Financial research report knowledge base construction method and device, equipment and storage medium

Abstract

The invention relates to the technical field of data processing, in particular to a financial research report knowledge base construction method and device, equipment and a storage medium. Comprising the following steps: constructing a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions; correcting the preliminary knowledge extraction framework according to the target application scene to obtain a knowledge extraction framework corresponding to the target application scene; obtaining a plurality of financial research and report files corresponding to the target application scene; performing extraction processing on the plurality of financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; the ontology item is related to business ontology information in the knowledge extraction framework; and according to the knowledge extraction framework and the ontology item, constructing a knowledge base of the plurality of financial research and report files. According to the embodiment of the invention, the problem that the currently constructed financial research report knowledge base is inaccurate can be solved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, equipment and storage medium for constructing a financial research report knowledge base. Background Art

[0002] Parsing financial research reports and building a knowledge base of financial research reports can support policy analysis and related calculations in the financial field. Most financial research reports are in the portable document format (PDF), and the content is relatively uniform and fixed. The current way to build a knowledge base of financial research reports is generally to first perform document layout detection, that is, to distinguish which are paragraphs, which are tables, bar charts, line charts, etc., and then further disassemble and store them according to the results of the analysis.

[0003] However, the above-mentioned method for constructing the financial research report knowledge base is not accurate enough because the layout detection itself is not accurate enough, resulting in the constructed financial research report knowledge base being inaccurate. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides a method, device, equipment and storage medium for constructing a financial research report knowledge base, which can solve the problem of inaccuracy of the currently constructed financial research report knowledge base.

[0005] In a first aspect, the present application provides a method for constructing a financial research report knowledge base, including: constructing a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions; modifying the preliminary knowledge extraction framework according to a target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario; obtaining multiple financial research report files corresponding to the target application scenario; extracting and processing the multiple financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; the ontology items are related to the business ontology information in the knowledge extraction framework; and constructing a knowledge base for the multiple financial research report files based on the knowledge extraction framework and the ontology items.

[0006] In the second aspect, the present application provides a device for constructing a financial research report knowledge base, including: a construction module, which is used to construct a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions; a correction module, which is used to correct the preliminary knowledge extraction framework according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario; an acquisition module, which is used to obtain multiple financial research report files corresponding to the target application scenario; an extraction module, which is used to extract and process multiple financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; the ontology items are related to the business ontology information in the knowledge extraction framework; the construction module is also used to construct a knowledge base for multiple financial research report files based on the knowledge extraction framework and the ontology items.

[0007] In a third aspect, the present application provides an electronic device, comprising: a memory and a processor, the memory storing a computer program, and the processor implementing the steps of a method for constructing a financial research report knowledge base provided in any embodiment of the first aspect of the present application when executing the computer program.

[0008] In a fourth aspect, the present application provides a computer-readable storage medium, comprising: a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method for constructing a financial research report knowledge base provided in any embodiment of the first aspect of the present application are implemented.

[0009] In a fifth aspect, the present application provides a computer program product. When the computer program product runs on a computer, it enables the computer to execute the steps of the method for constructing a financial research report knowledge base as provided in any embodiment of the first aspect.

[0010] The technical solution provided by the present application has the following advantages compared with the prior art: First, a preliminary knowledge extraction framework is constructed based on the business ontology information of multiple business dimensions; then, the preliminary knowledge extraction framework is modified according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario. Afterwards, multiple financial research report files corresponding to the target application scenario are obtained, and the multiple financial research report files are extracted and processed according to the knowledge extraction framework to obtain the ontology items in each financial research report file. Among them, the ontology items are related to the business ontology information in the knowledge extraction framework. Finally, a knowledge base of multiple financial research report files is constructed based on the knowledge extraction framework and the ontology items. In this way, a knowledge extraction framework can be constructed based on the business ontology information of multiple business dimensions combined with actual application scenarios, and the ontology items in the financial research report files corresponding to the application scenario can be directly extracted according to the knowledge extraction framework to construct a knowledge base. It avoids the use of layout detection to construct a knowledge base, thereby improving the accuracy of constructing a financial research report knowledge base. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0012] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0013] Figure 1 This is an application scenario diagram of a method for constructing a financial research report knowledge base in one embodiment;

[0014] Figure 2 One of the flowcharts of the method for constructing a financial research report knowledge base provided in the embodiment of the present application;

[0015] Figure 3 The second flowchart of the method for constructing a financial research report knowledge base provided in the embodiment of the present application;

[0016] Figure 4 A structural block diagram of a device for constructing a financial research report knowledge base provided in an embodiment of the present application;

[0017] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0019] The construction method of the financial research report knowledge base provided in this application can be applied to Figure 1 In the application environment shown. The construction method of the financial research report knowledge base is applied to the construction system of the financial research report knowledge base. The construction system of the financial research report knowledge base includes a terminal 102 and a server 101. Among them, the terminal 102 can communicate with the server 101 through a network. The terminal 102 can include but is not limited to various personal computers, laptops, smart phones, tablet computers, portable wearable devices, etc. The server 101 can be implemented as an independent server or a server cluster composed of multiple servers.

[0020] It should be noted that the method for constructing the financial research report knowledge base provided in the embodiment of the present application can be executed by an electronic device. The electronic device can be a server 101, a terminal 102, or a device that integrates the functions of the server 101 and the terminal 102. The present application does not limit this.

[0021] In addition, the method for constructing the financial research report knowledge base provided in the embodiment of the present application can be executed by the construction device of the financial research report knowledge base, and the construction device of the financial research report knowledge base can be hardware or software. When the construction device of the financial research report knowledge base is hardware, it can be various electronic devices with external speaker function, including but not limited to mobile phones, computers, computers, tablet computers, televisions, smart TVs, vehicle-mounted devices, portable wearable devices, laser projection equipment, etc. When the construction device of the financial research report knowledge base is software, it can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules, or it can be implemented as a single software or software module. No specific limitation is made here.

[0022] In one embodiment, Figure 2 As shown, a method for constructing a financial research report knowledge base is provided. This embodiment mainly applies this method to Figure 1 Take terminal 102 in FIG. 1 as an example.

[0023] S11. Construct a preliminary knowledge extraction framework based on business ontology information from multiple business dimensions.

[0024] Among them, the business ontology information of a business dimension can be the knowledge set corresponding to the business under the dimension. For example, the business dimension can include business entities, business processes, corporate behaviors and events, financial derivatives, indexes, mortgages, securities, etc., and the corresponding business ontology information can be business entity ontology information, business process ontology information, corporate behaviors and events ontology information, financial derivatives ontology information, index ontology information, mortgage ontology information, securities ontology information, etc. The preliminary knowledge extraction framework at least includes the business ontology information corresponding to each business dimension in multiple business dimensions.

[0025] Specifically, the method of constructing a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions can be to directly determine all business ontology information as the preliminary knowledge extraction framework; or, based on the correspondence between multiple business dimensions and business ontology information, to construct a preliminary knowledge extraction framework that includes the correspondence between business dimensions and business ontology information.

[0026] For example, when the business dimension includes business entities, corporate behaviors and events, and the business ontology information corresponding to the business entities includes business entity names and business entity locations; when the business ontology information corresponding to corporate behaviors and events includes procurement and training, the way to construct a preliminary knowledge extraction framework based on the business ontology information of multiple business dimensions can be to directly determine the business entity name, business entity location, procurement, and training as the preliminary knowledge extraction framework; or to determine the business entity-business entity name, business entity-business entity location, corporate behaviors and events-procurement, and corporate behaviors and events-training as the preliminary knowledge extraction framework.

[0027] In some embodiments, before constructing a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions, the method for constructing a financial research report knowledge base further includes: obtaining business ontology information of multiple business dimensions based on an open source Internet ontology architecture. Specifically, business ontology information of multiple business dimensions can be obtained based on an open source Internet ontology architecture, such as fibo.

[0028] S12. Modify the preliminary knowledge extraction framework according to the target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario.

[0029] Among them, the target application scenario is used to indicate the scenario corresponding to building the knowledge base, which can be an industry scenario, an industrial scenario, etc.

[0030] Specifically, since the preliminary knowledge extraction framework includes business ontology information of multiple business dimensions, which is complex and numerous, and a single company will not cover all business dimensions, when revising the preliminary knowledge extraction framework, it is necessary to integrate and summarize according to the needs of the actual application scenario. When integrating and summarizing, it is not necessary to cover all business dimensions, but to extract various business ontology information related to the target application scenario.

[0031] The method for modifying the preliminary knowledge extraction framework according to the target application scenario can be to use a large language model to perform intent recognition according to the requirements of the target application scenario, delete and merge the preliminary knowledge extraction framework, and obtain a revised knowledge system and framework. The processing principle of deletion and merging is to retain the business ontology information required by the target application scenario, so that the revised knowledge extraction framework is more compatible with the target application scenario.

[0032] Of course, the situation and direction of each actual application scenario are different, and the business ontology focused on is also different, so the revised knowledge extraction framework is also different.

[0033] S13. Obtain multiple financial research report files corresponding to the target application scenario.

[0034] Specifically, financial research reports can be obtained from open source websites or think tanks on the Internet. The financial research reports correspond to the target application scenarios. For example, the open source websites on the Internet can be financial websites. Financial research reports are documents that securities companies and securities investment consulting companies use to analyze the value of securities and securities-related products or the factors that affect their market prices based on an independent and objective standpoint. They usually contain value analysis of specific securities and securities-related products, investment rating opinions, and other content. Financial research reports can be roughly divided into macro research, industry research, strategy research, company research, quantitative research, and other types.

[0035] For example, the content of a financial research report usually includes but is not limited to the following aspects: Investment highlights: including long-term and short-term views; Fundamental analysis: involving industry attributes, industry competitiveness, development trends, company business models, competitive advantages, operating activities, etc.; Valuation: including financial status, relative valuation, absolute valuation; Earnings forecast: making assumptions about key factors affecting performance such as product prices, sales volume, and raw material prices, and conducting sensitivity analysis; Risk warning: explaining the risk factors that may affect the investment; Disclaimer: legal disclaimer of the research report, etc.

[0036] S14. Extract and process multiple financial research report files according to the knowledge extraction framework to obtain the ontology items in each financial research report file.

[0037] Among them, the ontology items are related to the business ontology information in the knowledge extraction framework.

[0038] Specifically, after obtaining the financial research report file, extract the information in the financial research report file that corresponds to the business ontology information in the knowledge extraction framework to obtain the ontology items in each financial research report file. The method of extracting and processing multiple financial research report files according to the knowledge extraction framework to obtain the ontology items in each financial research report file can be to use a pre-trained extraction model to extract and process multiple financial research report files according to the knowledge extraction framework to obtain the ontology items in each financial research report file; or to first construct a prompt word according to the knowledge extraction framework, and then use the large language model and the prompt word to extract the ontology items from each financial research report file.

[0039] Among them, prompt words are constructed according to the knowledge extraction framework, and then the large language model and the prompt words are used to extract ontology items from each financial research report file. The prompt words can be constructed according to the knowledge extraction framework, and the large language model is allowed to directly perform named entity recognition (NER) and / or relation extraction (RE) extraction in a manner of following the instructions. That is, the value of the business ontology information in the knowledge extraction framework is input to locate the specific position of the value in the financial research report file, and then the corresponding value (item value) is extracted.

[0040] For example, after obtaining the financial research report file, the following prompt words can be constructed according to the knowledge extraction framework: "This is a text extraction task. Please identify and extract information related to the following knowledge list from the provided financial research report PDF document: company name, company rating, market performance, operating income / year, growth rate / year, price-earnings ratio / year, current assets / year, net profit / year. Please read the content of the research report carefully and find paragraphs or sentences related to any item in the above knowledge list. The extracted information should be specific text fragments, and please note that the content of the financial research report should be respected during the extraction process, and no form of tampering or distortion should be performed. If the financial research report contains charts, images or other non-text, please present the results in a clear and orderly manner after the extraction is completed, ensuring that the relevant information of each knowledge point is clearly marked and organized."

[0041] In this way, the results returned by Big Language may include: "(1) Company name: A; (2) Company rating: Overweight; (3) Market performance: April 23 (-20%), June 23: (-8%), August 23: (4%), October 23: (16%), December 23: (28%), February 24: (40%); (4) Operating income / year: RMB 18,658 million in 2000, RMB 22,091 million in 2001, RMB 26,046 million in 2002, and RMB 26,046 million in 2003. 28,548 million yuan in 2003 and 29,933 million yuan in 2004; (5) Growth rate / year: 46.9% in 2000, 18.4% in 2001, 17.9% in 2002, 9.6% in 2003 and 4.9% in 2004; (6) P / E ratio / year: 39.43 in 2000, 17.43 in 2001, 12.11 in 2002, 9.33 in 2003 and 8.36 in 2004; (7) Current assets / year: 8,383 million yuan in 2022…”

[0042] S15. Construct a knowledge base of multiple financial research report documents based on the knowledge extraction framework and ontology items.

[0043] Specifically, the method of constructing a knowledge base for multiple financial research report files based on the knowledge extraction framework and ontology items can be to directly use the business ontology information and the extracted ontology items in the knowledge extraction framework as the knowledge base for multiple financial research report files; or to obtain a vector knowledge base for multiple financial research report files after vector processing of the business ontology information and the extracted ontology items in the knowledge extraction framework; or to construct a knowledge base for multiple financial research report files based on the correspondence between the business ontology information and the ontology items in the knowledge extraction framework.

[0044] In some embodiments, a knowledge base of multiple financial research report files can be constructed based on the correspondence between business ontology information and ontology items in a knowledge extraction framework. The business ontology information and ontology items in each correspondence can be vectorized to obtain a vectorized knowledge base; or, the business ontology information in each correspondence can be used as the previous statement and the ontology item as the following statement to generate a question-and-answer statement knowledge base.

[0045] Specifically, the business ontology information and ontology items in each corresponding relationship are vectorized to obtain a vectorized knowledge base. For each corresponding relationship, the business ontology information and ontology items in the corresponding relationship are vectorized to obtain vectorized business ontology information and ontology items, and then the vectorized business ontology information and ontology items are stored in the knowledge base according to the corresponding relationship to obtain a knowledge base with retrieval-augmented generation (RAG) function. Among them, vectorization can be performed through embedding models, such as a deep learning model (word2vec) for generating word vectors, a model (GloVe) for generating word vectors, a pre-trained language representation model based on Transformer, etc. These embedding models can convert text data into vector form. Based on the correspondence between business ontology information and ontology items in the knowledge extraction framework, a knowledge base of multiple financial research report files can also be constructed by pre-processing the content in each corresponding relationship to remove useless information, such as punctuation marks and stop words. Then, vector processing techniques, such as the bag-of-words model and the term frequency-inverse document frequency (TF-IDF) statistical algorithm, are used to convert the content in the corresponding relationship into a vector representation. In this way, the content in each corresponding relationship can be represented by a vector, and each dimension of the vector corresponds to a specific word or phrase.

[0046] In the above scheme, first, a preliminary knowledge extraction framework is constructed based on the business ontology information of multiple business dimensions; then, the preliminary knowledge extraction framework is modified according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario. After that, multiple financial research report files corresponding to the target application scenario are obtained, and the multiple financial research report files are extracted and processed according to the knowledge extraction framework to obtain the ontology items in each financial research report file. Among them, the ontology items are related to the business ontology information in the knowledge extraction framework. Finally, a knowledge base of multiple financial research report files is constructed based on the knowledge extraction framework and the ontology items. In this way, a knowledge extraction framework can be constructed based on the business ontology information of multiple business dimensions combined with actual application scenarios, and the ontology items in the financial research report files corresponding to the application scenario can be directly extracted according to the knowledge extraction framework to construct a knowledge base. It avoids the use of layout detection to construct a knowledge base, thereby improving the accuracy of constructing a financial research report knowledge base.

[0047] In some embodiments, the preliminary knowledge extraction framework is modified according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario. The method can be to determine the scenario information corresponding to the target application scenario, and use a preset model to modify the preliminary knowledge extraction framework according to the scenario information to obtain the knowledge extraction framework corresponding to the target application scenario.

[0048] Among them, the preset model is a pre-trained correction model, or a large language model with intent recognition capability.

[0049] Specifically, when the preset model is a pre-trained correction model, the preset model is used to correct the preliminary knowledge extraction framework according to the scene information to obtain the knowledge extraction framework corresponding to the target application scenario. The scene information and the preliminary knowledge extraction framework can be directly input into the correction model, and the correction model corrects the preliminary knowledge extraction framework according to the scene information and then outputs the knowledge extraction framework corresponding to the target application scenario. When the preset model is a large language model with intent recognition capability, the preset model is used to correct the preliminary knowledge extraction framework according to the scene information to obtain the knowledge extraction framework corresponding to the target application scenario. The method can be to first construct a prompt word for the correction intent according to the scene information and the preliminary knowledge extraction framework, and then input the prompt word for the correction intent and the preliminary knowledge extraction framework into the large language model with intent recognition capability, and the large language model corrects the preliminary knowledge extraction framework according to the prompt word and then outputs the knowledge extraction framework corresponding to the target application scenario.

[0050] In the above scheme, the preliminary knowledge extraction framework can be modified according to the actual application scenario to obtain a knowledge extraction framework corresponding to the actual application scenario, making the generated knowledge extraction framework more directional and indirectly improving the accuracy of the knowledge base generated according to the knowledge extraction framework.

[0051] In some embodiments, Figure 3 As shown, the method of modifying the preliminary knowledge extraction framework according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario may also include the following steps:

[0052] S121. Determine scenario keywords corresponding to the target application scenario.

[0053] Specifically, the method for determining the scenario keywords corresponding to the target application scenario may be that the scenario keywords are set by relevant personnel according to actual conditions; or may be that pre-stored scenario keywords corresponding to the target application scenario are obtained.

[0054] S122. Calculate the correlation between scenario keywords and business ontology information in the preliminary knowledge extraction framework.

[0055] Specifically, a correlation algorithm may be used to calculate the correlation between the scenario keywords and the business ontology information in the preliminary knowledge extraction framework, for example, a cosine similarity algorithm, a Jaccard similarity algorithm, and the like.

[0056] S123. Construct a knowledge extraction framework corresponding to the target application scenario based on the business ontology information whose relevance is greater than a preset threshold.

[0057] Among them, the preset threshold is preset, such as a default value, or a value set by relevant personnel according to actual conditions.

[0058] In the above scheme, on the one hand, the preliminary knowledge extraction framework can be modified according to the actual application scenario to obtain the knowledge extraction framework corresponding to the actual application scenario, so that the generated knowledge extraction framework is more targeted, which indirectly improves the accuracy of the knowledge base generated according to the knowledge extraction framework. On the other hand, the preliminary knowledge extraction framework can be modified based on the calculation of similarity, which can be applied to a wider range of modification scenarios and improves the flexibility of modifying the preliminary knowledge extraction framework.

[0059] The embodiment of the present application can divide the functional modules of the construction device of the financial research report knowledge base according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0060] like Figure 4As shown, a device for constructing a financial research report knowledge base provided by an embodiment of the present application includes: a construction module 61, a correction module 62, an acquisition module 63, and an extraction module 64, wherein:

[0061] A construction module 61 is used to construct a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions; a correction module 62 is used to correct the preliminary knowledge extraction framework according to the target application scenario to obtain the knowledge extraction framework corresponding to the target application scenario; an acquisition module 63 is used to obtain multiple financial research report files corresponding to the target application scenario; an extraction module 64 is used to extract and process multiple financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; the ontology items are related to the business ontology information in the knowledge extraction framework; the construction module 61 is also used to construct a knowledge base of multiple financial research report files based on the knowledge extraction framework and the ontology items.

[0062] In some embodiments, the correction module 62 is specifically used to determine the scenario information corresponding to the target application scenario, and use a preset model to correct the preliminary knowledge extraction framework according to the scenario information to obtain the knowledge extraction framework corresponding to the target application scenario; the preset model is a pre-trained correction model, or a large language model with intent recognition capabilities.

[0063] In some embodiments, the correction module 62 is specifically used to: determine the scenario keywords corresponding to the target application scenario; calculate the correlation between the scenario keywords and the business ontology information in the preliminary knowledge extraction framework; and construct a knowledge extraction framework corresponding to the target application scenario based on the business ontology information whose correlation is greater than a preset threshold.

[0064] In some embodiments, the extraction module 64 is specifically used to construct prompt words according to the knowledge extraction framework, and use the large language model and the prompt words to extract ontology items from each financial research report file.

[0065] In some embodiments, the construction module 61 is specifically used to construct a knowledge base of multiple financial research report files based on the correspondence between business ontology information and ontology items in the knowledge extraction framework.

[0066] In some embodiments, the construction module 61 is specifically used to vectorize the business ontology information and ontology items in each corresponding relationship to obtain a vectorized knowledge base; or, the business ontology information in each corresponding relationship is used as the previous statement and the ontology item is used as the following statement to generate a question and answer statement knowledge base.

[0067] For the specific definition of the construction device of the financial research report knowledge base, please refer to the definition of the construction method of the financial research report knowledge base in the above text, which will not be repeated here. Each module in the above-mentioned construction device of the financial research report knowledge base can be implemented in whole or in part by software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0068] Figure 5 An electronic device according to an exemplary embodiment may include a processor 802, and the processor 802 is used to execute application code to implement the method for constructing a financial research report knowledge base in this application.

[0069] The processor 802 may be a central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present application.

[0070] like Figure 5 As shown, the electronic device may further include a memory 803. The memory 803 is used to store application program codes for executing the solution of the present application, and the execution is controlled by the processor 802.

[0071] The memory 803 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 803 may exist independently and be connected to the processor 802 via the bus 804. The memory 803 may also be integrated with the processor 802.

[0072] like Figure 5As shown, the electronic device may further include a communication interface 801, wherein the communication interface 801, the processor 802, and the memory 803 may be coupled to each other, for example, via a bus 804. The communication interface 801 is used to exchange information with other devices, for example, to support information exchange between the electronic device and other devices.

[0073] It should be pointed out that Figure 5 The device structure shown in the figure does not constitute a limitation on the electronic device, except Figure 5 In addition to the components shown, the electronic device may include more or fewer components than shown, or combine certain components, or arrange the components differently. The electronic device provided in this embodiment can execute the method for constructing the financial research report knowledge base provided in the above method embodiment, and its implementation principle and technical effect are similar to the above method, which will not be repeated here.

[0074] An embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, each process of the method for constructing a financial research report knowledge base in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0075] The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0076] An embodiment of the present application provides a computer program product, which stores a computer program. When the computer program is executed by a processor, each process of the method for constructing a financial research report knowledge base in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0077] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media that include computer-usable program code.

[0078] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static random access memory (SRAM) and dynamic random access memory (DRAM).

[0079] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0080] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0081] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a financial research report knowledge base, characterized in that: include: Based on the business ontology information of multiple business dimensions, a preliminary knowledge extraction framework is constructed; Modifying the preliminary knowledge extraction framework according to the target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario; Obtain multiple financial research report files corresponding to the target application scenario; Extracting and processing the plurality of financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; The ontology items are related to the business ontology information in the knowledge extraction framework; A knowledge base of the plurality of financial research report documents is constructed according to the knowledge extraction framework and the ontology items.

2. The method for constructing a financial research report knowledge base according to claim 1, characterized in that: The step of modifying the preliminary knowledge extraction framework according to the target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario includes: Determine the scenario information corresponding to the target application scenario, and use a preset model to correct the preliminary knowledge extraction framework according to the scenario information to obtain the knowledge extraction framework corresponding to the target application scenario; the preset model is a pre-trained correction model, or a large language model with intent recognition capability.

3. The method for constructing a financial research report knowledge base according to claim 1, characterized in that: The step of modifying the preliminary knowledge extraction framework according to the target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario includes: Determine the scenario keywords corresponding to the target application scenario; Calculating the correlation between the scenario keywords and the business ontology information in the preliminary knowledge extraction framework; Based on the business ontology information whose relevance is greater than a preset threshold, a knowledge extraction framework corresponding to the target application scenario is constructed.

4. The method for constructing a financial research report knowledge base according to claim 1, characterized in that: The extracting and processing the plurality of financial research report files according to the knowledge extraction framework to obtain the ontology items in each financial research report file includes: Prompt words are constructed according to the knowledge extraction framework, and ontology items are extracted from each financial research report file using a large language model and the prompt words.

5. The method for constructing a financial research report knowledge base according to claim 1, characterized in that: The step of constructing a knowledge base of the plurality of financial research report files according to the knowledge extraction framework and the ontology items includes: Based on the correspondence between the business ontology information in the knowledge extraction framework and the ontology items, a knowledge base of the multiple financial research report files is constructed.

6. The method for constructing a financial research report knowledge base according to claim 5, characterized in that: Based on the correspondence between the business ontology information in the knowledge extraction framework and the ontology items, a knowledge base of the plurality of financial research report files is constructed, including: Vectorize the business ontology information and ontology items in each corresponding relationship to obtain a vectorized knowledge base; or, The business ontology information in each corresponding relationship is used as the previous sentence, and the ontology item is used as the following sentence to generate a question and answer sentence knowledge base.

7. A device for constructing a financial research report knowledge base, characterized in that: include: A construction module is used to build a preliminary knowledge extraction framework based on business ontology information of multiple business dimensions; A correction module, used to correct the preliminary knowledge extraction framework according to the target application scenario to obtain a knowledge extraction framework corresponding to the target application scenario; An acquisition module, used to acquire multiple financial research report files corresponding to the target application scenario; An extraction module, configured to extract and process the plurality of financial research report files according to the knowledge extraction framework to obtain ontology items in each financial research report file; the ontology items are related to the business ontology information in the knowledge extraction framework; The construction module is also used to construct a knowledge base of the multiple financial research report files based on the knowledge extraction framework and the ontology items.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the method according to any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that When the computer program product is executed on a computer, the computer is enabled to implement the steps of the method according to any one of claims 1 to 6.

Images