Document interpretation method and device, electronic equipment and computer readable storage medium

Abstract

The invention relates to the technical field of financial science and technology, and provides a document interpretation method and device, electronic equipment and a computer readable storage medium. Analyzing and processing the system document to obtain system attribute information; matching the system attribute information with a preset service category knowledge base, and determining a system type; extracting historical system information from the business category knowledge base according to the system type; generating an initial interpretation query according to the system document, and performing enhancement processing on the initial interpretation query according to the historical system information to obtain an enhanced query prompt; and inputting the enhanced query prompt into a preset large language model for analysis and processing to obtain system interpretation information. By means of the technical scheme, intelligent document interpretation can be achieved, the document interpretation efficiency is improved, and therefore development of banking business is better facilitated.

Description

Technical Field

[0001] The embodiments of the present application relate to but are not limited to the field of financial technology, and in particular to a document interpretation method, device, electronic device and computer-readable storage medium. Background Art

[0002] Currently, in banking business, after the regulatory system is issued, personnel in the industry will interpret the tasks according to the task assignment, and after the interpretation, it will be summarized and distributed to relevant personnel for study; manual interpretation requires strong professional knowledge, and requires a very good familiarity with the current system, related historical systems and internal systems, etc., in order to correctly and comprehensively interpret; however, the efficiency of manual interpretation is low, which will affect the development of banking business and increase the risk of regulatory penalties. Summary of the invention

[0003] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0004] In order to solve the problems mentioned in the above background technology, the embodiments of the present application provide a document interpretation method, device, electronic device and computer-readable storage medium, which can realize intelligent document interpretation, improve the efficiency of document interpretation, and thus be more conducive to the development of banking business.

[0005] In a first aspect, an embodiment of the present application provides a document interpretation method, comprising:

[0006] Obtaining system documents;

[0007] Analyzing and processing the system document to obtain system attribute information;

[0008] Matching the system attribute information with a preset business category knowledge base to determine the system type;

[0009] extracting historical system information from the business category knowledge base according to the system type;

[0010] generating an initial interpretation query according to the system document, and enhancing the initial interpretation query according to the historical system information to obtain an enhanced query prompt;

[0011] The enhanced query prompt is input into a preset large language model for analysis and processing to obtain system interpretation information.

[0012] According to some embodiments of the present application, the step of analyzing and processing the system document to obtain system attribute information includes:

[0013] Determine attribute keywords based on a preset attribute database;

[0014] The system attribute information is obtained by querying the system document according to the attribute keywords.

[0015] According to some embodiments of the present application, matching the policy attribute information with a preset business category knowledge base to determine the policy type includes:

[0016] Performing information splitting processing on the system attribute information to obtain multiple basic attributes;

[0017] The system type of the system document is obtained by matching a plurality of the basic attributes with the business category knowledge base.

[0018] According to some embodiments of the present application, the business category knowledge base includes a plurality of original system information, and each of the original system information corresponds to a type information, and the extracting of historical system information from the business category knowledge base according to the system type includes:

[0019] Matching the system type with the type information of the original system information to obtain the target original system information;

[0020] The target original system information is determined as the historical system information.

[0021] According to some embodiments of the present application, the step of performing enhancement processing on the initial interpretation query according to the historical system information to obtain an enhanced query prompt includes:

[0022] fusing the historical system information with the initial interpretation query to obtain context information;

[0023] Compressing the context information to obtain an initial query prompt;

[0024] The initial query prompt is combined with preset query guidance information to obtain the enhanced query prompt.

[0025] According to some embodiments of the present application, the step of inputting the enhanced query prompt into a preset large language model for analysis and processing to obtain the system interpretation information includes:

[0026] Preprocessing the enhanced query prompt to obtain coded prompt information;

[0027] Converting the encoded prompt information into a high-dimensional prompt vector;

[0028] Inputting the high-dimensional prompt vector into a preset self-attention mechanism network model for text semantic analysis and processing to obtain initial text semantic information;

[0029] Inputting the initial text semantic information into a preset decoder for prediction processing to obtain text prediction information;

[0030] The text prediction information is converted to obtain the system interpretation information.

[0031] According to some embodiments of the present application, the preprocessing of the enhanced query prompt to obtain the encoded prompt information includes:

[0032] Performing word segmentation processing on the enhanced query prompt to obtain multiple query prompt segments;

[0033] Cleaning the plurality of query prompt segments to obtain a plurality of sorted prompt segments;

[0034] The coded prompt information is obtained by performing coding conversion processing on the plurality of the sorting prompt segments.

[0035] In a second aspect, an embodiment of the present application further provides a document interpretation device, the device comprising:

[0036] The first processing module is used to obtain system documents;

[0037] The second processing module is used to analyze and process the system document to obtain system attribute information;

[0038] The third processing module is used to match the system attribute information with a preset business category knowledge base to determine the system type;

[0039] A fourth processing module, configured to extract historical system information from the business category knowledge base according to the system type;

[0040] A fifth processing module, configured to generate an initial interpretation query according to the system document, and enhance the initial interpretation query according to the historical system information to obtain an enhanced query prompt;

[0041] The sixth processing module is used to input the enhanced query prompt into a preset large language model for analysis and processing to obtain system interpretation information.

[0042] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the document interpretation method as described in the first aspect above is implemented.

[0043] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the document interpretation method described in the first aspect above.

[0044] According to the document interpretation method of the embodiment provided by the present application, at least the following beneficial effects are achieved: in the process of document interpretation, firstly, the system document is obtained; then, the system attribute information can be obtained by analyzing and processing the system document; then, the system attribute information and the preset business category knowledge base are matched to determine the system type; then, historical system information is extracted from the business category knowledge base according to the system type; then, an initial interpretation query is generated according to the system document, and the initial interpretation query is enhanced according to the historical system information to obtain an enhanced query prompt; finally, the enhanced query prompt is input into the preset large language model for analysis and processing to obtain the corresponding system interpretation information. Through the above method, intelligent document interpretation can be achieved, the efficiency of document interpretation is improved, and it is more conducive to the development of banking business. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to provide further understanding of the technical solution of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present application and do not constitute a limitation on the technical solution of the present application.

[0046] Figure 1 is a flowchart of a document interpretation method provided by an embodiment of the present application;

[0047] Figure 2 is a specific flow chart of step S200 provided in one embodiment of the present application;

[0048] Figure 3 is a specific flow chart of step S300 provided in one embodiment of the present application;

[0049] Figure 4 is a specific flow chart of step S400 provided in one embodiment of the present application;

[0050] Figure 5 is a specific flow chart of step S500 provided in one embodiment of the present application;

[0051] Figure 6 is a specific flow chart of step S600 provided in one embodiment of the present application;

[0052] Figure 7 is a specific flow chart of step S610 provided in one embodiment of the present application;

[0053] Figure 8 is a schematic diagram of a document interpretation device provided by an embodiment of the present application;

[0054] Fig. 9 It is a schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0055] 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.

[0056] It should be noted that although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification and claims and the above drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0057] It should be noted that, unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0058] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0059] AI is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. Artificial intelligence is a branch of computer science. Artificial intelligence attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Research in this field includes robots, language recognition, image recognition, natural language processing and expert systems. Artificial intelligence can simulate the information process of human consciousness and thinking. Artificial intelligence is also a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0060] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0061] Artificial intelligence is AI. AI is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0062] The servers involved in artificial intelligence technology can be independent servers or cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), as well as big data and artificial intelligence platforms.

[0063] The present application provides a document interpretation method, device, electronic device and computer-readable storage medium. In the process of document interpretation, first obtain the system document; then analyze and process the system document to obtain the system attribute information; then match the system attribute information with the preset business category knowledge base to determine the system type; then extract the historical system information from the business category knowledge base according to the system type; then generate an initial interpretation query according to the system document, and enhance the initial interpretation query according to the historical system information to obtain an enhanced query prompt; finally, input the enhanced query prompt into the preset large language model for analysis and processing to obtain the corresponding system interpretation information. Through the above method, intelligent document interpretation can be achieved, the efficiency of document interpretation is improved, and it is more conducive to the development of banking business.

[0064] The document interpretation method provided in the embodiment of the present application relates to the field of financial technology. The document interpretation method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the document interpretation method, etc., but is not limited to the above forms.

[0065] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0066] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0067] The embodiments of the present application are further described below in conjunction with the accompanying drawings.

[0068] like Figure 1 As shown, Figure 1 It is a flowchart of a document interpretation method provided by an embodiment of the present application, and the document interpretation method includes but is not limited to steps S100 to S600.

[0069] Step S100, obtaining system documents;

[0070] Step S200, analyzing and processing the system document to obtain system attribute information;

[0071] Step S300, matching the system attribute information with the preset business category knowledge base to determine the system type;

[0072] Step S400, extracting historical system information from the business category knowledge base according to the system type;

[0073] Step S500, generating an initial interpretation query according to the system document, and enhancing the initial interpretation query according to the historical system information to obtain an enhanced query prompt;

[0074] Step S600: input the enhanced query prompt into a preset large language model for analysis and processing to obtain system interpretation information.

[0075] It should be noted that in the process of document interpretation, first obtain the system document; then analyze and process the system document to obtain the system attribute information; then match the system attribute information with the preset business category knowledge base to determine the system type; then extract the historical system information from the business category knowledge base according to the system type; then generate the initial interpretation query according to the system document, and enhance the initial interpretation query according to the historical system information to obtain the enhanced query prompt; finally, input the enhanced query prompt into the preset large language model for analysis and processing to obtain the corresponding system interpretation information. Through the above method, intelligent document interpretation can be realized, the efficiency of document interpretation is improved, and it is more conducive to the development of banking business.

[0076] It is worth noting that the system document in the embodiment of the present application can be a relevant regulatory document for banking business, and the interpretation task of the corresponding regulatory document includes the following aspects: extraction of basic attributes of the system: extracting the system name, issuing unit, regulatory system document number, issuing date, and implementation date; system business category mapping: classifying articles according to business categories; system historical evolution: extracting the historical evolution information of the system; system clause interpretation: interpreting the system clauses, outputting the interpretation content, and highlighting the key points and key information of the content; system clause change point interpretation: analyzing two different versions of the same system, associating the content of the new and old versions of the clauses, and interpreting the content change points; system clause association penalties: extracting punitive clauses and regulations, and associating system clauses; system clause association penalty cases: system clause association regulatory penalty cases; system association with our bank system: regulatory system association with the system clauses within the bank, comparing differences, and helping to improve the bank's internal system. For the technical solution of the embodiment of the present application, introducing artificial intelligence into the interpretation process of system documents can well solve the inefficiency caused by traditional interpretation methods, as well as the problems of understanding deviation or omission, and well improve the efficiency and accuracy of document interpretation.

[0077] It is worth noting that in the process of obtaining system documents, the user's permission or consent must be obtained first, and the collection, use and processing of these data will comply with relevant laws, regulations and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or jump to a confirmation page. After the user's separate permission or consent is clearly obtained, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0078] It should be noted that by analyzing and processing the system document, the system attribute information can be obtained, which is a prerequisite for the subsequent determination of the system type; the system attribute information obtained by analysis is matched with the preset business category knowledge base to obtain the system type corresponding to the system document; subsequently, the historical system information can be extracted from the business category knowledge base according to the system type; then, the extracted historical system information can be used to enhance the initial interpretation query generated according to the system document to obtain enhanced query prompts; finally, the enhanced query prompts obtained are input into the preset large language model for analysis and processing to obtain the corresponding system interpretation information; wherein, in the embodiment of the present application, the large language model is used as a basic tool, which can process a large amount of information and provide in-depth understanding; and the initial interpretation query can be enhanced to retrieve key information from a large amount of relevant materials, and the integrated enhanced query prompts can be input into the large language model for interpretation processing, thereby providing a more comprehensive and accurate interpretation result.

[0079] For example, in the process of interpreting the bank's system documents, the basic attributes of the system are extracted to obtain the system name, issuing unit, regulatory system number, issuing date, implementation date, etc. The system business categories are mapped to classify the system. The historical evolution of the system is extracted to identify and extract the release date of each version and the major changes compared with the previous version. For the interpretation of system clauses, each clause in the system is interpreted separately, and the output interpretation content includes the main meaning of the clause, scope of application, etc. For the interpretation of the change points of system clauses, two different versions of the same system are compared to find out the change points of the clauses, and interpret the significance and impact of these change points. For the association of system clauses with penalties, the punitive clauses in the system are extracted and associated with relevant laws and regulations for further interpretation and understanding. For the association of system clauses with penalty cases, according to the system clauses, relevant regulatory penalty cases are searched and associated, which helps to understand the application and effect of the clauses in actual operations.

[0080] It is worth noting that through the above technical solutions, an efficient and intelligent interpretation process is achieved: this solution can realize an automated, efficient and accurate interpretation process by using a large model to interpret the regulatory system, greatly improving the efficiency and quality of interpretation. The current accuracy and sales efficiency of interpretation have greatly exceeded that of manual work. Comprehensive coverage of regulatory systems: this solution not only includes the interpretation of the regulatory system text, but also includes the extraction and association of basic attributes, business categories, historical evolution, penalties and cases of the system, achieving comprehensive coverage and management of the regulatory system. Real-time update of regulatory systems: this solution can monitor the update of regulatory systems in real time, and automatically interpret and update management to ensure that relevant personnel can grasp the latest regulatory policies and requirements in a timely manner. Flexible application methods: this solution can be flexibly applied to various business scenarios, such as business control, risk assessment, knowledge inheritance, etc., providing strong support for the development of banking business. Powerful intelligent reasoning and suggestion capabilities: this solution uses the intelligent reasoning and suggestion capabilities of the large model to provide intelligent suggestions and guidance to relevant personnel based on the interpretation results and business needs, further improving business efficiency and quality.

[0081] In some embodiments, Figure 2 As shown, the above step S200 may include but is not limited to steps S210 to S220.

[0082] Step S210, determining attribute keywords based on a preset attribute database;

[0083] Step S220: query the system document according to the attribute keywords to obtain system attribute information.

[0084] It should be noted that in the process of analyzing and processing system documents to obtain system attribute information, the attribute keywords are first determined based on the preset attribute database; then the system documents are queried according to the attribute keywords to obtain the corresponding system attribute information; after obtaining the system attribute information, the system attributes can be matched with the pre-set business category knowledge base to obtain the corresponding system attribute information, which is the prerequisite for the subsequent document interpretation.

[0085] It is worth noting that the attribute database includes multiple attribute keywords, and in the subsequent query process, the multiple attribute keywords can be matched with the attribute information of the system document, and then the system attribute information corresponding to the system document can be determined.

[0086] In some embodiments, Figure 3 As shown, the above step S300 may include but is not limited to step S310 to step S320.

[0087] Step S310, splitting the system attribute information to obtain multiple basic attributes;

[0088] Step S320, matching multiple basic attributes with the business category knowledge base to obtain the policy type of the policy document.

[0089] It should be noted that in the process of matching the system attribute information with the preset business category knowledge base to determine the system type, the system attribute information can be split to obtain multiple basic attributes; then the multiple basic attributes and the business category knowledge base can be matched to obtain the system type corresponding to the system document, thereby making preparations for the subsequent reading of historical system information.

[0090] It is worth noting that by splitting the system attribute information, multiple basic attributes can be obtained; then, the multiple basic attributes and the business category knowledge base are matched to obtain the system type corresponding to the system document; illustratively, for banking business, the system type corresponding to the system document can be obtained by matching multiple basic attributes and the business category knowledge base, for example, the system type of banking business is the deposit business supervision system.

[0091] In some embodiments, Figure 4 As shown, the business category knowledge base includes a plurality of original system information, and each original system information corresponds to a type information. The above step S400 may include but is not limited to step S410 to step S420.

[0092] Step S410, matching the system type with the type information of the original system information to obtain the target original system information;

[0093] Step S420: determining the target original system information as historical system information.

[0094] It should be noted that in the process of extracting historical system information from the business category knowledge base according to the system type, the system type is first matched with the type information of the original system information to obtain the target original system information; then the matched target original system information is determined as the historical system information, thereby making the prerequisite preparation for the subsequent generation of enhanced query prompts.

[0095] It is worth noting that in the process of determining historical institutional information, the institutional type and the type information of the original institutional information in the business category knowledge base can be matched and processed, and then the target original institutional information can be determined from the business category knowledge base; finally, the target original institutional information can be determined as historical institutional information.

[0096] It is understandable that the business category knowledge base includes multiple original system information, which is the previously existing system information. By matching the system type with the type information of the original system information, the corresponding target original system information can be obtained. Subsequently, the screened target original system information can be determined as historical system information to prepare for the generation of enhanced query prompts. For example, the business category knowledge base can include multiple original system information related to banking business.

[0097] In some embodiments, Figure 5 As shown, the above step S500 may include but is not limited to step S510 to step S530.

[0098] Step S510, fusing the historical system information and the initial interpretation query to obtain context information;

[0099] Step S520, compressing the context information to obtain an initial query prompt;

[0100] Step S530: The initial query prompt is combined with the preset query guidance information to obtain an enhanced query prompt.

[0101] It should be noted that in the process of enhancing the initial interpretation query according to the historical system information to obtain the enhanced query prompt, the historical system information and the initial interpretation query are firstly integrated to obtain the context information; then the context information is compressed to obtain the initial query prompt; finally, the initial query prompt is combined with the preset query guidance information to obtain the corresponding enhanced query prompt. By enhancing the initial interpretation query, key information can be retrieved from a large amount of relevant information, and the integrated enhanced query prompt is input into the large language model for interpretation processing, thereby providing a more comprehensive and accurate interpretation result.

[0102] It is worth noting that Retrieval-Augmented Generation (RAG) is a technology that combines information retrieval and neural network generation models. It achieves accurate, diverse and contextual text generation based on a large-scale knowledge base by introducing relevant external information in the generation process. The purpose of RAG technology is to improve the performance of the generation model. Its key innovation is to break the limitation of traditional generation models that only rely on their own parameters to predict outputs, and instead introduce retrieval strategies to obtain relevant information in external knowledge bases, and then use this information to guide the model to generate more accurate and more informative content. This innovation has greatly improved the performance of the model in various tasks such as question answering, dialogue systems, and text summarization. The implementation process of RAG technology usually includes three stages: retrieval stage, fusion stage, and generation stage. In the retrieval stage, information related to the input text is retrieved from the external knowledge base through information retrieval technology. In the fusion stage, the retrieved information is fused with the input text to provide rich context for the generation model. Finally, in the generation stage, a pre-trained natural language generation model is used to generate the corresponding output based on the input information. The role and value of RAG technology are reflected in improving generation quality, increasing diversity, and reducing factual errors (alleviating model hallucinations). It has broad application prospects in open domain question answering, dialogue systems, document summarization and generation, etc. In addition, RAG technology also faces some challenges, such as the retrieval process may be limited by the quality and coverage of the knowledge base, and the way of information fusion and the performance of the generation model have an important impact on the quality of the final generated results, which need to be carefully designed and adjusted.

[0103] In some embodiments, Figure 6 As shown, the above step S600 may also include but is not limited to steps S610 to S650.

[0104] Step S610, pre-processing the enhanced query prompt to obtain coded prompt information;

[0105] Step S620, converting the encoded prompt information into a high-dimensional prompt vector;

[0106] Step S630, inputting the high-dimensional prompt vector into a preset self-attention mechanism network model to perform text semantic analysis and processing to obtain initial text semantic information;

[0107] Step S640, inputting the initial text semantic information into a preset decoder for prediction processing to obtain text prediction information;

[0108] Step S650: converting the text prediction information to obtain system interpretation information.

[0109] It should be noted that in the process of inputting the enhanced query prompt into the preset large language model for analysis and processing to obtain the system interpretation information, the enhanced query prompt is first preprocessed to obtain the encoded prompt information; then the encoded prompt information is converted into a high-dimensional prompt vector; then the high-dimensional prompt vector is input into the preset self-attention mechanism network model for text semantic analysis and processing to obtain the initial text semantic information; then the initial text semantic information is input into the preset decoder for prediction processing to obtain text prediction information; finally, the text prediction information is converted and processed to obtain the system interpretation information. Through the above technical solution, intelligent analysis and interpretation processing can be achieved through the large language model, which can greatly improve the efficiency of system interpretation, and make document interpretation more accurate, bringing a good experience to users.

[0110] It is worth noting that large language models have been a hot topic in the field of artificial intelligence in recent years. They have been trained on a large amount of text data through deep learning technology, especially models based on the Transformer architecture, to understand and generate human language; these models learn the general laws of language in the pre-training stage, and then optimize for specific tasks in the fine-tuning stage. As the size of the model increases, they have significantly improved the fluency and accuracy of the generated text and the ability to understand complex semantics. The application of large language models is very wide, including but not limited to natural language processing tasks such as text classification, machine translation, speech recognition, text generation, information retrieval, etc. They can also support chatbots, intelligent customer service, intelligent dialogue systems, etc., and even play a role in multiple industries such as medical care, finance, and entertainment. For example, AI agents such as smart assistants, AI health butlers, and AI financial butlers have been applied in travel, government affairs, people's livelihood, catering, scenic spots, medical care, finance, and other fields.

[0111] It is worth noting that the self-attention mechanism network model is an important concept in the field of deep learning, which is inspired by the human visual system's selective attention to information. The self-attention mechanism allows the model to dynamically focus on the importance of different parts of the sequence when processing sequence data, thereby improving the efficiency and accuracy of the model in processing information. This mechanism has applications in a variety of tasks, such as image recognition, natural language processing, etc. The core of the self-attention mechanism is to calculate the relationship between each element in the input sequence and all other elements, usually through the interaction of queries, keys, and values. In natural language processing, the self-attention mechanism can help the model understand the dependencies between words, especially in the case of long-distance dependencies. It can effectively capture and utilize these features by calculating the direct relationship between any two positions in the sequence.

[0112] In some embodiments, Figure 7As shown, the above step S610 may also include but is not limited to steps S611 to S613.

[0113] Step S611, performing word segmentation processing on the enhanced query prompt to obtain multiple query prompt segments;

[0114] Step S612, cleaning the multiple query prompt segments to obtain multiple sorted prompt segments;

[0115] Step S613: performing encoding conversion processing on the multiple sorting prompt segments to obtain encoding prompt information.

[0116] It should be noted that in the process of preprocessing the enhanced query prompt to obtain the encoded prompt information, the enhanced query prompt is first segmented to obtain multiple query prompt fragments; then the multiple query prompt fragments are cleaned to obtain multiple sorted prompt fragments; finally, the multiple sorted prompt fragments are encoded and converted to obtain the encoded prompt information; through the above technical solution, the prerequisite preparation for obtaining a high-dimensional prompt vector can be made.

[0117] In addition, if Figure 8 As shown, an embodiment of the present application further provides a document interpretation device 10, comprising:

[0118] The first processing module 100 is used to obtain the system document;

[0119] The second processing module 200 is used to analyze and process the system document to obtain system attribute information;

[0120] The third processing module 300 is used to match the system attribute information with the preset business category knowledge base to determine the system type;

[0121] The fourth processing module 400 is used to extract historical system information from the business category knowledge base according to the system type;

[0122] The fifth processing module 500 is used to generate an initial interpretation query according to the system document, and enhance the initial interpretation query according to the historical system information to obtain an enhanced query prompt;

[0123] The sixth processing module 600 is used to input the enhanced query prompt into a preset large language model for analysis and processing to obtain system interpretation information.

[0124] It should be noted that in the process of document interpretation, first obtain the system document; then analyze and process the system document to obtain the system attribute information; then match the system attribute information with the preset business category knowledge base to determine the system type; then extract the historical system information from the business category knowledge base according to the system type; then generate the initial interpretation query according to the system document, and enhance the initial interpretation query according to the historical system information to obtain the enhanced query prompt; finally, input the enhanced query prompt into the preset large language model for analysis and processing to obtain the corresponding system interpretation information. Through the above method, intelligent document interpretation can be realized, the efficiency of document interpretation is improved, and it is more conducive to the development of banking business.

[0125] The specific implementation of the document interpretation device 10 is substantially the same as the specific implementation of the document interpretation method described above, and will not be described in detail herein.

[0126] In addition, if Fig. 9 As shown, an embodiment of the present application further provides an electronic device 700 , which includes: a memory 720 , a processor 710 , and a computer program stored in the memory 720 and executable on the processor 710 .

[0127] The processor 710 and the memory 720 may be connected via a bus or in other ways.

[0128] The non-transitory software programs and instructions required to implement the document interpretation methods of the above embodiments are stored in the memory 720 , and when executed by the processor 710 , the document interpretation methods of the above embodiments are executed.

[0129] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0130] In addition, an embodiment of the present application further provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are executed by a processor 710 or a controller, for example, by a processor 710 in the above-mentioned device embodiment, so that the above-mentioned processor 710 can execute the document interpretation method in the above-mentioned embodiment.

[0131] The above embodiments may be used in combination, and modules with the same name in different embodiments may be the same or different.

[0132] The above describes specific embodiments of the present application, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily have to be performed in the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0133] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and computer-readable storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0134] The apparatus, device, computer-readable storage medium and method provided in the embodiments of the present application correspond to each other. Therefore, the apparatus, device and non-volatile computer storage medium also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the corresponding apparatus, device and computer storage medium will not be repeated here.

[0135] In the 1990s, it was very clear whether the improvement of a technology was hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented by hardware entity modules. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to ask chip manufacturers to design and make dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs. The original code before compilation must also be written in a specific programming language, which is called Hardware Description Language (HDL). There is not only one kind of HDL, but many kinds, such as ABEL (Advanced Bootean Express Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell Unity Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used one is VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language). Those skilled in the art should also be aware that by programming the method flow in the above-mentioned hardware description languages ​​and programming it into an integrated circuit, a hardware circuit for implementing the logic method flow can be easily obtained.

[0136] The controller can be implemented in any appropriate manner, for example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) that can be executed by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91 SAM, Microchip PIC18F26K20, and Silicon Labs C8051 F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in a purely computer-readable program code manner, the controller can be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, this controller can be considered as a hardware component, and the devices included therein for implementing various functions can also be regarded as structures within the hardware component. Or even, the means for realizing various functions may be regarded as both a software module for realizing the method and a structure within a hardware component.

[0137] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0138] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing the embodiments of the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0139] 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 embodiments of the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0140] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0141] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0143] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0144] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash memory (Flash RAM). The memory is an example of a computer-readable medium.

[0145] Computer readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (Transitory Med ia), such as modulated data signals and carrier waves.

[0146] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0147] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c can be single or multiple.

[0148] Embodiments of the present application may be described in the general context of computer executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. Embodiments of the present application may also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0149] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0150] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A document interpretation method, characterized in that: The method comprises: Obtaining system documents; Analyzing and processing the system document to obtain system attribute information; Matching the system attribute information with a preset business category knowledge base to determine the system type; extracting historical system information from the business category knowledge base according to the system type; generating an initial interpretation query according to the system document, and enhancing the initial interpretation query according to the historical system information to obtain an enhanced query prompt; The enhanced query prompt is input into a preset large language model for analysis and processing to obtain system interpretation information.

2. The document interpretation method according to claim 1, characterized in that: The analyzing and processing of the system document to obtain system attribute information includes: Determine attribute keywords based on a preset attribute database; The system attribute information is obtained by querying the system document according to the attribute keywords.

3. The document interpretation method according to claim 1, characterized in that: The matching of the system attribute information with a preset business category knowledge base to determine the system type includes: Performing information splitting processing on the system attribute information to obtain multiple basic attributes; The system type of the system document is obtained by matching a plurality of the basic attributes with the business category knowledge base.

4. The document interpretation method according to claim 1, characterized in that: The business category knowledge base includes a plurality of original system information, and each of the original system information corresponds to a type information. The extracting of historical system information from the business category knowledge base according to the system type includes: Matching the system type with the type information of the original system information to obtain the target original system information; The target original system information is determined as the historical system information.

5. The document interpretation method according to claim 1, characterized in that: The step of performing enhancement processing on the initial interpretation query according to the historical system information to obtain an enhanced query prompt includes: fusing the historical system information with the initial interpretation query to obtain context information; Compressing the context information to obtain an initial query prompt; The initial query prompt is combined with preset query guidance information to obtain the enhanced query prompt.

6. The document interpretation method according to claim 1, characterized in that: The enhanced query prompt is input into a preset large language model for analysis and processing to obtain system interpretation information, including: Preprocessing the enhanced query prompt to obtain coded prompt information; Converting the encoded prompt information into a high-dimensional prompt vector; Inputting the high-dimensional prompt vector into a preset self-attention mechanism network model for text semantic analysis and processing to obtain initial text semantic information; Inputting the initial text semantic information into a preset decoder for prediction processing to obtain text prediction information; The text prediction information is converted to obtain the system interpretation information.

7. The document interpretation method according to claim 6, characterized in that: The preprocessing of the enhanced query prompt to obtain the coded prompt information includes: Performing word segmentation processing on the enhanced query prompt to obtain multiple query prompt segments; Cleaning the plurality of query prompt segments to obtain a plurality of sorted prompt segments; The coded prompt information is obtained by performing coding conversion processing on the plurality of the sorting prompt segments.

8. A document interpretation device, characterized in that: include: The first processing module is used to obtain system documents; The second processing module is used to analyze and process the system document to obtain system attribute information; The third processing module is used to match the system attribute information with a preset business category knowledge base to determine the system type; A fourth processing module, configured to extract historical system information from the business category knowledge base according to the system type; A fifth processing module, used for generating an initial interpretation query according to the system document, and performing enhancement processing on the initial interpretation query according to the historical system information to obtain an enhanced query prompt; The sixth processing module is used to input the enhanced query prompt into a preset large language model for analysis and processing to obtain system interpretation information.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the document interpretation method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing computer-executable instructions, characterized in that: The computer executable instructions are used to execute the document interpretation method described in any one of claims 1 to 7.

Images