described in any one of claims 5 to 10.

12. A processor for realizing intelligent generation and processing of industry research reports based on large models and causal graphs, characterized in that: The processor is configured to execute computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method for intelligently generating and processing industry research reports based on a large model and a causal graph as described in any one of claims 5 to 10 are implemented.

13. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program can be executed by a processor to implement the various steps of the method for intelligently generating and processing industry research reports based on a large model and a causal graph as described in any one of claims 5 to 10.

Images