Question answering method and system based on knowledge graph and large language model

Abstract

This embodiment discloses a question-answering method and system based on a knowledge graph and a large language model. Among them, the method includes: obtaining a directional data set corresponding to the application requirements of the target discipline or target industry; processing the directional data in the directional data set to obtain the recognition results and classification results of the artificial intelligence application scenario; based on the recognition results and classification results of the artificial intelligence application scenario, construct a knowledge graph of artificial intelligence application scenarios that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment; combining the large-scale language model with the knowledge graph to obtain an updated large-scale language model; inputting the user's query request into the updated large-scale language model to obtain the question-answering results output by the model. This method can enhance the user experience and improve the efficiency of user interaction, so that non-technical users can easily apply the question-answering system, reducing the difficulty of use for non-technical users.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a question-answering method and system based on a knowledge graph and a large language model. Background Art

[0002] The rapid development of artificial intelligence technology has brought the potential for change to multiple disciplines and industries. Although artificial intelligence technology has shown great innovation potential in interdisciplinary applications, its complexity and professionalism often become obstacles for non-technical personnel to understand and apply it. Professionals in many fields lack a deep understanding of the working principles and implementation methods of artificial intelligence, which limits the potential of artificial intelligence technology in interdisciplinary innovation.

[0003] In related technologies, question-answering systems based on artificial intelligence technology face significant challenges when processing massive amounts of data. The knowledge graph of the question-answering system expresses limited types of knowledge, and its construction and maintenance usually require a lot of expertise and manual intervention, which results in the question-answering system's processing capabilities being inefficient in practical applications and unable to effectively convert complex query questions, thus affecting user experience and interaction efficiency, and increasing the difficulty of use for non-technical users. Summary of the invention

[0004] In view of this, the embodiments of the present disclosure provide a question-answering method and system based on a knowledge graph and a large language model, which can effectively integrate the knowledge graph and the large-scale language model into a cross-industry and cross-disciplinary intelligent question-answering system, which can not only provide users with powerful knowledge understanding and generation capabilities, realize the customized application of artificial intelligence to specific industries and specific disciplines, but also provide users with existing cases of artificial intelligence empowering industries and disciplines. It can also infer the exploration and discovery of new scenarios in the cross-industry and cross-disciplinary empowerment of artificial intelligence, as well as the exploration and discovery of new knowledge and new models through the relationship between the existing massive industries, massive disciplines and artificial intelligence, thereby enhancing the user experience and improving the efficiency of user interaction, so that non-technical users can easily apply the question-answering system, reducing the difficulty of use for non-technical users.

[0005] In a first aspect, the embodiments of the present disclosure provide a question-answering method based on a knowledge graph and a large language model, which adopts the following technical solutions:

[0006] Obtain targeted data sets corresponding to the application requirements of the target discipline or target industry;

[0007] Processing the directional data in the directional data set based on a deep learning model and pattern recognition technology to obtain recognition results and classification results of artificial intelligence application scenarios;

[0008] Based on the recognition results and classification results of the artificial intelligence application scenario, a knowledge graph of the artificial intelligence application scenario is constructed, which includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment;

[0009] Combining the large-scale language model with the knowledge graph to obtain an updated large-scale language model;

[0010] The user's query request is input into the updated large-scale language model to obtain a question-answering result output by the updated large-scale language model.

[0011] In some embodiments, after obtaining a directional data set of application requirements corresponding to a target discipline or target industry, the method further includes:

[0012] Determining sensitive data based on the directional data in the directional data set;

[0013] The sensitive data is desensitized to obtain desensitized data corresponding to the sensitive data.

[0014] In some embodiments, based on the recognition results and classification results of the artificial intelligence application scenario, a knowledge graph of the artificial intelligence application scenario is constructed, which includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment, including:

[0015] Text analysis technology based on large-scale language models to extract multiple entities and relations from the text to be extracted;

[0016] Obtaining the relevance of a plurality of the entities in different target disciplines or target industries;

[0017] The knowledge graph is constructed based on the recognition results and classification results of the artificial intelligence application scenario, as well as the entities, relationships and correlations between multiple entities.

[0018] In some embodiments, the large-scale language model is combined with the knowledge graph to obtain an updated large-scale language model, including:

[0019] Acquire implicit knowledge of the large-scale language model and explicit relationships of the knowledge graph;

[0020] Merging the implicit knowledge of the large-scale language model with the explicit relationships of the knowledge graph;

[0021] The graph neural network model is used to optimize the embedding of entities and relationships in the knowledge graph to obtain an updated large-scale language model.

[0022] In some embodiments, the method further comprises:

[0023] According to the received user customization request for the user interface, obtaining the interface configuration parameters carried in the customization request;

[0024] Based on the interface configuration parameters, the data query function or business information display of the user interface customization template is revised to obtain the target user interface.

[0025] In some embodiments, the method further comprises:

[0026] Receiving a trigger operation from a user on a feedback button on the target user interface;

[0027] According to the trigger operation, obtaining user satisfaction feedback information on the question and answer result;

[0028] Based on the satisfaction feedback information, sorting the preset question and answer results in the updated large-scale language model;

[0029] Generate iterative data or optimized data according to the satisfaction feedback information;

[0030] The iterative data or the optimized data is used to perform corresponding iterative operations or optimization operations on the large-scale language model.

[0031] In some embodiments, the method further comprises:

[0032] Use the reinforcement learning algorithm to train the initial reward model to obtain a trained reward model;

[0033] Inputting at least two question-answering results output by the updated large-scale language model into the trained reward model;

[0034] Comparing and evaluating the at least two question-answering results using the trained reward model to obtain an evaluation result;

[0035] Converting the evaluation result into a reward value of the trained reward model;

[0036] Based on the reward value, the reinforcement learning algorithm is used to adjust the parameters in the updated large-scale language model, and the updated large-scale language model is continued to be trained to optimize the question and answer result output by the updated large-scale language model.

[0037] In a second aspect, the embodiment of the present disclosure further provides a question-answering system based on a knowledge graph and a large language model, which adopts the following technical solutions:

[0038] An acquisition unit is configured to acquire a directional data set corresponding to application requirements of a target discipline or a target industry;

[0039] A recognition and classification unit is configured to process the directional data in the directional data set based on a deep learning model and pattern recognition technology to obtain a recognition result and a classification result of an artificial intelligence application scenario;

[0040] A construction unit is configured to construct a knowledge graph of the artificial intelligence application scenario that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment based on the recognition result and classification result of the artificial intelligence application scenario;

[0041] An integration unit, configured to combine the large-scale language model with the knowledge graph to obtain an updated large-scale language model;

[0042] The question-answering unit is configured to input the user's query request into the updated large-scale language model to obtain the question-answering result output by the updated large-scale language model.

[0043] In a third aspect, the embodiments of the present disclosure further provide a computer device, which adopts the following technical solution:

[0044] The computer device comprises:

[0045] at least one processor; and,

[0046] a memory communicatively connected to the at least one processor; wherein,

[0047] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any of the above-mentioned question-answering methods based on knowledge graphs and large language models.

[0048] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute any of the above-mentioned question-answering methods based on knowledge graphs and large language models.

[0049] In a fifth aspect, an embodiment of the present disclosure further provides a computer program product, including a computer program / instruction, which implements the steps of any of the above methods when executed by a processor.

[0050] The question-answering method based on knowledge graph and large language model provided in the embodiment of the present disclosure processes the directional data in the directional data set through deep learning model and pattern recognition technology to obtain recognition results and classification results of artificial intelligence application scenarios, and constructs a knowledge graph of artificial intelligence application scenarios that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and cross-empowerment of artificial intelligence. The knowledge graph and large-scale language model can be effectively integrated into a cross-industry and cross-disciplinary intelligent question-answering system, which can not only provide users with powerful knowledge understanding and generation capabilities, realize customized application of artificial intelligence to specific industries and specific disciplines, but also provide users with existing cases of artificial intelligence empowering industries and empowering disciplines. In addition, through the association between the existing massive industries, massive disciplines and artificial intelligence, it can infer the exploration and discovery of new scenarios in the cross-industry and cross-disciplinary empowerment of artificial intelligence, as well as the exploration and discovery of new knowledge and new models, thereby enhancing the user experience and improving the efficiency of user interaction, so that non-technical users can easily apply the question-answering system generated by the question-answering method, reducing the difficulty of use for non-technical users.

[0051] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the following preferred embodiments are specifically cited and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0053] Figure 1 A flowchart of a question-answering method based on a knowledge graph and a large language model provided in an embodiment of the present disclosure;

[0054] Figure 2 A schematic diagram of the structure of a question-answering system based on a knowledge graph and a large language model provided in an embodiment of the present disclosure;

[0055] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0056] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0057] It should be clear that the following embodiments of the present disclosure are described by specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other in the absence of conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0058] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0059] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0060] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.

[0061] like Figure 1 As shown, Figure 1 A flowchart of a question-answering method based on a knowledge graph and a large language model provided in an embodiment of the present disclosure is provided. A question-answering method based on a knowledge graph and a large language model provided in an embodiment of the present disclosure includes the following steps:

[0062] S101. Obtain a targeted data set corresponding to application requirements of a target discipline or target industry.

[0063] For example, the target disciplines and target industries include healthcare, education and training, financial technology services, intelligent manufacturing, etc. Qualitative data collection is performed on the above-mentioned target disciplines or target industries to obtain a targeted data set to ensure that the data relevance and data quality meet the application requirements of the embodiments of the present disclosure.

[0064] S102. Process the directional data in the directional data set based on the deep learning model and pattern recognition technology to obtain recognition results and classification results of artificial intelligence application scenarios.

[0065] Optionally, deep learning models and pattern recognition technology are used to automatically analyze targeted data to identify and classify potential application scenarios of artificial intelligence. For example, the identified and classified artificial intelligence application scenarios may include intelligent diagnosis, risk assessment, personalized education, etc.

[0066] S103. Based on the recognition and classification results of artificial intelligence application scenarios, construct a knowledge graph of artificial intelligence application scenarios that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and cross-empowers artificial intelligence.

[0067] For example, construct a knowledge graph of AI application scenarios that cross-empowers the subject knowledge of multiple target disciplines with AI. Another example is to construct a knowledge graph of AI application scenarios that cross-empowers the industry knowledge of multiple target industries with AI. Another example is to construct a knowledge graph of AI application scenarios that cross-empowers the subject knowledge of multiple target disciplines and the industry knowledge of multiple target industries with AI.

[0068] By constructing a knowledge graph of AI application scenarios that can reflect the subject knowledge of multiple target disciplines or the industry knowledge of multiple target industries and cross-empower AI, it is possible to visualize specific scenarios and solutions for how AI technology can assist other fields.

[0069] S104. Combine the large-scale language model and the knowledge graph to obtain an updated large-scale language model.

[0070] By combining large-scale language models with knowledge graphs and optimizing the embedding of entities and relationships in knowledge graphs through technical means such as graph neural network models, the semantic parsing capabilities of large-scale language models can be enhanced.

[0071] S105: Input the user's query request into the updated large-scale language model to obtain the question-answering result output by the updated large-scale language model.

[0072] When a user has a question-and-answer need, the question-and-answer method of the embodiment of the present disclosure can integrate the knowledge graph and the large-scale language model to automatically generate case studies describing the innovative application of artificial intelligence technology in specific fields.

[0073] The question-answering method based on knowledge graph and large language model provided in the embodiment of the present disclosure processes the directional data in the directional data set through deep learning model and pattern recognition technology to obtain recognition results and classification results of artificial intelligence application scenarios, and constructs a knowledge graph of artificial intelligence application scenarios that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and cross-empowerment of artificial intelligence. The knowledge graph and large-scale language model can be effectively integrated into a cross-industry and cross-disciplinary intelligent question-answering system, which can not only provide users with powerful knowledge understanding and generation capabilities, realize customized application of artificial intelligence to specific industries and specific disciplines, but also provide users with existing cases of artificial intelligence empowering industries and empowering disciplines. In addition, through the association between the existing massive industries, massive disciplines and artificial intelligence, it can infer the exploration and discovery of new scenarios in the cross-industry and cross-disciplinary empowerment of artificial intelligence, as well as the exploration and discovery of new knowledge and new models, thereby enhancing the user experience and improving the efficiency of user interaction, so that non-technical users can easily apply the question-answering system generated by the question-answering method, reducing the difficulty of use for non-technical users.

[0074] In some embodiments, after obtaining a directional data set of application requirements corresponding to a target discipline or target industry, the method further includes:

[0075] Determine sensitive data based on the targeted data in the targeted data set;

[0076] The sensitive data is desensitized to obtain desensitized data corresponding to the sensitive data.

[0077] The embodiments of the present disclosure mainly realize how artificial intelligence empowers scenarios. Therefore, in the process of targeted data collection and analysis involving sensitive data, there is no need to involve some privacy information. Therefore, the sensitive data can be desensitized by anonymization and other desensitization processing to obtain desensitized data corresponding to the sensitive data. For example, anonymized data can be obtained by anonymizing sensitive data to ensure the security of user data and user privacy.

[0078] In some embodiments, based on the recognition results and classification results of artificial intelligence application scenarios, a knowledge graph of artificial intelligence application scenarios is constructed that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment, including:

[0079] Text analysis technology based on large-scale language models to extract multiple entities and relations from the text to be extracted;

[0080] Capture the relevance of multiple entities in different target disciplines or target industries;

[0081] Construct a knowledge graph based on the recognition and classification results of artificial intelligence application scenarios, as well as the correlation between entities, relationships and multiple entities.

[0082] The disclosed embodiments perform in-depth mining of entities and relationships through advanced text analysis technology based on large-scale language models, which can not only extract basic entities and relationships from the text to be extracted, but also obtain the interactions and influences of these entities in different target disciplines and target industries, thereby improving the knowledge understanding and generation capabilities of the knowledge graph, as well as improving the efficiency and accuracy of users' question-answering during the question-answering process.

[0083] In some embodiments, the large-scale language model is combined with the knowledge graph to obtain an updated large-scale language model, including:

[0084] Capture implicit knowledge from large-scale language models and explicit relationships from knowledge graphs;

[0085] Merge the implicit knowledge of large-scale language models with the explicit relationships of knowledge graphs;

[0086] The graph neural network model is used to optimize the embedding of entities and relations in the knowledge graph to obtain an updated large-scale language model.

[0087] Since the traditional knowledge graph organizes and presents the complex relationships between entities in a graphical way, it explicitly expresses knowledge through symbolic methods and has good semantic clarity and interpretability. However, the types of knowledge expressed by traditional knowledge graphs are limited, and their construction and maintenance usually require a lot of expertise and manual intervention. Large-scale language models can automatically learn and mine knowledge from massive data and store it implicitly in model parameters. They not only cover a wide range of knowledge types (for example, linguistic knowledge, world knowledge, and common sense knowledge, etc.), but also can flexibly respond to the challenges of open domains. Therefore, the disclosed embodiments can not only provide powerful knowledge understanding and generation capabilities by effectively integrating large-scale language models and knowledge graphs, but also can achieve customized applications for specific industries and scenarios through user interactive learning and optimization.

[0088] In some embodiments, the method further comprises:

[0089] According to the received user customization request for the user interface, obtaining the interface configuration parameters carried in the customization request;

[0090] Based on the interface configuration parameters, the data query function or business information display of the user interface customization template is revised to obtain the target user interface.

[0091] The disclosed embodiments also provide an interactive interface customization solution, that is, providing a customizable user interface to support users in various industries to configure information display and data query functions as required. For example, users can add natural language interaction functions or graphical interface interaction functions to existing user interface customization templates, thereby enhancing user experience and improving user interaction efficiency, allowing non-technical users to easily apply the question-and-answer system and reducing the difficulty of use for non-technical users.

[0092] For example, in the medical field, non-technical medical professionals can use the question-and-answer system to query and obtain information about the application of artificial intelligence in disease diagnosis and treatment; in the education field, teachers and course designers can use the question-and-answer system to explore the possibilities of artificial intelligence in personalized teaching and learning assessment.

[0093] Through this question-and-answer system, users can not only gain an in-depth understanding of artificial intelligence technology, but also actually participate in the creation and optimization of artificial intelligence solutions, which greatly reduces the application threshold of artificial intelligence technology and accelerates its innovative application in various industries and disciplines. In addition, with the advancement of technology and the continuous optimization of models, the question-and-answer system of the disclosed embodiment is expected to be more widely used in more fields, promoting the popularization of artificial intelligence technology and industry innovation.

[0094] In some embodiments, the method further comprises:

[0095] Receiving a trigger operation of a user on a feedback button on a target user interface;

[0096] According to the triggering operation, obtain the user's satisfaction feedback information on the question and answer results;

[0097] Based on satisfaction feedback information, the preset question-and-answer results in the updated large-scale language model are ranked;

[0098] Generate iterative data or optimization data based on satisfaction feedback information;

[0099] Using the iterative data or the optimized data, corresponding iterative operations or optimization operations are performed on the large-scale language model.

[0100] The embodiment of the present disclosure can obtain satisfaction feedback information submitted by the user on the target user interface through a feedback button triggered by the user on the target user interface, sort the preset question and answer results in the large-scale language model according to the satisfaction feedback information, and use the satisfaction feedback information to iterate or optimize the large-scale language model, thereby achieving the purpose of optimizing the first question and answer result generated by the question and answer method of the embodiment of the present disclosure.

[0101] In some embodiments, the method further comprises:

[0102] Use the reinforcement learning algorithm to train the initial reward model to obtain a trained reward model;

[0103] Input at least two question-answer results output by the updated large-scale language model into the trained reward model;

[0104] Use the trained reward model to compare and evaluate at least two question-answering results to obtain an evaluation result;

[0105] Convert the evaluation results into reward values ​​for the trained reward model;

[0106] Based on the reward value, the reinforcement learning algorithm is used to adjust the parameters in the updated large-scale language model, and the updated large-scale language model is continued to be trained to optimize the question and answer results output by the updated large-scale language model.

[0107] The disclosed embodiment takes into account that the question-and-answer results generated by the large-scale language model may not fully meet the specific needs of the user, and therefore introduces a trained reward model obtained by model training based on a reinforcement learning algorithm. At least two question-and-answer results generated by the large-scale language model, for example, the first question-and-answer result and the second question-and-answer result, are compared and evaluated by the trained reward model to obtain an evaluation result, and the obtained evaluation result is converted into a reward value of the trained reward model. Finally, the parameters of the updated large-scale language model are fine-tuned using reinforcement learning technology to optimize the question-and-answer results generated by the updated large-scale language model, so as to maximize the reward function that meets the user's evaluation, thereby realizing continuous self-evaluation and optimization of the large-scale language model.

[0108] The question-answering method of the disclosed embodiment strengthens the application potential of artificial intelligence technology in different industries and disciplines by intelligently integrating and utilizing interdisciplinary knowledge, integrating knowledge graphs and large-scale language models, and not only provides a powerful knowledge and data processing tool for scientific researchers and industry experts, but also opens the door for non-professional users to understand and apply artificial intelligence technology, enabling them to intuitively explore the potential value and application prospects of artificial intelligence. Similarly, the development and application of the question-answering system generated by the question-answering method of the disclosed embodiment will greatly promote the popularization and innovation of artificial intelligence technology, especially in the field of cross-empowerment scenarios of artificial intelligence, so that non-technical users can easily apply the question-answering system generated by the question-answering method, reducing the difficulty of use for non-technical users.

[0109] Figure 2 A schematic diagram of the structure of a question-answering system based on a knowledge graph and a large language model provided in an embodiment of the present disclosure. The embodiment of the present disclosure also provides a question-answering system based on a knowledge graph and a large language model, including:

[0110] An acquisition unit 21 is configured to acquire a directional data set corresponding to an application requirement of a target discipline or a target industry;

[0111] The recognition and classification unit 22 is configured to process the directional data in the directional data set based on the deep learning model and pattern recognition technology to obtain the recognition result and classification result of the artificial intelligence application scenario;

[0112] A construction unit 23 is configured to construct a knowledge graph of artificial intelligence application scenarios that includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment based on the recognition results and classification results of the artificial intelligence application scenarios;

[0113] An integration unit 24 is configured to combine the large-scale language model with the knowledge graph to obtain an updated large-scale language model;

[0114] The question-answering unit 25 is configured to input the user's query request into the updated large-scale language model to obtain the question-answering result output by the updated large-scale language model.

[0115] The computer device according to the embodiment of the present disclosure includes a memory and a processor. The memory is used to store non-temporary computer-readable instructions. Specifically, the memory may include one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc.

[0116] The processor may be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and may control other components in the computer device to perform desired functions. In one embodiment of the present disclosure, the processor is used to run the computer-readable instructions stored in the memory, so that the computer device executes all or part of the steps of a question-answering method based on a knowledge graph and a large language model in the aforementioned embodiments of the present disclosure.

[0117] Those skilled in the art should be able to understand that in order to solve the technical problem of how to obtain a good user experience, the present embodiment may also include well-known structures such as a communication bus and an interface, and these well-known structures should also be included in the protection scope of the present disclosure.

[0118] like Figure 3A schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure is shown, which is a schematic diagram of the structure of a computer device suitable for implementing the embodiment of the present disclosure. Figure 3 The computer device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0119] like Figure 3 As shown, the computer device may include a processor (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage device into a random access memory (RAM). In the RAM, various programs and data required for the operation of the computer device are also stored. The processor, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0120] Typically, the following devices can be connected to the I / O interface: input devices such as sensors or visual information acquisition devices; output devices such as display screens; storage devices such as tapes, hard disks, etc.; and communication devices. The communication device can allow the computer device to communicate with other devices (such as edge computing devices) wirelessly or by wire to exchange data. Figure 3 A computer device having various devices is shown, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead.

[0121] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processor, all or part of the steps of a question-answering method based on a knowledge graph and a large language model of an embodiment of the present disclosure are executed.

[0122] For detailed description of this embodiment, reference may be made to the corresponding descriptions in the aforementioned embodiments, which will not be repeated here.

[0123] According to a computer-readable storage medium of an embodiment of the present disclosure, non-transitory computer-readable instructions are stored thereon. When the non-transitory computer-readable instructions are executed by a processor, all or part of the steps of a question-answering method based on a knowledge graph and a large language model of each embodiment of the present disclosure are executed.

[0124] The above-mentioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or mobile hard disk), media with built-in rewritable non-volatile memory (e.g., memory card) and media with built-in ROM (e.g., ROM box).

[0125] For detailed description of this embodiment, reference may be made to the corresponding descriptions in the aforementioned embodiments, which will not be repeated here.

[0126] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0127] In the present disclosure, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The block diagrams of the devices, devices, equipment, and systems involved in the present disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.

[0128] Additionally, as used herein, "or" used in a list of items beginning with "at least one" indicates a separate list, so that, for example, a list of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the example described is preferred or better than other examples.

[0129] It should also be noted that in the system and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0130] Various changes, substitutions, and modifications of the techniques described herein may be made without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of the present disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of events, means, methods, and actions described above. Currently existing or later to be developed processes, machines, manufactures, compositions of events, means, methods, or actions that perform substantially the same functions or achieve substantially the same results as the corresponding aspects described herein may be utilized. Thus, the appended claims include such processes, machines, manufactures, compositions of events, means, methods, or actions within their scope.

[0131] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0132] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A question-answering method based on knowledge graph and large language model, characterized in that: include: Obtain targeted data sets corresponding to the application requirements of the target discipline or target industry; Processing the directional data in the directional data set based on a deep learning model and pattern recognition technology to obtain recognition results and classification results of artificial intelligence application scenarios; Based on the recognition results and classification results of the artificial intelligence application scenario, a knowledge graph of the artificial intelligence application scenario is constructed, which includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment; Among them, based on the recognition results and classification results of the artificial intelligence application scenario, a knowledge graph of the artificial intelligence application scenario is constructed, which includes subject knowledge of multiple target disciplines or industry knowledge of multiple target industries and artificial intelligence for cross-empowerment, including: Text analysis technology based on large-scale language models to extract multiple entities and relations from the text to be extracted; Obtaining the relevance of a plurality of the entities in different target disciplines or target industries; Based on the recognition results and classification results of the artificial intelligence application scenario, as well as the entities, relationships and correlations between the entities, the knowledge graph is constructed; Combining the large-scale language model with the knowledge graph to obtain an updated large-scale language model; The large-scale language model is combined with the knowledge graph to obtain an updated large-scale language model, including: Acquire implicit knowledge of the large-scale language model and explicit relationships of the knowledge graph; Merging the implicit knowledge of the large-scale language model with the explicit relationships of the knowledge graph; Optimizing the embedding of entities and relationships in the knowledge graph using a graph neural network model to obtain an updated large-scale language model; The user's query request is input into the updated large-scale language model to obtain a question-answering result output by the updated large-scale language model.

2. The question-answering method based on knowledge graph and large language model according to claim 1, characterized in that: After obtaining a directional data set corresponding to the application requirements of the target discipline or target industry, the method further includes: Determining sensitive data based on the directional data in the directional data set; The sensitive data is desensitized to obtain desensitized data corresponding to the sensitive data.

3. The question-answering method based on knowledge graph and large language model according to claim 1, characterized in that: The method further comprises: According to the received user customization request for the user interface, obtaining the interface configuration parameters carried in the customization request; Based on the interface configuration parameters, the data query function or business information display of the user interface customization template is revised to obtain the target user interface.

4. The question-answering method based on knowledge graph and large language model according to claim 3, characterized in that: The method further comprises: Receiving a trigger operation from a user on a feedback button on the target user interface; According to the trigger operation, obtaining user satisfaction feedback information on the question and answer result; Based on the satisfaction feedback information, sorting the preset question and answer results in the updated large-scale language model; Generate iterative data or optimized data according to the satisfaction feedback information; The iterative data or the optimized data is used to perform corresponding iterative operations or optimization operations on the large-scale language model.

5. The question-answering method based on knowledge graph and large language model according to claim 1, characterized in that: The method further comprises: Use the reinforcement learning algorithm to train the initial reward model to obtain a trained reward model; Inputting at least two question-answering results output by the updated large-scale language model into the trained reward model; Comparing and evaluating the at least two question-answering results using the trained reward model to obtain an evaluation result; Converting the evaluation result into a reward value of the trained reward model; Based on the reward value, the reinforcement learning algorithm is used to adjust the parameters in the updated large-scale language model, and the updated large-scale language model is continued to be trained to optimize the question and answer result output by the updated large-scale language model.

6. A computer device, characterized in that: The computer device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the question-answering method based on the knowledge graph and the large language model as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which are used to enable a computer to execute the question-answering method based on a knowledge graph and a large language model as described in any one of claims 1 to 5.

8. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images