Security compliance-oriented intelligent question and answer implementation method and device

Abstract

The invention discloses an intelligent question and answer implementation method and device for security compliance, and the method comprises the steps: determining a security compliance large model based on a target data set and a predetermined initial general large model; building an internal and external regulation knowledge base based on the regulation data set; and when a security compliance question is received, determining a security compliance answer corresponding to the security compliance question based on the security compliance question, the internal and external compliance knowledge base and the security compliance large model. Visibly, the security compliance answer can be determined based on the acquired security compliance question, the internal and external regulatory knowledge base established by the regulation data set and the security compliance large model determined by the target data set, so that the response accuracy and efficiency for the security compliance question can be improved; and therefore, security business decision and operation can be ensured to closely follow the latest security supervision dynamic state, and the security compliance risk can be reduced.

Description

Technical Field

[0001] The present invention relates to the field of securities compliance technology, and in particular to a method and device for implementing intelligent question and answer for securities compliance. Background Art

[0002] As a highly regulated industry, the securities industry faces numerous and complex regulatory policies, laws and regulations. Under the strict regulatory framework, securities practitioners will inevitably frequently encounter various legal compliance issues in the course of their daily business. These issues cover a wide range, from compliance judgment of business operations to the application of regulations in complex trading scenarios, and the slightest carelessness may lead to violations. This urgently requires securities companies to have a complete and efficient compliance Q&A function to provide practitioners with accurate and timely legal guidance to ensure that business activities are always on the track of compliance.

[0003] At present, in order to deal with compliance issues, securities companies usually set up a special regulatory research team to conduct in-depth interpretation of regulations and provide legal support and compliance Q&A services to company employees. However, this traditional model has high labor costs, is prone to errors and is inefficient, and it is difficult to meet the needs of rapid changes in securities business and massive compliance consulting. It can be seen that it is particularly important to provide an intelligent Q&A implementation solution for securities compliance to improve the accuracy and efficiency of responding to securities compliance issues. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method and device for realizing intelligent question and answer for securities compliance, which can improve the accuracy and efficiency of responding to securities compliance issues.

[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a method for implementing intelligent question and answer for securities compliance, the method comprising: Determine a securities compliance big model based on a target data set and a predetermined initial general big model; the target data set includes one or more of a securities industry question and answer data set, a legal knowledge question and answer data set, and a general encyclopedia question and answer data set; Building an internal and external regulation knowledge base based on a regulatory data set; the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; When a securities compliance question is received, a securities compliance answer corresponding to the securities compliance question is determined based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model.

[0006] As an optional implementation, in the first aspect of the present invention, determining the securities compliance big model based on the target data set and the predetermined initial general big model includes: Cleaning the target data set to remove non-target data in the target data set, and obtaining a cleaned data set corresponding to the target data set; the non-target data includes duplicate data, irrelevant data, and data whose data quality does not meet a preset quality condition in the target data set; Annotating the cleaned data set corresponding to the target data set to obtain a fine-tuning data set corresponding to the target data set; Adjusting the predetermined initial general macro model based on the fine-tuning data set to obtain a securities compliance macro model; Furthermore, before the forming of the domestic and foreign regulations knowledge base based on the regulatory data set, the method further includes: An initial regulatory data set is obtained and a correction operation is performed on the initial regulatory data set to obtain a regulatory data set corresponding to the initial regulatory data set; the correction operation includes an operation of removing redundant information, an operation of correcting format errors, and an operation of unifying terminology expressions.

[0007] As an optional implementation, in the first aspect of the present invention, before determining the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base and the securities compliance big model, the method further includes: Perform block processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a number of text blocks corresponding to each regulatory data; for each text block of each regulatory data, perform vectorization operation on the text block to obtain a text block vector corresponding to the text block; based on the text block vectors corresponding to all the text blocks, construct a knowledge vector library corresponding to the internal and external regulations knowledge base.

[0008] As an optional implementation manner, in the first aspect of the present invention, determining the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model includes: Performing a vectorization operation on the securities compliance issue to obtain a problem vector corresponding to the securities compliance issue; for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, performing a similarity calculation based on the text block vector and the problem vector to obtain a correlation parameter corresponding to the text block vector; Determining target regulatory data from the internal and external regulations knowledge base based on correlation parameters corresponding to all text block vectors in the knowledge vector base corresponding to the internal and external regulations knowledge base; All the target regulatory data are input into the securities compliance big model, and securities compliance answers corresponding to the securities compliance questions are generated based on the securities compliance big model.

[0009] As an optional implementation, in the first aspect of the present invention, for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge library, performing similarity calculation based on the text block vector and the question vector to obtain a correlation parameter corresponding to the text block vector includes: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, a similarity calculation is performed between the text block vector and the question vector based on the cosine similarity method and the matching algorithm in this paper to obtain the correlation parameter corresponding to the text block vector; the cosine similarity method is used to determine the cosine distance between the text block vector and the question vector; the matching algorithm in this paper is used to determine the distance between the securities compliance issue corresponding to the question vector and the text block corresponding to each of the text block vectors.

[0010] As an optional implementation, in the first aspect of the present invention, the determining the target regulatory data from the internal and external regulations knowledge base based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge base includes: Based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge library, a plurality of target text block vectors corresponding to the question vector are determined from all text block vectors in the knowledge vector library; For each of the target text block vectors, a text backtracking operation is performed on the target text block vector based on the internal and external regulations knowledge base corresponding to the knowledge vector library to obtain the regulatory data corresponding to the target text block vector; Determine the regulatory data corresponding to all the target text block vectors as candidate regulatory data; Based on the target attribute of each candidate regulatory data, at least one target regulatory data is determined from all the candidate regulatory data; the target attribute includes at least one of the document integrity, document length and relevance parameters of the corresponding target text block vector of the candidate regulatory data.

[0011] As an optional implementation, in the first aspect of the present invention, adjusting the predetermined initial general large model based on the fine-tuning data set to obtain the securities compliance large model includes: Injecting a low-rank matrix into a self-attention layer of a predetermined initial general large model, and adding a prompt vector in each layer of the initial general large model; the prompt vector is a vector related to securities compliance and risk; Based on the fine-tuning dataset, the low-rank matrix in the self-attention layer of the initial general large model and the prompt vector in each layer, the initial general large model is adjusted by gradient descent and back propagation methods to obtain a securities compliance large model.

[0012] The second aspect of the present invention discloses an intelligent question-answering device for securities compliance, the device comprising: A first determination module is used to determine a securities compliance big model based on a target data set and a predetermined initial general big model; the target data set includes one or more of a securities industry question and answer data set, a legal knowledge question and answer data set, and a general encyclopedia question and answer data set; A building module, used to build an internal and external regulation knowledge base based on a regulatory data set; the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; The second determination module is used to determine, when receiving a securities compliance question, a securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base and the securities compliance big model.

[0013] As an optional implementation, in the second aspect of the present invention, the first determination module determines the securities compliance big model based on the target data set and the predetermined initial general big model, specifically in the following manners: Cleaning the target data set to remove non-target data in the target data set, and obtaining a cleaned data set corresponding to the target data set; the non-target data includes duplicate data, irrelevant data, and data whose data quality does not meet a preset quality condition in the target data set; Annotating the cleaned data set corresponding to the target data set to obtain a fine-tuning data set corresponding to the target data set; Adjusting the predetermined initial general macro model based on the fine-tuning data set to obtain a securities compliance macro model; And, the device also includes: The correction module is used to obtain an initial regulatory data set and perform correction operations on the initial regulatory data set before the building module builds the domestic and foreign regulatory knowledge base based on the regulatory data set to obtain a regulatory data set corresponding to the initial regulatory data set; the correction operations include removing redundant information operations, revising format errors, and unifying terminology expressions.

[0014] As an optional implementation, in the second aspect of the present invention, the device further includes: A block cutting module is used to perform block cutting processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a number of text blocks corresponding to each regulatory data before the second determination module determines the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the internal and external regulations knowledge base and the securities compliance big model; for each text block of each regulatory data, perform vectorization operation on the text block to obtain a text block vector corresponding to the text block; and construct a knowledge vector library corresponding to the internal and external regulations knowledge base based on the text block vectors corresponding to all the text blocks.

[0015] As an optional implementation, in the second aspect of the present invention, the second determination module determines the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base and the securities compliance big model, specifically in the following manners: Performing a vectorization operation on the securities compliance issue to obtain a problem vector corresponding to the securities compliance issue; for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, performing a similarity calculation based on the text block vector and the problem vector to obtain a correlation parameter corresponding to the text block vector; Determining target regulatory data from the internal and external regulations knowledge base based on correlation parameters corresponding to all text block vectors in the knowledge vector base corresponding to the internal and external regulations knowledge base; All the target regulatory data are input into the securities compliance big model, and securities compliance answers corresponding to the securities compliance questions are generated based on the securities compliance big model.

[0016] As an optional implementation, in the second aspect of the present invention, the second determination module performs similarity calculation on each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base based on the text block vector and the question vector to obtain a correlation parameter corresponding to the text block vector, and the specific method includes: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, a similarity calculation is performed between the text block vector and the question vector based on the cosine similarity method and the matching algorithm in this paper to obtain the correlation parameter corresponding to the text block vector; the cosine similarity method is used to determine the cosine distance between the text block vector and the question vector; the matching algorithm in this paper is used to determine the distance between the securities compliance issue corresponding to the question vector and the text block corresponding to each of the text block vectors.

[0017] As an optional implementation, in the second aspect of the present invention, the second determination module determines the target regulatory data from the internal and external regulations knowledge base based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge base, and the specific method includes: Based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge library, a plurality of target text block vectors corresponding to the question vector are determined from all text block vectors in the knowledge vector library; For each of the target text block vectors, a text backtracking operation is performed on the target text block vector based on the internal and external regulations knowledge base corresponding to the knowledge vector library to obtain the regulatory data corresponding to the target text block vector; Determine the regulatory data corresponding to all the target text block vectors as candidate regulatory data; Based on the target attribute of each candidate regulatory data, at least one target regulatory data is determined from all the candidate regulatory data; the target attribute includes at least one of the document integrity, document length and relevance parameters of the corresponding target text block vector of the candidate regulatory data.

[0018] As an optional implementation, in the second aspect of the present invention, the first determination module adjusts the predetermined initial general large model based on the fine-tuning data set to obtain the securities compliance large model, specifically in the following manner: Injecting a low-rank matrix into a self-attention layer of a predetermined initial general large model, and adding a prompt vector in each layer of the initial general large model; the prompt vector is a vector related to securities compliance and risk; Based on the fine-tuning dataset, the low-rank matrix in the self-attention layer of the initial general large model and the prompt vector in each layer, the initial general large model is adjusted by gradient descent and back propagation methods to obtain a securities compliance large model.

[0019] The third aspect of the present invention discloses an intelligent question-answering implementation system for securities compliance, the system comprising: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the steps in the intelligent question-and-answer implementation method for securities compliance disclosed in the first aspect of the present invention.

[0020] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the intelligent question and answer implementation method for securities compliance disclosed in the first aspect of the present invention.

[0021] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The present invention determines the securities compliance big model based on the target data set and the predetermined initial general big model; builds the internal and external regulations knowledge base based on the regulatory data set; when receiving securities compliance issues, determines the securities compliance answers corresponding to the securities compliance issues based on the securities compliance issues, the internal and external regulations knowledge base and the securities compliance big model. It can be seen that the present invention can determine the securities compliance answers based on the acquired securities compliance issues, the internal and external regulations knowledge base built by the regulatory data set and the securities compliance big model determined by the target data set, which is conducive to improving the accuracy and efficiency of responding to securities compliance issues, and further can ensure that securities business decisions and operations keep up with the latest securities regulatory trends and reduce securities compliance risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a flowchart of a method for implementing intelligent question and answer for securities compliance disclosed in an embodiment of the present invention; Figure 2 It is a process framework diagram of a method for implementing intelligent question and answer for securities compliance disclosed in an embodiment of the present invention; Figure 3 It is a structural schematic diagram of a smart question-answering implementation device for securities compliance disclosed in an embodiment of the present invention; Figure 4 It is a structural diagram of an intelligent question-answering implementation system for securities compliance disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or terminal including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or terminals.

[0026] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0027] The present invention discloses a method and device for realizing intelligent question and answer for securities compliance. The method described in the embodiment of the present invention can determine securities compliance answers based on the acquired securities compliance questions, the internal and external regulations knowledge base formed by the regulatory data set, and the securities compliance big model determined by the target data set, which is conducive to improving the accuracy and efficiency of responding to securities compliance issues, thereby ensuring that securities business decisions and operations keep up with the latest securities regulatory trends and reduce securities compliance risks. The following are detailed descriptions.

[0028] Embodiment 1 See also Figure 1 , Figure 1 This is a flow chart of a method for implementing intelligent question-answering for securities compliance disclosed in an embodiment of the present invention. Figure 1 The described method can be applied to intelligent question-answering scenarios for securities compliance, and the embodiments of the present invention are not limited thereto. Figure 1 As shown, the intelligent question-answering implementation method for securities compliance includes the following operations: 101. Determine the securities compliance big model based on the target data set and the predetermined initial general big model; In an embodiment of the present invention, the target data set includes one or more of a securities industry question-and-answer data set, a legal knowledge question-and-answer data set, and a general encyclopedia question-and-answer data set; 102. Establish a knowledge base of domestic and foreign regulations based on regulatory data sets; In the embodiment of the present invention, the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; 103. When a securities compliance issue is received, the securities compliance answer corresponding to the securities compliance issue is determined based on the securities compliance issue, the internal and external regulations knowledge base, and the securities compliance big model.

[0029] In the embodiment of the present invention, in order to ensure that the large language model can accurately understand and generate answers that meet the characteristics of the securities industry, fine-tuning datasets are carefully selected and constructed from multiple sources. The securities industry question-and-answer dataset covers questions and answers in securities business operations, market trading rules, risk management, etc.; the legal knowledge question-and-answer dataset contains a wide range of legal provisions interpretations, case analysis, and application examples to help the large model understand complex legal logic; the general encyclopedia question-and-answer dataset provides basic knowledge background support, enhances the generalization ability of the large model, and enables it to reason in a wider knowledge background.

[0030] In the embodiment of the present invention, the internal and external regulations knowledge base is collected and organized for the special needs of the securities industry. Up to now, a total of 9,510 pieces of external regulatory data have been collected and integrated with 873 pieces of internal regulatory data of the company to form a comprehensive and authoritative internal and external regulations knowledge base.

[0031] In actual operation, the embodiments of the present invention can assist the compliance department in efficiently answering employees' regulatory questions and ensure that business decisions and operations keep up with the latest regulatory developments. When faced with changes in regulatory policies, the technical solution of the present invention can adapt quickly, effectively reducing compliance risks caused by misunderstanding or ignoring new regulations, and ensuring the compliance of the company's business. In the product development or business model innovation stage, the technical solution of the present invention can provide accurate compliance assessments to ensure that each step of innovation is on a legal and compliant track.

[0032] It can be seen that the present invention can determine securities compliance answers based on the obtained securities compliance issues, the internal and external regulations knowledge base constructed by the regulatory data set, and the securities compliance big model determined by the target data set, which is conducive to improving the accuracy and efficiency of responding to securities compliance issues, and further ensure that securities business decisions and operations keep up with the latest securities regulatory developments and reduce securities compliance risks.

[0033] In an optional embodiment, determining the securities compliance big model based on the target data set and the predetermined initial general big model in step 101 may include: Cleaning the target data set to remove non-target data in the target data set, and obtaining a cleaned data set corresponding to the target data set; the non-target data includes duplicate data, irrelevant data, and data whose data quality does not meet preset quality conditions in the target data set; Annotate the cleaned data set corresponding to the target data set to obtain a fine-tuned data set corresponding to the target data set; The predetermined initial general macro model is adjusted based on the fine-tuning data set to obtain the securities compliance macro model; Furthermore, before building a domestic and foreign regulations knowledge base based on the regulatory data set in step 102, the method may further include: An initial regulatory data set is obtained and a correction operation is performed on the initial regulatory data set to obtain a regulatory data set corresponding to the initial regulatory data set; the correction operation includes an operation of removing redundant information, an operation of revising format errors, and an operation of unifying terminology expressions.

[0034] In this optional embodiment, it is understood that the target data set undergoes a rigorous cleaning process to remove duplicates, irrelevant or low-quality data points to obtain a cleaned data set to ensure the purity of the data set. Subsequently, a team of domain experts can manually annotate the data to clarify information such as question types, keywords, and expected answers to obtain a fine-tuning data set, providing high-quality learning materials for fine-tuning the large language model.

[0035] In this optional embodiment, the regulatory data set used to construct the internal and external regulations knowledge base also undergoes a rigorous manual cleaning process, including removing redundant information, correcting format errors, unifying terminology, etc., to ensure the consistency and readability of regulatory texts. In this way, not only a detailed regulatory index is established, but also a solid foundation is laid for subsequent retrieval enhancement generation technology. Ultimately, this internal and external regulations knowledge base will become a core resource for compliance questions and answers, supporting users to query the most relevant regulatory provisions and their interpretations, while tracing the source of information to ensure the authority and accuracy of the answers. Through the meticulous processing of the fine-tuning data set and the internal and external regulations knowledge base, strong data support can be provided for securities companies' compliance questions and answers.

[0036] It can be seen that this optional embodiment can clean the target data set to remove non-target data to obtain a fine-tuning data set, and adjust the initial general large model based on the fine-tuning data set to obtain a securities compliance large model, which is beneficial to improving the determination accuracy of the securities compliance large model; it can also obtain the initial regulatory data set and perform correction operations on the initial regulatory data set to obtain a regulatory data set for use in establishing internal and external regulations knowledge bases, which is beneficial to improving the accuracy of the regulatory data set, and then improving the determination accuracy of the internal and external regulations knowledge bases, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0037] In another optional embodiment, before determining the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model in step 103, the method may further include: Each regulatory data contained in the internal and external regulations knowledge base is segmented to obtain a number of text blocks corresponding to each regulatory data; for each text block of each regulatory data, a vectorization operation is performed on the text block to obtain a text block vector corresponding to the text block; based on the text block vectors corresponding to all text blocks, a knowledge vector library corresponding to the internal and external regulations knowledge base is constructed.

[0038] In this optional embodiment, it is understood that the chunking process divides a large amount of text data into smaller, easy-to-manage and retrievable units. This process can be implemented through natural language processing technology, such as using sentence and paragraph methods, or document chunking technology based on topic models. Each knowledge unit after chunking will be converted into a vector and stored in a vector library. Further, the process can use text embedding technologies such as BERT, TF-IDF, etc. to convert text into vectors in a high-dimensional space.

[0039] It can be seen that this optional embodiment is able to perform block processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a text block corresponding to each regulatory data, and further perform vectorization operations on the text block to obtain a text block vector corresponding to the text block, and construct a knowledge vector library based on the text block vector, which is conducive to improving the accuracy of constructing the knowledge vector library, and further improving the accuracy of determining securities compliance answers based on the internal and external regulations knowledge base, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0040] In yet another optional embodiment, the above step 103, based on the securities compliance question, the knowledge base of domestic and foreign regulations, and the securities compliance big model, determining the securities compliance answer corresponding to the securities compliance question may include: Perform vectorization operations on securities compliance issues to obtain the problem vectors corresponding to the securities compliance issues; for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, perform similarity calculation based on the text block vector and the problem vector to obtain the correlation parameter corresponding to the text block vector; Based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge library, target regulatory data is determined from the internal and external regulations knowledge library; Input all target regulatory data into the securities compliance big model, and generate securities compliance answers corresponding to securities compliance questions based on the securities compliance big model.

[0041] In this optional embodiment, it can be understood that the original text of the regulatory data is input into the securities compliance big model instead of sending the text block to the big model in order to avoid the problem of the quality of knowledge input to the big model due to the lack of text segmentation strategy in the traditional retrieval enhanced generation (RAG), which leads to the inability to effectively answer questions. Retrieving the relevant original text directly with the help of the powerful recall retrieval ability of the big model can effectively solve this problem. Finally, these screened target regulatory data are input into the securities compliance big model, and the securities compliance big model will generate the final answer in combination with the document content of these target regulatory data.

[0042] It can be seen that this optional embodiment can perform vectorization operations on securities compliance issues to obtain question vectors, perform similarity calculations on each text block vector with the question vector to obtain correlation parameters, determine target regulatory data based on the correlation parameters, and input all target regulatory data into the securities compliance big model to generate securities compliance answers, which is beneficial to the accuracy of determining the target regulatory data, and further improves the accuracy of generating securities compliance answers, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0043] In yet another optional embodiment, in the above step, for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge library, performing similarity calculation based on the text block vector and the question vector to obtain the correlation parameter corresponding to the text block vector may include: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, a similarity calculation is performed between the text block vector and the question vector based on the cosine similarity method and the matching algorithm in this paper to obtain the correlation parameter corresponding to the text block vector; the cosine similarity method is used to determine the cosine distance between the text block vector and the question vector; the matching algorithm in this paper is used to determine the distance between the securities compliance issue corresponding to the question vector and the text block corresponding to each text block vector.

[0044] In this optional embodiment, the essence of the similarity calculation process is to quickly retrieve the most relevant target regulatory data from the knowledge vector library when the user asks a question. The matching algorithm in this article can be the BM25 algorithm, and the similarity calculation can refer to the following formula, combining cosine similarity and BM25 text matching algorithm similarity to achieve this goal.

[0045]

[0046] In the above formula, M represents the number of vectors in the knowledge vector library of the knowledge base, w represents the weight, represents the cosine similarity calculation, Indicates BM25 similarity calculation, Represents a text block vector i , represents the problem vector, Represents a text block vector i The corresponding text block, A block of text representing the securities compliance issue corresponding to the issue vector, Similarity i Represents a text block vector i The similarity calculation result with the question vector. Cosine similarity is used to measure the cosine distance between vectors. It can quickly find the vector closest to the question vector in the vector space. The BM25 text matching algorithm is a probability-based ranking function algorithm. It takes into account the frequency of query words in the document and the length of the document, and can effectively calculate the distance between the question and the text block.

[0047] It can be seen that this optional embodiment can complete the calculation based on the cosine similarity method and the matching algorithm of this article when performing similarity calculation based on the text block vector and the question vector, which is conducive to improving the calculation accuracy of the correlation parameters, and then improving the accuracy of determining the target regulatory data, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0048] In yet another optional embodiment, the above step of determining the target regulatory data from the internal and external regulatory knowledge base based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulatory knowledge base may include: Based on the correlation parameters corresponding to all the text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge base, several target text block vectors corresponding to the question vector are determined from all the text block vectors in the knowledge vector library; For each target text block vector, a text backtracking operation is performed on the target text block vector based on the internal and external regulations knowledge base corresponding to the knowledge vector library to obtain the regulatory data corresponding to the target text block vector; Determine the regulatory data corresponding to all target text block vectors as candidate regulatory data; Based on the target attribute of each candidate regulatory data, at least one target regulatory data is determined from all candidate regulatory data; the target attribute includes at least one of the document integrity, document length and relevance parameter of the corresponding target text block vector of the candidate regulatory data.

[0049] In this optional embodiment, based on the correlation parameters obtained by similarity calculation, the most relevant K (TOP-K) target text block vectors can be retrieved from the knowledge vector library (in actual operation, K is generally set to 10); further, based on the target attributes of each candidate regulatory data, at least one (TOP-N) target regulatory data can be determined from all candidate regulatory data (in actual operation, it is generally set to 4) to be input into the securities compliance model for generation.

[0050] In this optional embodiment, the text backtracking operation involves mapping the regulatory data from the knowledge vector library to the domestic and foreign regulatory knowledge library, which can be achieved by maintaining a mapping table when the knowledge vector library is constructed. In this way, the original text content corresponding to each vector can be quickly located, thereby providing the necessary context information for the generation stage.

[0051] Based on the above optional embodiments, the process framework diagram of the present invention can be as follows: Figure 2 As shown, the fine-tuned large model is first used as the base, and then based on the constructed internal and external regulations knowledge base, the retrieval enhanced generation (RAG) technology route is adopted to realize the compliance question-answering function. The core of the technology route is to combine the two stages of retrieval (Retrieval) and generation (Generation). First, the most relevant knowledge to the question is retrieved from the large-scale internal and external regulations knowledge base, and then the large model generates answers based on this. The goal of this invention is to improve the understanding and answering capabilities of securities compliance issues through the efficient retrieval and utilization of the internal and external regulations knowledge base.

[0052] It can be seen that this optional embodiment can determine several target text block vectors corresponding to the question vector based on the correlation parameters corresponding to all text block vectors; perform text backtracking operations on the target text block vectors based on the internal and external regulations knowledge base to obtain the regulatory data corresponding to the target text block vectors and determine them as candidate regulatory data; and determine at least one target regulatory data based on the target attributes of the candidate regulatory data, which is conducive to improving the accuracy of determining the target regulatory data, and then improving the accuracy of generating securities compliance answers, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0053] In yet another optional embodiment, the above step of adjusting the predetermined initial general macro model based on the fine-tuning data set to obtain the securities compliance macro model includes: Injecting a low-rank matrix into the self-attention layer of a predetermined initial general large model, and adding a prompt vector in each layer of the initial general large model; the prompt vector is a vector related to securities compliance and risk; Based on the fine-tuning dataset, the low-rank matrix in the self-attention layer of the initial general large model, and the prompt vector in each layer, the initial general large model is adjusted through gradient descent and back propagation methods to obtain the securities compliance large model.

[0054] In this optional embodiment, in terms of technical implementation, injecting a low-rank matrix into the self-attention layer of the predetermined initial general large model can be based on the LoRA (low-rank adaptive) fine-tuning method, and adding a prompt vector in each layer of the initial general large model can be based on the Prompt-tuning v2 fine-tuning method; based on this, on the basis of the constructed large model fine-tuning dataset, the technical solution of the present invention can integrate the two fine-tuning methods of LoRA and Prompt-tuning v2 to train the large model. LoRA is a parameter-efficient fine-tuning technology that completes fine-tuning by adding low-rank matrices between key layers of the pre-trained model. The core of this method is that it does not require the entire model to be retrained, but adjusts the behavior of the model by introducing additional trainable parameters, thereby reducing memory requirements and training time. This means that the model can be efficiently adjusted to adapt to specific securities compliance tasks without sacrificing model performance. Prompt-tuning v2 is an efficient method based on deep prompt adjustment, which achieves universality for different tasks and model sizes by adding continuous prompts to each layer of the pre-trained model. The key feature of this approach is that hints are added at every layer of the model, allowing the model to respond more directly to task-specific requirements. This means that securities compliance-related hints can be embedded in every layer of the model, further improving the model’s understanding and predictive capabilities for these specific areas.

[0055] In this optional embodiment, it can be further understood that the initial general large model used as the fine-tuning base model can be Alibaba's Tongyi Qianwen 2.5 open source large model, or the open source DeepSeek-R1, or other open source large models, which are not limited in the embodiments of the present invention. Based on the initial general large model, the LoRA method is used to inject a low-rank matrix into the self-attention layer of the model to simulate the effect of full parameter fine-tuning. At the same time, the Prompt-tuning v2 method is applied to add prompt vectors related to securities compliance and risk in each layer of the model. The overall fine-tuning architecture is as follows: Figure 2 As shown. In the subsequent fine-tuning process, the compliance question-answer pair dataset will be used to learn the low-rank matrix in LoRA and the prompt vector in Prompt-tuning v2 through the gradient descent combined with the back-propagation optimization algorithm. Such fine-tuning training will make the model more adaptable to the compliance question-answering scenarios of securities companies while maintaining the performance of the base large model. This method can not only improve the accuracy of the model's answers, but also ensure that the model's decision-making process meets the compliance requirements of the securities industry.

[0056] It can be seen that this optional embodiment can adjust the initial general big model based on the fine-tuning data set, the low-rank matrix in the self-attention layer of the initial general big model, and the prompt vector in each layer through gradient descent and back propagation methods to obtain the securities compliance big model, which is beneficial to improve the determination accuracy of the securities compliance big model, and then improve the generation accuracy of securities compliance answers based on the securities compliance big model, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0057] Embodiment 2 See also Figure 3 , Figure 3 : is a schematic diagram of a structure of an intelligent question-answering device for securities compliance disclosed in an embodiment of the present invention. Figure 3 The described device can be applied to any intelligent question-answering scenario for securities compliance, and the embodiments of the present invention are not limited thereto. Figure 3 As shown, the intelligent question-answering implementation device for securities compliance may include: A first determination module 201 is used to determine a securities compliance big model based on a target data set and a predetermined initial general big model; the target data set includes one or more of a securities industry question and answer data set, a legal knowledge question and answer data set, and a general encyclopedia question and answer data set; A building module 202 is used to build an internal and external regulation knowledge base based on a regulatory data set; the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; The second determination module 203 is used to determine, when receiving a securities compliance question, a securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model.

[0058] It can be seen that the device described in the embodiment of the present invention can determine securities compliance answers based on the obtained securities compliance issues, the internal and external regulations knowledge base constructed by the regulatory data set, and the securities compliance big model determined by the target data set, which is conducive to improving the accuracy and efficiency of responding to securities compliance issues, and further ensure that securities business decisions and operations keep up with the latest securities regulatory developments and reduce securities compliance risks.

[0059] In an optional embodiment, the first determination module 201 determines the securities compliance big model based on the target data set and the predetermined initial general big model, specifically in the following manners: Cleaning the target data set to remove non-target data in the target data set, and obtaining a cleaned data set corresponding to the target data set; the non-target data includes duplicate data, irrelevant data, and data whose data quality does not meet preset quality conditions in the target data set; Annotate the cleaned data set corresponding to the target data set to obtain a fine-tuned data set corresponding to the target data set; The predetermined initial general macro model is adjusted based on the fine-tuning data set to obtain the securities compliance macro model; And, the device may also include: The correction module is used to obtain the initial regulatory data set and perform correction operations on the initial regulatory data set before the building module 202 builds the internal and external regulatory knowledge base based on the regulatory data set to obtain a regulatory data set corresponding to the initial regulatory data set; the correction operations include removing redundant information operations, revising format errors, and unifying terminology expressions.

[0060] It can be seen that this optional embodiment can clean the target data set to remove non-target data to obtain a fine-tuning data set, and adjust the initial general large model based on the fine-tuning data set to obtain a securities compliance large model, which is beneficial to improving the determination accuracy of the securities compliance large model; it can also obtain the initial regulatory data set and perform correction operations on the initial regulatory data set to obtain a regulatory data set for use in establishing internal and external regulations knowledge bases, which is beneficial to improving the accuracy of the regulatory data set, and then improving the determination accuracy of the internal and external regulations knowledge bases, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0061] In another optional embodiment, the device may further include: The block cutting module is used to perform block cutting processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a number of text blocks corresponding to each regulatory data before the second determination module 203 determines the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the internal and external regulations knowledge base and the securities compliance big model; for each text block of each regulatory data, perform vectorization operation on the text block to obtain a text block vector corresponding to the text block; based on the text block vectors corresponding to all text blocks, construct a knowledge vector library corresponding to the internal and external regulations knowledge base.

[0062] It can be seen that this optional embodiment is able to perform block processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a text block corresponding to each regulatory data, and further perform vectorization operations on the text block to obtain a text block vector corresponding to the text block, and construct a knowledge vector library based on the text block vector, which is conducive to improving the accuracy of constructing the knowledge vector library, and further improving the accuracy of determining securities compliance answers based on the internal and external regulations knowledge base, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0063] In another optional embodiment, the second determination module 203 determines the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model, and the specific method includes: Perform vectorization operations on securities compliance issues to obtain the problem vectors corresponding to the securities compliance issues; for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, perform similarity calculation based on the text block vector and the problem vector to obtain the correlation parameter corresponding to the text block vector; Based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge library, target regulatory data is determined from the internal and external regulations knowledge library; Input all target regulatory data into the securities compliance big model, and generate securities compliance answers corresponding to securities compliance questions based on the securities compliance big model.

[0064] It can be seen that this optional embodiment can perform vectorization operations on securities compliance issues to obtain question vectors, perform similarity calculations on each text block vector with the question vector to obtain correlation parameters, determine target regulatory data based on the correlation parameters, and input all target regulatory data into the securities compliance big model to generate securities compliance answers, which is beneficial to the accuracy of determining the target regulatory data, and further improves the accuracy of generating securities compliance answers, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0065] In another optional embodiment, the second determination module 203 performs similarity calculation on each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge library based on the text block vector and the question vector to obtain a correlation parameter corresponding to the text block vector, specifically in the following manner: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, a similarity calculation is performed between the text block vector and the question vector based on the cosine similarity method and the matching algorithm in this paper to obtain the correlation parameter corresponding to the text block vector; the cosine similarity method is used to determine the cosine distance between the text block vector and the question vector; the matching algorithm in this paper is used to determine the distance between the securities compliance issue corresponding to the question vector and the text block corresponding to each text block vector.

[0066] It can be seen that this optional embodiment can complete the calculation based on the cosine similarity method and the matching algorithm of this article when performing similarity calculation based on the text block vector and the question vector, which is conducive to improving the calculation accuracy of the correlation parameters, and then improving the accuracy of determining the target regulatory data, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0067] In another optional embodiment, the second determination module 203 determines the target regulatory data from the internal and external regulatory knowledge base based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulatory knowledge base, specifically in the following manners: Based on the correlation parameters corresponding to all the text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge base, several target text block vectors corresponding to the question vector are determined from all the text block vectors in the knowledge vector library; For each target text block vector, a text backtracking operation is performed on the target text block vector based on the internal and external regulations knowledge base corresponding to the knowledge vector library to obtain the regulatory data corresponding to the target text block vector; Determine the regulatory data corresponding to all target text block vectors as candidate regulatory data; Based on the target attribute of each candidate regulatory data, at least one target regulatory data is determined from all candidate regulatory data; the target attribute includes at least one of the document integrity, document length and relevance parameter of the corresponding target text block vector of the candidate regulatory data.

[0068] It can be seen that this optional embodiment can determine several target text block vectors corresponding to the question vector based on the correlation parameters corresponding to all text block vectors; perform text backtracking operations on the target text block vectors based on the internal and external regulations knowledge base to obtain the regulatory data corresponding to the target text block vectors and determine them as candidate regulatory data; and determine at least one target regulatory data based on the target attributes of the candidate regulatory data, which is conducive to improving the accuracy of determining the target regulatory data, and then improving the accuracy of generating securities compliance answers, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0069] In another optional embodiment, the first determination module 201 adjusts the predetermined initial general macro model based on the fine-tuning data set to obtain the securities compliance macro model, specifically in the following manners: Injecting a low-rank matrix into the self-attention layer of a predetermined initial general large model, and adding a prompt vector in each layer of the initial general large model; the prompt vector is a vector related to securities compliance and risk; Based on the fine-tuning dataset, the low-rank matrix in the self-attention layer of the initial general large model, and the prompt vector in each layer, the initial general large model is adjusted through gradient descent and back propagation methods to obtain the securities compliance large model.

[0070] It can be seen that this optional embodiment can adjust the initial general big model based on the fine-tuning data set, the low-rank matrix in the self-attention layer of the initial general big model, and the prompt vector in each layer through gradient descent and back propagation methods to obtain the securities compliance big model, which is beneficial to improve the determination accuracy of the securities compliance big model, and then improve the generation accuracy of securities compliance answers based on the securities compliance big model, thereby improving the accuracy and efficiency of responding to securities compliance issues.

[0071] Embodiment 3 See also Figure 4 , Figure 4 Schematic diagram of a smart question-answering system for securities compliance disclosed in an embodiment of the present invention. Figure 4 The intelligent question-answering implementation system for securities compliance may include: A memory 301 storing executable program codes; a processor 302 coupled to the memory 301; The processor 302 calls the executable program code stored in the memory 301 to execute the steps in the intelligent question-and-answer implementation method for securities compliance described in the first embodiment of the present invention.

[0072] Embodiment 4 An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the intelligent question-and-answer implementation method for securities compliance described in the first embodiment of the present invention.

[0073] Embodiment 5 An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the intelligent question-and-answer implementation method for securities compliance described in Example 1.

[0074] The system embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative labor.

[0075] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0076] Finally, it should be noted that the intelligent question-answering implementation method and device for securities compliance disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for implementing intelligent question and answer for securities compliance, characterized in that: The method comprises: Determine a securities compliance big model based on a target data set and a predetermined initial general big model; the target data set includes one or more of a securities industry question and answer data set, a legal knowledge question and answer data set, and a general encyclopedia question and answer data set; Building an internal and external regulation knowledge base based on a regulatory data set; the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; When a securities compliance question is received, a securities compliance answer corresponding to the securities compliance question is determined based on the securities compliance question, the domestic and foreign regulations knowledge base, and the securities compliance big model.

2. The intelligent question-answering method for securities compliance according to claim 1, characterized in that: The step of determining the securities compliance big model based on the target data set and the predetermined initial general big model includes: Cleaning the target data set to remove non-target data in the target data set, and obtaining a cleaned data set corresponding to the target data set; the non-target data includes duplicate data, irrelevant data, and data whose data quality does not meet a preset quality condition in the target data set; Annotating the cleaned data set corresponding to the target data set to obtain a fine-tuning data set corresponding to the target data set; Adjusting the predetermined initial general macro model based on the fine-tuning data set to obtain a securities compliance macro model; Furthermore, before the forming of the domestic and foreign regulations knowledge base based on the regulatory data set, the method further includes: An initial regulatory data set is obtained and a correction operation is performed on the initial regulatory data set to obtain a regulatory data set corresponding to the initial regulatory data set; the correction operation includes an operation of removing redundant information, an operation of correcting format errors, and an operation of unifying terminology expressions.

3. The intelligent question-answering method for securities compliance according to claim 1 or 2, characterized in that: Before determining the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the internal and external regulations knowledge base and the securities compliance big model, the method further includes: Perform block processing on each regulatory data contained in the internal and external regulations knowledge base to obtain a number of text blocks corresponding to each regulatory data; for each text block of each regulatory data, perform vectorization operation on the text block to obtain a text block vector corresponding to the text block; based on the text block vectors corresponding to all the text blocks, construct a knowledge vector library corresponding to the internal and external regulations knowledge base.

4. The intelligent question-answering method for securities compliance according to claim 3, characterized in that: Determining the securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the internal and external regulations knowledge base, and the securities compliance big model includes: Performing a vectorization operation on the securities compliance issue to obtain a problem vector corresponding to the securities compliance issue; for each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, performing a similarity calculation based on the text block vector and the problem vector to obtain a correlation parameter corresponding to the text block vector; Determining target regulatory data from the internal and external regulations knowledge base based on correlation parameters corresponding to all text block vectors in the knowledge vector base corresponding to the internal and external regulations knowledge base; All the target regulatory data are input into the securities compliance big model, and securities compliance answers corresponding to the securities compliance questions are generated based on the securities compliance big model.

5. The intelligent question-answering method for securities compliance according to claim 4, characterized in that: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge library, similarity calculation is performed based on the text block vector and the question vector to obtain a correlation parameter corresponding to the text block vector, including: For each text block vector in the knowledge vector library corresponding to the internal and external regulations knowledge base, a similarity calculation is performed between the text block vector and the question vector based on the cosine similarity method and the matching algorithm in this paper to obtain the correlation parameter corresponding to the text block vector; the cosine similarity method is used to determine the cosine distance between the text block vector and the question vector; the matching algorithm in this paper is used to determine the distance between the securities compliance issue corresponding to the question vector and the text block corresponding to each of the text block vectors.

6. The intelligent question-answering method for securities compliance according to claim 4 or 5, characterized in that: The step of determining target regulatory data from the internal and external regulations knowledge base based on the correlation parameters corresponding to all text block vectors in the knowledge vector base corresponding to the internal and external regulations knowledge base includes: Based on the correlation parameters corresponding to all text block vectors in the knowledge vector library corresponding to the internal and external regulations knowledge library, a plurality of target text block vectors corresponding to the question vector are determined from all text block vectors in the knowledge vector library; For each of the target text block vectors, a text backtracking operation is performed on the target text block vector based on the internal and external regulations knowledge base corresponding to the knowledge vector library to obtain the regulatory data corresponding to the target text block vector; Determine the regulatory data corresponding to all the target text block vectors as candidate regulatory data; Based on the target attribute of each candidate regulatory data, at least one target regulatory data is determined from all the candidate regulatory data; the target attribute includes at least one of the document integrity, document length and relevance parameters of the corresponding target text block vector of the candidate regulatory data.

7. The intelligent question-answering method for securities compliance according to claim 2, characterized in that: The step of adjusting the predetermined initial general macro model based on the fine-tuning data set to obtain the securities compliance macro model includes: Injecting a low-rank matrix into a self-attention layer of a predetermined initial general large model, and adding a prompt vector in each layer of the initial general large model; the prompt vector is a vector related to securities compliance and risk; Based on the fine-tuning dataset, the low-rank matrix in the self-attention layer of the initial general large model and the prompt vector in each layer, the initial general large model is adjusted by gradient descent and back propagation methods to obtain a securities compliance large model.

8. An intelligent question-answering device for securities compliance, characterized in that: The device comprises: A first determination module is used to determine a securities compliance big model based on a target data set and a predetermined initial general big model; the target data set includes one or more of a securities industry question and answer data set, a legal knowledge question and answer data set, and a general encyclopedia question and answer data set; A building module, used to build an internal and external regulation knowledge base based on a regulatory data set; the regulatory data set includes a number of regulatory data, and the regulatory data includes external regulatory data or internal regulatory data; The second determination module is used to determine, when receiving a securities compliance question, a securities compliance answer corresponding to the securities compliance question based on the securities compliance question, the domestic and foreign regulations knowledge base and the securities compliance big model.

9. An intelligent question-answering implementation system for securities compliance, characterized in that: The system comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the intelligent question-and-answer implementation method for securities compliance as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, which, when called, are used to execute the intelligent question-and-answer implementation method for securities compliance as described in any one of claims 1 to 7.

Images