Knowledge base generation method and device, electronic equipment and medium
By obtaining, reviewing and structuring knowledge data, and using large models to update and search the knowledge base, the problems of untimely update and unclear classification in the traditional knowledge base generation method are solved, and the frequency of updates and management efficiency of the knowledge base are improved. It realizes the need for enterprises to quickly acquire and apply knowledge.
Patent Information
- Application Number
- CN202411194175.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-28
- Publication Date
- 2024-11-15
AI Technical Summary
The traditional knowledge base generation method has problems such as untimely update, unclear knowledge classification, and difficulty in retrieval, which is difficult to meet the needs of enterprises to quickly acquire and apply knowledge.
By obtaining knowledge data, reviewing and structured processing according to preset rules, updating and retrieving data classification and large models, ensuring the timeliness and accuracy of the knowledge base and providing user interaction and decision-making support.
It improves the update frequency and management efficiency of the knowledge base, provides an accurate and convenient knowledge base usage experience, and meets the needs of enterprises to quickly acquire and apply knowledge.
Smart Images

Figure CN118966346A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a knowledge base generation method, device, electronic device and medium. Background Art
[0002] With the expansion of enterprise scale and diversification of business, a large amount of knowledge resources have been accumulated within the enterprise. However, the traditional knowledge base generation method has problems such as untimely updates, unclear knowledge classification, and difficult retrieval, which makes it difficult to meet the needs of enterprises to quickly acquire and apply knowledge. Summary of the invention
[0003] The present disclosure provides a knowledge base generation method, device, electronic device and medium to solve the management difficulties caused by the untimely update of the knowledge base, the wide variety of content, the complex knowledge sources and the like in the related art.
[0004] A first aspect of the present disclosure provides a method for generating a knowledge base, the method comprising:
[0005] Acquiring knowledge data;
[0006] Review the knowledge data according to preset rules, and use the data that passes the review in the knowledge data as the first knowledge data;
[0007] Based on the first knowledge data, the second knowledge data is obtained through data classification and structuring, and stored in the target database;
[0008] At a first moment, the original knowledge data is updated using the second knowledge data, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and the update period;
[0009] In response to the user's target interaction instruction, the target database is used as the knowledge base of the large model to output the response information of the target interaction instruction.
[0010] In one embodiment of the present disclosure, based on the first knowledge data, data classification and structuring are performed to obtain the second knowledge data, and the second knowledge data is stored in the target database, including:
[0011] Inferring, according to the pre-trained knowledge classification model, the category of the first knowledge data is the first knowledge category;
[0012] According to the first storage form corresponding to the first knowledge category, the first knowledge data is structured into second knowledge data;
[0013] The second knowledge data is saved in a storage area corresponding to the first storage format in the target database.
[0014] In one embodiment of the present disclosure, the original knowledge data is updated using the second knowledge data at a first moment, including:
[0015] Based on the update cycle, regularly monitor the difference between the second knowledge data and the original knowledge data;
[0016] If the difference is less than or equal to the preset difference threshold, the second knowledge data is used to replace the original knowledge data, and the first knowledge category is set as the knowledge category of the original knowledge data; otherwise, the second knowledge data is saved in the storage area corresponding to the first knowledge category in the target database.
[0017] In one embodiment of the present disclosure, in response to a user's target interaction instruction, the target database is used as a knowledge base of a large model, and response information of the target interaction instruction is output, including:
[0018] In response to the target interaction instruction, using the large model, the semantics of the target interaction instruction is understood as target semantic information;
[0019] Based on the knowledge base, a large model is used to generate response information corresponding to the target semantic information, and the response information is strongly related to the corpus in the knowledge base.
[0020] In one embodiment of the present disclosure, the method further includes:
[0021] In response to a user's search request for target information, generating a target word vector for the target information;
[0022] From the word vector library corresponding to the knowledge base, determine the knowledge data corresponding to the word vector with the highest similarity to the target word vector.
[0023] In one embodiment of the present disclosure, the method further includes:
[0024] Collect statistics on the usage of knowledge base by different users;
[0025] Based on usage data, the usage of the knowledge base is analyzed through a large model to generate decision-making recommendations for the enterprise.
[0026] In one embodiment of the present disclosure, the method further includes:
[0027] In response to the update instruction of the knowledge base, it is checked whether the user authority is at a preset level. If so, the user's knowledge data is classified and structured into second knowledge data, and the knowledge base is updated using the second knowledge data.
[0028] A second aspect of the present disclosure provides a knowledge base generation device, the device comprising:
[0029] An acquisition module, used to acquire knowledge data;
[0030] An audit module, used to audit the knowledge data according to preset rules, and use the data in the knowledge data that pass the audit as the first knowledge data;
[0031] A storage module, used for converting the first knowledge data into second knowledge data through data classification and structuring, and storing the second knowledge data in a target database;
[0032] An updating module, used to update the original knowledge data with the second knowledge data at a first moment, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and the updating period;
[0033] The interaction module is used to respond to the user's target interaction instruction, use the target database as the knowledge base of the large model, and output the response information of the target interaction instruction.
[0034] The third aspect embodiment of the present disclosure proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable 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 one of the methods in the first aspect embodiment of the present disclosure.
[0035] The fourth aspect embodiment of the present disclosure proposes a non-transitory computer-readable storage medium storing computer instructions, characterized in that the computer instructions are used to enable a computer to execute the method in the first aspect embodiment of the present disclosure.
[0036] In summary, according to the knowledge base generation method proposed in the present disclosure, knowledge data is acquired to obtain the original knowledge corpus data; the knowledge data is audited according to preset rules, and the audited data in the knowledge data is used as the first knowledge data, thereby ensuring the reliability of the knowledge data; based on the first knowledge data, the second knowledge data is processed through data classification and structured processing, and stored in the target database, thereby clarifying the category of the knowledge data and improving the data storage efficiency; the original knowledge data is updated at the first moment using the second knowledge data, and the original knowledge data is stored in the target database at the second moment, wherein the first moment is the sum of the second moment and the update cycle, and the knowledge base is updated in time; in response to the user's target interaction instruction, the target database is used as the knowledge base of the large model, and the response information of the target interaction instruction is output, thereby providing the user with an accurate and convenient knowledge base usage experience. Through the knowledge base generation method disclosed in the present disclosure, the update frequency and management efficiency of the knowledge base are improved.
[0037] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.
[0039] Figure 1 A flowchart of a knowledge base generation method according to an embodiment of the present disclosure;
[0040] Figure 2 A flowchart of an embodiment of the present disclosure, based on first knowledge data, through data classification and structuring processing to obtain second knowledge data, and storing it in a target database;
[0041] Figure 3 A flow chart of knowledge update according to an embodiment of the present disclosure;
[0042] Figure 4 A flow chart of knowledge interaction according to an embodiment of the present disclosure;
[0043] Figure 5 A flow chart of knowledge retrieval according to an embodiment of the present disclosure;
[0044] Figure 6 A flow chart of knowledge analysis according to an embodiment of the present disclosure;
[0045] Figure 7 A flowchart of authority management of a knowledge base according to an embodiment of the present disclosure;
[0046] Figure 8 A schematic diagram of a knowledge storage process of an embodiment of the present disclosure;
[0047] Fig. 9 A schematic diagram of knowledge base management according to an embodiment of the present disclosure;
[0048] Fig.10 A schematic diagram of a knowledge base system application architecture according to an embodiment of the present disclosure;
[0049] Fig.11 A schematic diagram of the structure of a knowledge base generating device according to an embodiment of the present disclosure;
[0050] Fig.12 is a block diagram of an electronic device for implementing the disclosed knowledge base generation method according to an exemplary embodiment;
[0051] Fig.13 It is a schematic diagram of the structure of the chip of the embodiment of the present disclosure. DETAILED DESCRIPTION
[0052] Embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout identify the same or similar originals or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.
[0053] Knowledge base plays a vital role in modern information management, both for enterprises and individual users. It can not only improve work efficiency, but also promote knowledge sharing and inheritance. The core of knowledge base lies in its data storage and retrieval functions, allowing users to efficiently find the information they need. At the same time, with the development of technology, the form and function of knowledge base are also constantly evolving. From the initial simple data storage to the current complex knowledge management system, knowledge base has become an indispensable part of the organization.
[0054] One of the main functions of a knowledge base is to serve as a self-service help desk for information. Users can find answers to their questions by consulting FAQs, operating manuals, etc. in the knowledge base without having to contact customer service directly. This not only improves the user experience, but also reduces the pressure on the customer service team. For example, many software companies provide external knowledge bases on their websites where users can get detailed guides on product usage, troubleshooting, etc.
[0055] In addition, the knowledge base is equally important within the enterprise. The internal knowledge base provides employees with a wealth of resources, such as best practice guides, onboarding training materials, technical documentation, etc., so that new employees can get started faster and improve the work efficiency of the entire team. These resources help employees quickly solve problems encountered at work, reduce repetitive work, and devote more energy to core tasks.
[0056] In the related technologies, the enterprise knowledge base is manually collected and built, which is labor-intensive, with untimely knowledge updates, low accuracy, and long time consumption. The traditional knowledge base generation method has problems such as low efficiency, unclear knowledge classification, and difficult retrieval, which makes it difficult to meet the needs of enterprises to quickly acquire and apply knowledge. In other words, the related technologies have problems such as untimely knowledge base updates, a wide variety of content, and complex knowledge sources, which lead to difficult management.
[0057] The present disclosure aims to solve the above-mentioned technical problems and provide a knowledge base generation method, which takes the construction of a high-quality enterprise knowledge base as the starting point and the end point, combines the knowledge base formed by collecting, reviewing, classifying, and regularly updating knowledge with a large model, realizes the efficient collection and classification management of multi-category knowledge, improves the updating frequency and management efficiency of the knowledge base, and thus improves the enterprise knowledge management and utilization efficiency.
[0058] The method proposed in the present disclosure is applied to the task of knowledge base generation, and its application has a rich variety of scenarios, and can be widely used in multiple fields and scenarios such as employee training, customer service, decision support, project management, R&D innovation, etc.
[0059] In the employee training scenario, knowledge standardization can be performed through the knowledge base. By centrally storing and standardizing training materials, rules and regulations, and operating manuals, new employees can quickly understand the company culture, business processes, and job responsibilities. This not only reduces resource dispersion and duplication, but also ensures information consistency and accuracy, thereby increasing the speed of onboarding and reducing training costs. The knowledge base can also be used to evaluate the effectiveness of training. Using the usage data of the knowledge base and employee feedback, companies can objectively evaluate the effectiveness of training. These data help to identify deficiencies in the training process, adjust the training content and methods in a timely manner, and improve the quality of training. At the same time, record employees' learning progress and achievements to provide basic data support for personalized training.
[0060] In customer service scenarios, knowledge bases can be used to quickly respond to needs. Customer service teams can quickly find relevant information through knowledge bases and answer various customer questions, such as FAQs, product usage guides, and troubleshooting steps. This speeds up the resolution of customer problems and improves customer satisfaction. It can also be used for knowledge sharing and collaboration. As an important knowledge sharing platform, knowledge bases allow service personnel to share and exchange experiences and form best practices. This not only improves the team's service level, but also reduces fluctuations in service quality caused by staff turnover.
[0061] In decision support scenarios, knowledge bases can be used for data integration and analysis. Knowledge bases can integrate various business data and market research reports. By analyzing these data, management can obtain comprehensive and accurate information to support decision-making. For example, sales data, customer feedback, and market trends can all be integrated and analyzed through knowledge bases to provide a reliable basis for strategic decision-making. It can also be used for continuous improvement and innovation. Continuously updating and optimizing the knowledge base can achieve continuous improvement and innovation of the enterprise. New ideas and success stories can be quickly shared in the knowledge base for more employees to refer to and learn from, promoting internal knowledge circulation and stimulating innovation capabilities.
[0062] In project management scenarios, knowledge bases can be used for project document management. In project management, a large number of documents, flow charts, and project plans need to be effectively managed. Knowledge bases can centrally store and classify this information, making it convenient for project team members to review and update at any time, such as project launch documents, requirement specifications, and test reports. It can also be used for progress tracking and problem tracking. Through the log function of the knowledge base, project managers can track project progress and problem solving in real time. The work progress and task completion status of each member will be recorded in the knowledge base, which is convenient for team coordination. At the same time, statistical reports are automatically generated to help management understand project progress.
[0063] In R&D innovation scenarios, the knowledge base can be used to summarize technical information. The R&D department needs a large amount of technical information and experimental data. The knowledge base can centrally store this information and provide convenient query and reference for R&D personnel, thereby improving R&D efficiency and innovation capabilities. It can also be used for intellectual property management. R&D results and intellectual property management are also very important. Patent documents, technical secrets, and copyright information can all be effectively managed and protected through the knowledge base to protect the company's intellectual property from leakage and infringement risks.
[0064] In sales and marketing scenarios, the knowledge base can be used for market research and competitive product analysis. Sales and marketing departments can obtain the latest market research reports and competitor analysis through the knowledge base. This can help sales staff develop more effective sales strategies, allowing the marketing team to adjust marketing plans in a timely manner and seize market opportunities. It can also be used for customer relationship management. The knowledge base can record customer purchase history and feedback information. Sales staff can predict purchase tendencies by analyzing customer purchase behavior, develop corresponding sales strategies, and improve sales success rates.
[0065] In administrative and human resources scenarios, knowledge bases can be used for employee self-service. In order to improve efficiency, many companies use knowledge bases to provide employee self-service. Employees can query personnel policies, welfare systems, and business processes through knowledge bases, reducing the burden on the human resources department and improving job satisfaction. It can also be used for recruitment and talent evaluation. The knowledge base stores recruitment and talent evaluation data to provide scientific decision-making support for the human resources department. For example, by analyzing interview records and employee performance data, the human resources department can better identify outstanding talents, formulate reasonable recruitment and training plans, and enhance the company's talent competitiveness.
[0066] In the financial and legal affairs scenario, the knowledge base can be used for financial data management. The financial department involves a large amount of data and reports. Through the knowledge base, this information can be effectively managed, such as the centralized storage of documents such as budget preparation, financial statements and tax declarations, which is convenient for query and analysis and ensures data security. It can also be used for legal document management. The knowledge base is also an important tool for legal affairs management. Documents such as contracts, agreements and legal opinions can be effectively managed and classified through the knowledge base, which facilitates the daily work of the legal department, improves legal compliance and reduces legal risks.
[0067] In the field of supply chain and logistics, knowledge bases can be used for supplier information management. In supply chain management, supplier information and cooperation records are very important. The knowledge base centrally stores all supplier qualification documents, cooperation agreements and evaluation records, which is convenient for the supply chain team to quickly access and manage, ensuring a stable and efficient supply chain. It can also be used for logistics information tracking. Logistics management involves a large amount of transportation and inventory information. The knowledge base provides real-time logistics information tracking services, such as the transportation status of goods, inventory levels and distribution records, etc., to improve the transparency of logistics management and help enterprises control the entire supply chain.
[0068] In summary, the enterprise knowledge base proposed in the present invention can play an important role in multiple fields and scenarios. Through intelligent management and application, the knowledge base can not only significantly improve the internal management and operation efficiency of the enterprise, but also provide customers with better services. Enterprises should make full use of this system, continuously optimize and upgrade the knowledge base, and ensure that they remain competitive in the ever-changing market environment. In the embodiments of the present disclosure, there is no limitation on the application scenario or field.
[0069] The knowledge base generation method provided by the present disclosure is described in detail below with reference to the accompanying drawings.
[0070] Figure 1 FIG. 1 is a flow chart of a method for generating a knowledge base according to an embodiment of the present disclosure. Figure 1 In the embodiment shown, the knowledge base generation method includes:
[0071] Step 101, acquiring knowledge data.
[0072] In this embodiment, knowledge data refers to information and materials related to the business operations of an enterprise, usually covering internal and external data of the enterprise. The form of knowledge data includes various documents and files, such as policy documents, operating manuals, technical documents, etc., as well as audio and video files related to the documents and files in other formats, etc. This disclosure does not limit the form of knowledge data. Knowledge data is collected manually, or the knowledge base automatically collects various types of knowledge documents from the internal or external data of the enterprise, for example, by integrating multiple types of documents from multiple data sources such as internal enterprise rules and regulations, management requirements, product introductions, system manuals, etc., including various structured and unstructured knowledge content such as word, pdf, ppt, excel, etc. Knowledge collection supports both automatic and manual methods. The automatic collection method is achieved by connecting to the internal system of the enterprise, such as OA, etc. Preferably, various types of knowledge documents are automatically collected from internal or external databases of the enterprise.
[0073] Step 102: review the knowledge data according to preset rules, and use the reviewed data in the knowledge data as the first knowledge data.
[0074] In this embodiment, the first knowledge data refers to knowledge data that complies with preset rules. After the knowledge data is collected, a series of predefined rules are applied to ensure the consistency, accuracy, and whether the collected knowledge content contains sensitive information. If the knowledge verification is not compliant, the reason for the failure of the knowledge data is returned to the knowledge collection stage. For manually collected knowledge documents, it is returned to the uploader of the document, and for knowledge automatically collected by the system, it is returned to the uploader of the affiliated department. If the verification passes, it will be further reviewed. In the knowledge review, the knowledge is reviewed manually or automatically, and the knowledge to be stored is reviewed and checked according to the preset review rules, focusing on the scope of knowledge openness, knowledge classification, whether sensitive information is involved, etc., to ensure the reliability and authority of the information.
[0075] Furthermore, when it is difficult to determine whether knowledge data complies with preset rules during automatic review, it is reported to the reviewer for review. If it complies with the review rules after reconfirmation, the knowledge data is used as the first knowledge data. Otherwise, the feedback acquisition knowledge data module reduces the collection amount of the knowledge data, thereby further ensuring the quality and reliability of the knowledge data.
[0076] Step 103: Based on the first knowledge data, the second knowledge data is processed through data classification and structuring, and stored in the target database.
[0077] In this embodiment, the second knowledge data refers to knowledge data with category labels and structure. The target database refers to a database used to store knowledge data for building a knowledge base, usually including a document knowledge base and a question-and-answer knowledge base. The first knowledge data is labeled through a knowledge classification model to indicate the category of the first knowledge data. The first knowledge data is structured so that it can be stored in large quantities or quickly identified in the target database. For example, various types of contract documents of an enterprise are classified into legal category files, and the contents are parsed and saved in a data table of the database dedicated to storing legal label files.
[0078] Step 104, at a first moment, the original knowledge data is updated using the second knowledge data, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and the update period.
[0079] In this embodiment, the first moment refers to the update time of the knowledge base. The second moment refers to the update time of the knowledge base earlier than the first moment. The update cycle refers to the time between two consecutive updates of the knowledge base, and the sum of the second moment and the update cycle equals the first moment. The original knowledge data refers to the knowledge data stored in the knowledge base before the knowledge base is updated at the first moment. At the first moment, the original knowledge data is updated using the second knowledge data, that is, the knowledge base is updated regularly to ensure the timeliness of the knowledge base.
[0080] Step 105, in response to the user's target interaction instruction, the target database is used as the knowledge base of the large model to output the response information of the target interaction instruction.
[0081] In this embodiment, the target interaction instruction refers to the user's operation instruction, such as a retrieval instruction and a conversation instruction. The target interaction instruction supports multiple interaction modes such as text input, voice input, and action gesture input. A large model refers to a machine learning model with large-scale parameters and complex computing structures. These models are usually built from deep neural networks and have billions or even hundreds of billions of parameters. The response information refers to the target interaction instruction for the user. The large model uses the target database as the knowledge base and outputs the answer information. The storage of the target database forms a document knowledge base and a question-and-answer knowledge base. The two knowledge bases are mapped according to certain rules to assist in subsequent document tracing. The generation of the question-and-answer knowledge base includes information extraction, knowledge representation, knowledge association, knowledge classification and annotation, and knowledge storage. Information extraction uses natural language processing (NLP) technology to extract key information from documents, such as entities, keywords, phrases, summaries, etc.; knowledge representation converts the extracted information into a structured format, such as using knowledge graphs, semantic networks, and other methods to represent document content; knowledge association builds connections between knowledge by linking and associating information in different documents to form a knowledge network; knowledge classification and annotation classifies knowledge, annotates different categories and attributes for easy retrieval and application, and finally stores knowledge in a structured way in a question-and-answer knowledge base. The system supports outputting knowledge in an appropriate format to a large model for in-depth analysis and generation of intelligent answers.
[0082] In summary, according to the knowledge base generation method proposed in the present disclosure, knowledge data is acquired to obtain the original knowledge corpus data; the knowledge data is audited according to preset rules, and the audited data in the knowledge data is used as the first knowledge data, thereby ensuring the reliability of the knowledge data; based on the first knowledge data, the second knowledge data is processed through data classification and structured processing, and stored in the target database, thereby clarifying the category of the knowledge data and improving the data storage efficiency; the original knowledge data is updated at the first moment using the second knowledge data, and the original knowledge data is stored in the target database at the second moment, wherein the first moment is the sum of the second moment and the update cycle, and the knowledge base is updated in time; in response to the user's target interaction instruction, the target database is used as the knowledge base of the large model, and the response information of the target interaction instruction is output, thereby providing the user with an accurate and convenient knowledge base usage experience. Through the knowledge base generation method disclosed in the present disclosure, the update frequency and management efficiency of the knowledge base are improved.
[0083] Figure 2 The present invention is a flowchart of an embodiment of the present invention, which is based on first knowledge data, is processed into second knowledge data through data classification and structuring, and is stored in a target database. Figure 2 Yes Figure 1 Further explanation of step 103 is based on Figure 2The embodiment shown comprises the following steps:
[0084] Step 201 : inferring that the category of first knowledge data is a first knowledge category based on a pre-trained knowledge classification model.
[0085] In this embodiment, the knowledge classification model refers to a model that uses a classification algorithm to classify knowledge data. It is a machine learning model that analyzes the characteristics and attributes of knowledge content and classifies it into different categories or labels. The first knowledge category refers to the category of the first knowledge data, which depends on the classification label of the knowledge data when training the knowledge classification model. According to the pre-trained knowledge classification model, the first knowledge data is used as the input of the knowledge classification model to infer the category of the first knowledge data, and the result obtained is the first knowledge category.
[0086] Step 202: According to the first storage form corresponding to the first knowledge category, the first knowledge data is structured into second knowledge data.
[0087] In this embodiment, the first storage form refers to the storage format of the first knowledge data, which is usually a data table of a database. The storage format corresponds to the first knowledge category, and the knowledge data of the same category can be saved as a whole, thereby improving storage efficiency. According to the first storage form determined by the first knowledge category, the first knowledge data is structured into second knowledge data, and the second knowledge data has the first storage form.
[0088] Step 203: Save the second knowledge data to a storage area corresponding to the first storage format in the target database.
[0089] In this embodiment, after the second knowledge data is obtained, the second knowledge data is saved in a storage area corresponding to the first storage format in the target database, thereby completing the storage of the second knowledge data.
[0090] In this embodiment, the first knowledge data is automatically classified and labeled using machine learning technology to facilitate management and retrieval, and structured processing is performed to help improve storage efficiency.
[0091] Figure 3 The present invention is a flowchart of a knowledge update according to an embodiment of the present disclosure. Figure 3 Yes Figure 1 Further description of step 104 is based on Figure 3 The embodiment shown comprises the following steps:
[0092] Step 301: Based on an update cycle, regularly monitor the difference between the second knowledge data and the original knowledge data.
[0093] In this embodiment, according to a preset update cycle, the second knowledge data and the original knowledge data are acquired at a fixed time, the knowledge data are extracted and a similarity comparison is performed to determine the difference or similarity between the second knowledge data and the original knowledge data.
[0094] Step 302, if the difference is less than or equal to the preset difference threshold, the original knowledge data is replaced by the second knowledge data, and the first knowledge category is set as the knowledge category of the original knowledge data; otherwise, the second knowledge data is saved in the storage area corresponding to the first knowledge category in the target database.
[0095] In this embodiment, the preset difference threshold is a numerical value used to reflect the similarity between the second knowledge data and the original knowledge data. The higher the similarity between the two, the smaller the difference, and vice versa. If the difference between the two is less than or equal to the preset difference threshold, it means that the second knowledge data is very close to the original knowledge data, and the second knowledge data has not been modified much relative to the original knowledge data. Then, the original knowledge data is replaced with the second knowledge data. In order to maintain data consistency, the first knowledge category corresponding to the second knowledge data is set to the knowledge category of the original knowledge data. If the difference between the two is greater than the preset difference threshold, it means that the second knowledge data has been greatly modified compared to the original knowledge data, and the second knowledge data is specially saved in the storage area corresponding to the first knowledge category.
[0096] In this embodiment, by updating knowledge data at regular intervals, the knowledge base is provided with a function of knowledge data with high timeliness and adaptability to rapid updates, making the content of the knowledge base more practical.
[0097] Figure 4 The present invention is a flowchart of a knowledge interaction according to an embodiment of the present disclosure. Figure 4 Yes Figure 1 The specific description of step 105 is based on Figure 4 The embodiment shown comprises the following steps:
[0098] Step 401 , in response to a target interaction instruction, using a large model, the semantics of the target interaction instruction is understood as target semantic information.
[0099] In this embodiment, the target semantic information refers to the interactive information that the target interaction instruction actually expresses. The semantic understanding ability of the big model provides the target interaction instruction with instruction understanding and escape functions. For example, the target interaction instruction is: query the customers received last month. It is actually a retrieval instruction, and the target semantics is to retrieve data with a time of last month and a customer as the object from the target data.
[0100] Step 402, based on the knowledge base, generate response information corresponding to the target semantic information through the large model, and the response information is strongly related to the corpus in the knowledge base.
[0101] In this embodiment, based on the target database as the knowledge base of the large model, the large model generates response information corresponding to the target semantic information. The response information is generated by the corpus of the knowledge base, that is, the response information is strongly related to the corpus in the knowledge base.
[0102] In this embodiment, through the question-answering of a large model with the target database as the knowledge base, personalized services are provided to users, and personalized answers are provided to different questions of different users, thereby improving user experience and enterprise operation efficiency.
[0103] Furthermore, the large amount of structured data accumulated by the knowledge base-based large model system during the construction and interaction with users provides rich material for training the large model. This data can be used to optimize the language model and improve its performance in specific fields and the accuracy of intelligent applications.
[0104] Figure 5 The present invention is a flowchart of a knowledge retrieval according to an embodiment of the present invention. Figure 5 Yes Figure 1 Further explanation is based on Figure 5 The embodiment shown comprises the following steps:
[0105] Step 501, in response to a user's search request for target information, generating a target word vector for the target information.
[0106] In this embodiment, the target word vector refers to the word vector obtained by converting the target information through the text-to-word vector tool. When the user searches for the target information, first, the target information is processed into a target word vector to facilitate similarity comparison with the word vector of the knowledge base.
[0107] Step 502, determining the knowledge data corresponding to the word vector having the highest similarity to the target word vector from the word vector library corresponding to the knowledge base.
[0108] In this embodiment, the target vector is compared one by one with the word vectors in the word vector library corresponding to the knowledge base, and the similarity is calculated one by one. The word vector with the highest similarity is used as the word vector corresponding to the retrieval result of the target word vector. The knowledge data corresponding to the word vector is indexed in the knowledge base and can be used as the retrieval result of the target information.
[0109] In this embodiment, an efficient search algorithm is used to quickly respond to user queries, thereby improving user experience.
[0110] Figure 6 The present invention is a flowchart of a knowledge analysis according to an embodiment of the present invention. Figure 6 Yes Figure 1 Further explanation is based on Figure 6 The embodiment shown comprises the following steps:
[0111] Step 601: Count usage data of different users on the knowledge base.
[0112] In this embodiment, in order to analyze users' use of the knowledge base, statistics are collected on the usage data of the knowledge base by different users. For example, the content in the knowledge base that is most visited by users and the content in the knowledge base that users have reported as needing quality improvement can all be counted as data sources for knowledge data usage analysis.
[0113] Step 602, based on the usage data, analyze the usage of the knowledge base through a large model to generate decision recommendations for the enterprise.
[0114] In this embodiment, based on the user's usage data of the knowledge base, the usage of knowledge data is analyzed through a large model to obtain content that users frequently use, content that needs to be provided to users urgently, content that users are interested in, and content that is rarely involved. These content can be summarized and summarized to form decision-making recommendations, providing decision support for decision makers of the enterprise.
[0115] In this embodiment, by analyzing the usage of the knowledge base data, deeper knowledge usage data is mined, which can provide suggestions for the optimization of the knowledge base and further enhance the value of the knowledge base.
[0116] Figure 7 The present invention is a flowchart of the authority management of a knowledge base according to an embodiment of the present invention. Fig. 9 Yes Figure 1 Further explanation is based on Figure 7 The embodiment shown comprises the following steps:
[0117] Step 701, in response to the update instruction of the knowledge base, check whether the user authority is at a preset level, if so, classify and structure the user's knowledge data into second knowledge data, and use the second knowledge data to update the knowledge base.
[0118] In this embodiment, during the manual update of the knowledge base content, it is checked whether the current user's authority reaches the preset level. If so, the knowledge data that the user is going to submit for update is classified, labeled, and structured into the second knowledge data. Since the current user has a brand new knowledge base update, the knowledge base is updated using the second knowledge data. If not, it means that the user does not have the authority to update the knowledge base. In order to ensure that the knowledge base is not tampered with at will, the user is not allowed to update the knowledge base. The data security of the knowledge base is guaranteed.
[0119] Figure 8 FIG. 1 is a schematic diagram of a knowledge storage process of an embodiment of the present disclosure. Figure 8As shown in the figure, the knowledge storage process includes: extracting information from the document knowledge base; representing the extracted knowledge data through word vectors and associating related knowledge together; classifying and annotating knowledge, labeling the knowledge data, and structuring it into a database that is easy to store, access, and retrieve. Based on the knowledge data in the database, a question-and-answer knowledge base is generated to provide knowledge database support for the large-model question-and-answer function based on the knowledge base.
[0120] Fig. 9 FIG. 1 is a schematic diagram of knowledge base management according to an embodiment of the present disclosure. Fig. 9 As shown, the knowledge base generation process is an efficient and automated system designed to build and maintain a comprehensive and accurate knowledge management system.
[0121] Knowledge collection integrates multiple types of documents from multiple data sources such as internal corporate rules and regulations, management requirements, product introductions, system manuals, etc., including various structured and unstructured knowledge content such as word, PDF, PPT, and excel. Knowledge collection supports both automatic and manual methods. The automatic collection method connects to the internal systems of the enterprise, such as OA, and obtains the knowledge files of the internal system according to the acquisition rules. According to the rules, the knowledge files are labeled with the department, category, open scope, upload time, effective time, and uploader for subsequent review and classification. The manual collection method supports authorized users to upload knowledge documents through the system front end and label the documents to enter the document verification stage.
[0122] Knowledge update is to update the collected knowledge documents, and it also supports two methods: automatic collection and manual collection. After the knowledge document is uploaded, the system will use optical character recognition to identify the file content and automatically compare it with the existing documents in the knowledge base. If the similarity is greater than the threshold, it is considered to be an update of the existing document knowledge. The knowledge document inherits the original version label attribute and is accompanied by an update mark to the knowledge verification stage. The knowledge update module ensures that the knowledge base can reflect the latest information and data.
[0123] In the knowledge verification stage, a series of predefined rules are applied to ensure the consistency, accuracy, and sensitive information of the collected knowledge content. If the knowledge verification fails to comply with the regulations, the reasons for the failure of the attachment will be returned to the knowledge collection stage. For manually collected knowledge documents, it will be returned to the uploader; for knowledge automatically collected by the system, it will be returned to the uploader of the affiliated department. If the verification passes, it will enter the knowledge review stage.
[0124] Knowledge review is a manual review of the knowledge that is about to be stored, with a focus on the scope of knowledge openness, knowledge classification, whether sensitive information is involved, etc., to ensure the reliability and authority of the information.
[0125] Knowledge classification uses machine learning technology to automatically classify and label knowledge for easy management and retrieval.
[0126] Knowledge storage includes the storage of document knowledge base and question-answer knowledge base. The two knowledge bases are mapped according to certain rules to assist in subsequent document tracing. The generation of question-answer knowledge base includes information extraction, knowledge representation, knowledge association, knowledge classification and annotation, and knowledge storage. Information extraction uses NLP technology to extract key information from documents, such as entities, keywords, phrases, summaries, etc.; knowledge representation converts the extracted information into a structured format, such as using knowledge graphs, semantic networks and other methods to represent document content; knowledge association links and associates information in different documents to build connections between knowledge and form a knowledge network; knowledge classification and annotation classify knowledge, annotate different categories and attributes for easy retrieval and application, and finally store knowledge in a structured manner in the question-answer knowledge base. The system supports outputting knowledge to the large model in an appropriate format for in-depth analysis and generation of intelligent answers.
[0127] Users can quickly query the content in the knowledge base through the knowledge retrieval module, and knowledge interaction allows users to interact with the knowledge base to provide feedback or supplement content; the knowledge analysis module monitors and analyzes the use of the knowledge base, providing enterprises with valuable insights and decision-making support.
[0128] The knowledge output module outputs knowledge to the big model through an interface for the big model to perform model training, user question and answer calls, etc.
[0129] The entire process forms a closed-loop system that continuously optimizes and enhances the value and utility of the knowledge base.
[0130] Fig.10 FIG. 1 is a schematic diagram of a knowledge base system application architecture according to an embodiment of the present disclosure. Fig.10As shown in the figure, the knowledge base application architecture is a comprehensive framework. It provides interactive experience through various system functions and provides users with various user operation services such as knowledge management and knowledge search. The system settings provide various functions such as organizational structure, personnel, permissions, roles, etc. to achieve refined management and security control, ensure the rational allocation and efficient operation of enterprise resources, and at the same time ensure data security and compliance. Data processing uses NLP and machine learning technologies to enhance content understanding and classification, and supports keyword search, semantic search and advanced query functions through efficient search algorithms. The operating environment of the knowledge base system relies on a high-performance server cluster, providing powerful computing and processing capabilities to ensure system stability and response speed. The system backend uses relational databases such as MySQL, NoSQL, and MongoDB to store and manage structured and unstructured data to ensure data consistency and scalability. In order to improve data access speed and cache efficiency, the system also integrates Redis memory data structure storage and distributed file system. For example, the knowledge base application architecture is a comprehensive framework that provides interactive experience through various system functions and provides users with various user operation services such as knowledge management and knowledge search. The system settings provide various functions such as organization, personnel, authority, and role to achieve refined management and security control, ensure the rational allocation and efficient operation of enterprise resources, and ensure data security and compliance. Data processing uses NLP and machine learning technology to enhance content understanding and classification, and supports keyword search, semantic search and advanced query functions through efficient search algorithms. The operating environment of the knowledge base system relies on a high-performance server cluster, providing powerful computing and processing capabilities to ensure the stability and response speed of the system. The system backend uses relational databases such as MySQL and NoSQL databases such as MongoDB to store and manage structured and unstructured data to ensure data consistency and scalability. In order to improve data access speed and cache efficiency, the system also integrates Redis memory data structure storage and distributed file systems, such as Hadoop Distributed File System (HDFS), to optimize the access and backup of large-scale data.
[0131] Corresponding to the methods provided in the above-mentioned embodiments, the present disclosure also provides a knowledge base generating device. Since the device provided in the embodiments of the present disclosure corresponds to the methods provided in the above-mentioned embodiments, the implementation method of the method is also applicable to the device provided in this embodiment and will not be described in detail in this embodiment.
[0132] Fig.11 FIG. 1 is a schematic diagram of the structure of a knowledge base generating device 1100 according to an embodiment of the present disclosure. Fig.11 As shown, the knowledge base generating device includes:
[0133] An acquisition module 1110, used for acquiring knowledge data;
[0134] The review module 1120 is used to review the knowledge data according to preset rules, and use the data in the knowledge data that pass the review as the first knowledge data;
[0135] The storage module 1130 is used to classify and structure the first knowledge data into second knowledge data, and store the second knowledge data in a target database;
[0136] An updating module 1140, configured to update the original knowledge data using the second knowledge data at a first moment, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and the updating period;
[0137] The interaction module 1150 is used to respond to the user's target interaction instruction, use the target database as the knowledge base of the large model, and output response information of the target interaction instruction.
[0138] In some embodiments, the storage module 1130 is used to:
[0139] Inferring, according to the pre-trained knowledge classification model, the category of the first knowledge data is the first knowledge category;
[0140] According to the first storage form corresponding to the first knowledge category, the first knowledge data is structured into second knowledge data;
[0141] The second knowledge data is saved in a storage area corresponding to the first storage format in the target database.
[0142] In some embodiments, the update module 1140 is used to:
[0143] Based on the update cycle, regularly monitor the difference between the second knowledge data and the original knowledge data;
[0144] If the difference is less than or equal to the preset difference threshold, the second knowledge data is used to replace the original knowledge data, and the first knowledge category is set as the knowledge category of the original knowledge data; otherwise, the second knowledge data is saved in the storage area corresponding to the first knowledge category in the target database.
[0145] In some embodiments, the interaction module 1150 is used to:
[0146] In response to the target interaction instruction, using the large model, the semantics of the target interaction instruction is understood as target semantic information;
[0147] Based on the knowledge base, a large model is used to generate response information corresponding to the target semantic information, and the response information is strongly related to the corpus in the knowledge base.
[0148] In some embodiments, the apparatus 1100 is further configured to:
[0149] In response to a user's search request for target information, generating a target word vector for the target information;
[0150] From the word vector library corresponding to the knowledge base, determine the knowledge data corresponding to the word vector with the highest similarity to the target word vector.
[0151] In some embodiments, the apparatus 1100 is further configured to:
[0152] Collect statistics on the usage of knowledge base by different users;
[0153] Based on usage data, the usage of the knowledge base is analyzed through a large model to generate decision-making recommendations for the enterprise.
[0154] In some embodiments, the apparatus 1100 is further configured to:
[0155] In response to the update instruction of the knowledge base, it is checked whether the user authority is at a preset level. If so, the user's knowledge data is classified and structured into second knowledge data, and the knowledge base is updated using the second knowledge data.
[0156] In summary, knowledge data is acquired through a knowledge base generation device; the knowledge data is reviewed according to preset rules, and the data that passes the review in the knowledge data is used as the first knowledge data; based on the first knowledge data, the second knowledge data is processed through data classification and structuring, and stored in the target database; the original knowledge data is updated using the second knowledge data at a first moment, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and the update cycle; in response to the user's target interaction instruction, the target database is used as the knowledge base of the large model, and the response information of the target interaction instruction is output. This device solves the management difficulties caused by the untimely update of the knowledge base, the wide variety of content, and the complex sources of knowledge in the related technology, and improves the update frequency and management efficiency of the knowledge base.
[0157] In the above embodiments provided by the present disclosure, the methods and devices provided by the embodiments of the present disclosure are introduced. In order to implement the functions in the methods provided by the above embodiments of the present disclosure, the electronic device may include a hardware structure and a software module, and implement the above functions in the form of a hardware structure, a software module, or a hardware structure plus a software module. One of the above functions may be executed in the form of a hardware structure, a software module, or a hardware structure plus a software module.
[0158] Fig.12 is a block diagram of an electronic device 1200 for implementing the above-mentioned knowledge base generating method according to an exemplary embodiment.
[0159] For example, the electronic device 1200 may be a mobile phone, a computer, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.
[0160] Reference Fig.12 , the electronic device 1200 may include one or more of the following components: a processing component 1202 , a memory 1204 , a power component 1206 , a multimedia component 1208 , an audio component 1210 , an input / output (I / O) interface 1212 , a sensor component 1214 , and a communication component 1216 .
[0161] The processing component 1202 generally controls the overall operation of the electronic device 1200, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 1202 may include one or more processors 1220 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 1202 may include one or more modules to facilitate the interaction between the processing component 1202 and other components. For example, the processing component 1202 may include a multimedia module to facilitate the interaction between the multimedia component 1208 and the processing component 1202.
[0162] The memory 1204 is configured to store various types of data to support operations on the electronic device 600. Examples of such data include instructions for any application or method operating on the electronic device 1200, contact data, phone book data, messages, pictures, videos, etc. The memory 1204 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.
[0163] The power supply component 1206 provides power to the various components of the electronic device 1200. The power supply component 1206 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device 1200.
[0164] The multimedia component 1208 includes a screen that provides an output interface between the electronic device 1200 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1208 includes a front camera and / or a rear camera. When the electronic device 1200 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
[0165] The audio component 1210 is configured to output and / or input audio signals. For example, the audio component 1210 includes a microphone (MIC), and when the electronic device 1200 is in an operation mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 1204 or sent via the communication component 1216. In some embodiments, the audio component 1210 also includes a speaker for outputting audio signals.
[0166] I / O interface 1212 provides an interface between processing component 1202 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.
[0167] The sensor assembly 1214 includes one or more sensors for providing various aspects of status assessment for the electronic device 1200. For example, the sensor assembly 1214 can detect the open / closed state of the electronic device 1200, the relative positioning of components, such as the display and keypad of the electronic device 1200, and the sensor assembly 1214 can also detect the position change of the electronic device 1200 or a component of the electronic device 1200, the presence or absence of user contact with the electronic device 1200, the orientation or acceleration / deceleration of the electronic device 1200, and the temperature change of the electronic device 1200. The sensor assembly 1214 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 1214 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1214 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
[0168] The communication component 1216 is configured to facilitate wired or wireless communication between the electronic device 1200 and other devices. The electronic device 1200 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, 4G LTE, 5G NR (NewRadio) or a combination thereof. In an exemplary embodiment, the communication component 1216 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 1216 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
[0169] In an exemplary embodiment, the electronic device 1200 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.
[0170] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 1204 including instructions, and the instructions can be executed by the processor 1220 of the electronic device 1200 to complete the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.
[0171] The embodiments of the present disclosure further provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the knowledge base generation method described in the above embodiments of the present disclosure.
[0172] An embodiment of the present disclosure further provides a computer program product, including a computer program, which executes the knowledge base generation method described in the above embodiment of the present disclosure when a processor is used to execute the computer program.
[0173] Fig.13 is a schematic structural diagram of a chip 1300 for implementing the above-mentioned knowledge base generation method according to an exemplary embodiment.
[0174] Reference Fig.13The chip 1300 includes at least one communication interface 1301 and a processor 1302; the communication interface 1301 is used to receive a signal input to the chip 1300 or a signal output from the chip 1300, and the processor 1302 communicates with the communication interface 1301 and implements the knowledge base generation method described in the above embodiment through a logic circuit or executing code instructions.
[0175] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0176] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0177] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present disclosure includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present disclosure belong.
[0178] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processing module, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (control method), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing in a suitable manner if necessary, and then stored in a computer memory.
[0179] It should be understood that the various parts of the embodiments of the present disclosure can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0180] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.
[0181] In addition, each functional unit in each embodiment of the present disclosure may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium. The above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.
[0182] Although the embodiments of the present disclosure have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present disclosure. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present disclosure.
Claims
1. A knowledge base generation method, characterized in that: The method comprises: Acquiring knowledge data; Reviewing the knowledge data according to preset rules, and taking the reviewed data among the knowledge data as first knowledge data; Based on the first knowledge data, the second knowledge data is obtained through data classification and structuring, and stored in a target database; At a first moment, the original knowledge data is updated using the second knowledge data, and the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and an update period; In response to the user's target interaction instruction, the target database is used as the knowledge base of the large model to output response information of the target interaction instruction.
2. The method according to claim 1, characterized in that The first knowledge data is processed into second knowledge data through data classification and structuring, and stored in a target database, including: Inferring, according to a pre-trained knowledge classification model, that the category of the first knowledge data is a first knowledge category; According to a first storage form corresponding to a first knowledge category, the first knowledge data is structured into the second knowledge data; The second knowledge data is saved in a storage area corresponding to the first storage format in the target database.
3. The method according to claim 1, characterized in that The updating of the original knowledge data by using the second knowledge data at the first moment includes: Based on the update cycle, regularly monitoring the difference between the second knowledge data and the original knowledge data; If the difference is less than or equal to a preset difference threshold, the second knowledge data is used to replace the original knowledge data, and the first knowledge category is set as the knowledge category of the original knowledge data; otherwise, the second knowledge data is saved in the storage area corresponding to the first knowledge category in the target database.
4. The method according to claim 1, characterized in that: The method of responding to the target interaction instruction of the user, using the target database as the knowledge base of the large model, and outputting the response information of the target interaction instruction includes: In response to the target interaction instruction, using the large model, the semantics of the target interaction instruction is understood as target semantic information; Based on the knowledge base, the response information corresponding to the target semantic information is generated by the large model, and the response information is strongly related to the corpus in the knowledge base.
5. The method according to claim 1, characterized in that The method further comprises: In response to a user's search request for target information, generating a target word vector for the target information; From the word vector library corresponding to the knowledge base, determine the knowledge data corresponding to the word vector with the highest similarity to the target word vector.
6. The method according to claim 1, characterized in that The method further comprises: Collecting statistics on usage data of the knowledge base by different users; Based on the usage data, the usage of the knowledge base is analyzed through the big model to generate decision recommendations for the enterprise.
7. The method according to claim 1, characterized in that The method further comprises: In response to the update instruction of the knowledge base, it is checked whether the user authority is at a preset level. If so, the user's knowledge data is classified and structured into the second knowledge data, and the knowledge base is updated using the second knowledge data.
8. A knowledge base generation device, characterized in that: The device comprises: An acquisition module, used to acquire knowledge data; An audit module, used to audit the knowledge data according to preset rules, and use the data in the knowledge data that pass the audit as the first knowledge data; A storage module, used for converting the first knowledge data into second knowledge data through data classification and structuring, and storing the second knowledge data in a target database; An updating module, configured to update original knowledge data using the second knowledge data at a first moment, wherein the original knowledge data is stored in the target database at a second moment, wherein the first moment is the sum of the second moment and an update period; The interaction module is used to respond to the user's target interaction instruction, use the target database as the knowledge base of the large model, and output the response information of the target interaction instruction.
9. An electronic device, characterized in that: include: at least one processor; as well as 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 to enable the at least one processor to perform the method according to any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.
Citation Information
Cited By
Knowledge base online updating method and device
CN119740650A
Incremental knowledge operation management method for ship design knowledge big data
CN120849613A
Incremental knowledge operation and management methods for big data in ship design knowledge
CN120849613B
Model-based agent creation method and device, equipment, medium and product
CN120851071A
Knowledge base file processing method and system, electronic device, medium, and product
WO2026118676A1
