Knowledge graph construction method and device based on data classification and grading

Abstract

The embodiment of the invention provides a knowledge graph construction method and device based on data classification and grading, and the method comprises the steps: obtaining historical enterprise data, and determining corresponding enterprise classification data according to a preset industry field classification rule and a business attribute rule in a preset industry field; grading label adding operation is carried out on the enterprise classification data, corresponding classification and grading characteristics are determined, the classification and grading characteristics are input into a preset initial model for model training, and a corresponding classification and grading model is determined; performing agent construction operation according to the classification and grading model and a set classification and grading engine, and determining a corresponding classification and grading agent; and performing classification and grading operation on the real-time enterprise data according to the classification and grading intelligent agent, determining corresponding enterprise classification and grading data, and constructing to obtain a corresponding classification and grading knowledge graph, so that the data classification and grading efficiency and accuracy can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing, and specifically to a method and device for constructing a knowledge graph based on data classification and grading. Background Art

[0002] With the rapid development of information technology, the amount of data faced by enterprises is growing exponentially, and the types of data are becoming increasingly complex and diverse. These data are not only the core assets of enterprises, but also the key elements to support enterprise decision-making, optimize operations, and enhance competitiveness. Therefore, how to efficiently and accurately classify, grade and protect sensitive information in these enterprise data has become a major challenge facing enterprises.

[0003] In recent years, with the rise of technologies such as artificial intelligence, big data, and machine learning, new opportunities have emerged in data classification and management. Among them, knowledge graphs, as a new method of representing and organizing knowledge, have gradually become an important tool for data management and utilization with their powerful knowledge representation and reasoning capabilities. However, data classification based on knowledge graphs in the market still faces many challenges. The current mainstream way of building knowledge graphs in the industry is based on internal enterprise data and public data, and graph service providers help customers customize and build knowledge graphs in the form of solutions. This method is undoubtedly very costly and inefficient, and usually takes a long time to complete.

[0004] Therefore, there is an urgent need for a knowledge graph construction method based on data classification and grading, which can efficiently and automatically classify and grade data, build a knowledge graph based on the classified and graded data, and launch a data classification knowledge base based on the knowledge graph to achieve efficient and accurate classification, grading and protection of enterprise data. Summary of the invention

[0005] In response to the problems in the prior art, the present application provides a method and device for constructing a knowledge graph based on data classification and grading, which can improve the efficiency and accuracy of data classification and grading.

[0006] In order to solve at least one of the above problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a method for constructing a knowledge graph based on data classification and grading, comprising:

[0008] Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data;

[0009] Performing feature extraction on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining a corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and a set classification and grading engine, and determining a corresponding classification and grading intelligent agent;

[0010] Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0011] Furthermore, classifying the historical enterprise data according to the preset industry field classification rules and the business attribute rules within the preset industry field to determine the corresponding enterprise classification data includes:

[0012] Performing data classification operations on the historical enterprise data according to preset industry field classification rules to determine corresponding industry field classification data, wherein the industry field classification data is obtained by classifying data according to industry fields;

[0013] The industry field classification data is refined and classified according to the business attribute rules in the preset industry field to determine the corresponding enterprise classification data, wherein the enterprise classification data is obtained by further classifying according to the business attributes in the industry field to which it belongs on the basis of the industry field classification.

[0014] Furthermore, the step of performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, and determining corresponding classification and classification features includes:

[0015] Performing a feature importance evaluation operation on the enterprise classification data according to a preset feature evaluation rule to determine corresponding important classification features, and performing a feature extraction operation on the important classification features to determine corresponding classification features;

[0016] A classification label adding operation is performed on the classification feature to determine the corresponding classification and classification feature.

[0017] Furthermore, the step of inputting the classification and grading features into a preset initial model for model training to determine a corresponding classification and grading model includes:

[0018] Input the classification and grading features into a preset long short-term memory model for model training, and determine the corresponding loss function;

[0019] The long short-term memory model is subjected to parameter tuning operations according to the minimum loss function and the back propagation algorithm to determine the corresponding classification and grading model.

[0020] Furthermore, before performing the agent construction operation according to the classification and grading model and setting the classification and grading engine to determine the corresponding classification and grading agent, it includes:

[0021] Performing a multi-dimensional index extraction operation on the historical enterprise data to determine corresponding multi-dimensional classification and grading indicators;

[0022] According to the multi-dimensional classification and grading indicators, preset indicator weights and preset judgment rules, a corresponding classification and grading engine is determined.

[0023] Furthermore, the classifying and grading operation is performed on the real-time enterprise data according to the classification and grading agent to determine the corresponding enterprise classification and grading data, including:

[0024] Classify and grade the real-time enterprise data according to the classification and grading model in the classification and grading agent to determine the corresponding first enterprise classification and grading data;

[0025] Performing classification and grading operations on the real-time enterprise data according to a set classification and grading engine in the classification and grading intelligent agent to determine corresponding second enterprise classification and grading data;

[0026] Corresponding enterprise classification and grading data are determined according to the first enterprise classification and grading data, the second enterprise classification and grading data, and a preset fusion rule.

[0027] Furthermore, the determining corresponding enterprise classification and grading data according to the first enterprise classification and grading data, the second enterprise classification and grading data and a preset fusion rule includes:

[0028] Determining whether the second enterprise classification and grading data is consistent with the first enterprise classification and grading data according to a priority fusion rule;

[0029] If they are inconsistent, the second enterprise classification and grading data is determined as the enterprise classification and grading data.

[0030] In a second aspect, the present application provides a knowledge graph construction device based on data classification and grading, comprising:

[0031] An enterprise classification data determination module is used to obtain historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine the corresponding enterprise classification data;

[0032] A classification and grading agent determination module is used to perform feature extraction operations on the enterprise classification data, perform grading label addition operations on the classification features after the feature extraction, determine the corresponding classification and grading features, input the classification and grading features into a preset initial model for model training, and determine the corresponding classification and grading model; perform agent construction operations according to the classification and grading model and the set classification and grading engine, and determine the corresponding classification and grading agent;

[0033] A classification and grading knowledge graph construction module is used to obtain real-time enterprise data, classify and grade the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0034] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for constructing a knowledge graph based on data classification and grading are implemented.

[0035] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for constructing a knowledge graph based on data classification and grading.

[0036] In a fifth aspect, the present application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the method for constructing a knowledge graph based on data classification and grading.

[0037] It can be seen from the above technical scheme that the present application provides a method and device for constructing a knowledge graph based on data classification and grading, which obtains historical enterprise data and determines the corresponding enterprise classification data according to preset industry classification rules and business attribute rules within the preset industry field; performs a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, inputs the classification and grading features into a preset initial model for model training, and determines the corresponding classification and grading model; performs an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; performs classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and constructs a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 This is one of the flow charts of the method for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0040] Figure 2 This is the second flow chart of the method for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0041] Figure 3 The third flowchart of the method for constructing a knowledge graph based on data classification and grading in the embodiment of the present application;

[0042] Figure 4 This is a fourth flow chart of a method for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0043] Figure 5 This is a fifth flow chart of a method for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0044] Figure 6 This is a sixth flow chart of the method for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0045] Figure 7 This is the seventh flow chart of the method for constructing a knowledge graph based on data classification and grading in the embodiment of the present application;

[0046] Figure 8 A structural diagram of a device for constructing a knowledge graph based on data classification and grading in an embodiment of the present application;

[0047] Fig. 9 It is a schematic diagram of the structure of an electronic device in an embodiment of the present application.

[0048] Reference numerals:

[0049] Electronic device 9600, central processing unit 9100, memory 9140, communication module 9110, input unit 9120, audio processor 9130, display 9160, power supply 9170, buffer memory 9141, application / function storage unit 9142, data storage unit 9143, driver program storage unit 9144, antenna 9111, speaker 9131, microphone 9132. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] The acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0052] Considering the low efficiency of existing knowledge graph construction and the inability to perform hierarchical protection on enterprise data, it is unable to meet the requirements of efficient and accurate classification and grading of enterprise data. The present application provides a method and device for constructing a knowledge graph based on data classification and grading, which obtains historical enterprise data, determines the corresponding enterprise classification data according to preset industry field classification rules and business attribute rules within preset industry fields; performs a classification label adding operation on the enterprise classification data to determine the corresponding classification and grading features, inputs the classification and grading features into a preset initial model for model training, and determines the corresponding classification and grading model; performs an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; performs classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and constructs a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0053] In order to improve the efficiency and accuracy of data classification and grading, this application provides an embodiment of a method for constructing a knowledge graph based on data classification and grading, see Figure 1 The method for constructing a knowledge graph based on data classification and grading specifically includes the following contents:

[0054] Step S101: Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data;

[0055] Optionally, in this embodiment, historical enterprise data refers mainly to data that can be assetized, including customer transaction records, market research data, production process data, employee performance data, etc. They exist in the form of text, numbers, pictures, videos, etc., and record important information such as the company's operating conditions, market trends, and customer needs.

[0056] Optionally, in this embodiment, when acquiring historical enterprise data, it is first necessary to analyze and determine the needs of various business areas within the enterprise, including data sources, data types, and related compliance requirements.

[0057] Specific, demand research: interview business departments to understand specific needs.

[0058] Business scenario definition: Clarify the purpose and usage of data classification.

[0059] Compliance Assessment: Evaluate compliance requirements within each business area.

[0060] Optionally, in this embodiment, industry field classification rules are preset, and data is divided into industrial data, telecommunications data, financial data, energy data, transportation data, natural resource data, health data, education data, scientific data, etc. according to the industry field.

[0061] Optionally, in this embodiment, business attribute rules within the industry field are preset, and the data in the industry field is classified in detail according to the business attributes of the industry field.

[0062] Common business attributes include but are not limited to:

[0063] 1) Business field: classified according to business scope, business type or business function;

[0064] 2) Responsible department: classified according to data management department or division of responsibilities;

[0065] 3) Description object: Classify data according to the object it describes;

[0066] Note 1: According to the description object, it is divided into user data, business data, business management data, and system operation and maintenance data.

[0067] 4) Process links: detailed classification according to business processes and industrial chain links;

[0068] Note 2: Energy data is divided into exploration, mining, production, processing, sales, and usage data according to the process links.

[0069] 5) Data subject: Detailed classification according to the data subject or owner;

[0070] Note 3: According to the data subject, it is divided into public data, organizational data and personal information.

[0071] 6) Content theme: Detailed classification according to the content theme described by the data;

[0072] 7) Data usage: Detailed classification according to data processing purpose and usage;

[0073] 8) Data processing: classified according to data processing activities or degree of data processing;

[0074] 9) Data source: Detailed classification according to data source and collection method.

[0075] Specifically, the data classification method is:

[0076] Data classification can be based on data management and usage requirements, combined with the existing data classification basis, and flexibly select business attributes to refine data classification. Please refer to the following steps to carry out industry data classification.

[0077] a) Clarify the scope of data: According to the responsibilities of the competent (regulatory) departments in the industry field, clarify the scope of data managed in this industry and field.

[0078] b) Refine business classification: Refine the business of this industry and field, including:

[0079] 1) Clarify the classification of industry fields or business lines in combination with the division of departmental responsibilities;

[0080] Note 1: Industrial sector data are divided into categories such as raw materials, equipment manufacturing, consumer goods, electronic information manufacturing, software and information technology services according to departmental responsibilities.

[0081] 2) According to business scope, operating model, business process, etc., refine the industry fields or clarify the key business classification of each business line.

[0082] c) Business attribute classification: Select appropriate business attributes and classify key business data in detail.

[0083] d) Determine classification rules: Sort out and analyze the data classification results of each key business, and determine the classification rules according to the data management and usage requirements of the industry.

[0084] Determine industry data classification rules, such as:

[0085] 1) Data classification rules can be given in the form of "business line - key business - business attribute classification";

[0086] 2) The data classification results of key businesses can also be classified and analyzed, and data subcategories with similar themes can be classified.

[0087] It is understandable that after the above data analysis, enterprise data can be accurately classified, laying a solid data foundation for the subsequent construction of knowledge graphs for data retrieval and management.

[0088] Step S102: performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining the corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining the corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and the set classification and grading engine, and determining the corresponding classification and grading intelligent agent;

[0089] Optionally, in this embodiment, the purpose of this step is to build a classification and grading intelligent entity to implement enterprise data classification and grading.

[0090] Optionally, in this embodiment, the technical means adopted in this step is to construct a classification and grading model and a classification and grading engine respectively, and to construct a classification and grading intelligent agent by integrating the two.

[0091] Optionally, in this embodiment, data classification and grading refers to dividing data into different categories and levels based on factors such as data sensitivity, value, legal requirements, etc., so as to adopt corresponding security measures and management strategies.

[0092] Optionally, the steps to build a classification and grading model are as follows:

[0093] First, data preprocessing is performed, which requires cleaning and preprocessing the raw data, including removing invalid data, filling missing values, converting data formats, etc. Then, the data is scaled to a smaller specific range (such as 0 to 1) or converted to a distribution with zero mean and unit variance.

[0094] Secondly, the feature importance of the preprocessed classification data is evaluated to obtain the importance level of the features, and the features with higher importance are extracted to obtain the extracted classification features.

[0095] Specifically, the feature evaluation rule uses statistical tests to select the features most relevant to the target variable. Preferably, for complex feature variables, a machine learning model is used to evaluate the importance of features, such as random forests, which iteratively improves the feature set by building a model and repeatedly selecting the most influential features.

[0096] More specifically, when using the model for importance assessment, feature engineering is performed based on the type of data:

[0097] For text data, the bag-of-words model is used to convert text into vectors of word occurrences, by evaluating the importance of a word for one of the documents in a text collection.

[0098] For numerical data, use binning to split continuous numerical data into discrete intervals and generate polynomial combinations of the original features.

[0099] For categorical data, one-hot encoding is used to convert categorical variables into binary (0 or 1) form. Label encoding maps categories to numerical values.

[0100] After the above feature evaluation, principal component analysis was performed on the features to transform the potentially correlated variables into a set of linearly uncorrelated variables with a smaller number of values ​​through orthogonal transformation.

[0101] It is understandable that after the above-mentioned importance evaluation of features with different characteristics, useful and important data can be accurately extracted in a targeted manner, improving the accuracy of subsequent model training. The subsequent main component analysis of the features is the process of feature dimensionality reduction. The main purpose is to visualize the data for easy observation and exploration. It also helps to simplify the training and prediction of subsequent machine learning models.

[0102] Then, feature extraction is performed on the important features obtained through the above evaluation, and feature extraction is performed on the classified data to extract key features that are helpful for subsequent processing. For example, for text data, features such as keywords and sentence structure can be extracted. Feature extraction involves different extraction methods for different data:

[0103] For text data, use pre-trained models such as the BERT model to extract embedded representations of sentences or paragraphs.

[0104] For image data, the Juanju neural network is used to extract local image features.

[0105] For audio data, Mel-frequency cepstral coefficients are used to extract audio signal features.

[0106] The different features extracted above are recombined into new features to discover the potential relationship between them and obtain the extracted classification features.

[0107] Then, hierarchical labels are added to the extracted classification features. According to the risk performance of historical data, labels (such as "low", "medium", "high") are added to each feature or feature combination to obtain classification and hierarchical features with hierarchical labels.

[0108] It can be understood that after the above steps, we obtain the classification and grading features for training the model, so that the model can learn the relationship between data categories and levels through the classification and grading features, and automatically classify and grade the data.

[0109] Next, the obtained classification and grading features are input into the long short-term memory model for model training. Through iterative training, the model parameters are optimized to improve the accuracy of classification and grading.

[0110] Specifically, the classification and grading features are divided into a training set, a validation set and a test set. The training set is used to train the model to obtain the predicted features of the model output, the predicted features are compared with the real features in the validation set, and the corresponding loss function is calculated. Through iterative training, the minimum loss function in a series of loss functions is used as the final training loss, which is used to tune and update the parameters of each functional layer of the model through back propagation according to the minimum training loss, so as to obtain an accurate model for classification and grading.

[0111] It can be understood that the above steps train an accurate and efficient classification and grading model, laying the foundation for the subsequent construction of a classification and grading intelligent agent.

[0112] Optionally, the steps to build a classification and grading engine are as follows:

[0113] First of all, in the rule engine, specific rules are formulated according to business needs, and these rules can be applied to various business scenarios to achieve automated decision-making and data processing. Therefore, the first task is the definition of rules in the rule engine. In this embodiment, business rules are defined using a rule language (such as a rule language dedicated to rule engines such as Drools and Jess) or a custom rule format. After the rule engine receives the data to be processed, it matches and executes according to the preset rules. It is worth noting that the definition of specific rules is defined according to the business scenario.

[0114] Secondly, in the process of data classification, multi-dimensional indicator judgment is introduced to identify and analyze data features from multiple dimensions.

[0115] Specifically, multi-dimensional indicator extraction and multi-dimensional indicator determination is a complex decision-making system that integrates multiple dimensions of determination indicators, such as data attributes, business logic, legal and regulatory requirements, etc., and performs multi-dimensional and multi-level verification and evaluation on data classification and grading results. Through this system, data classification results can be judged more comprehensively and accurately, improving the quality and efficiency of data management.

[0116] For multi-dimensional indicator determination, first conduct demand analysis and indicator definition to clarify the purpose, application scenarios and compliance requirements of data classification. Based on the results of demand analysis, define a series of multi-dimensional indicators for data classification determination and multi-dimensional indicators for data grading determination. These indicators can include basic attributes of data (such as data type, data source), business logic indicators (such as business processes, business scenarios), legal and regulatory indicators (such as data protection regulations, personal information processing specifications), confidentiality levels (low, medium, high), etc.

[0117] After extracting the above-mentioned multi-dimensional classification and grading indicators, assign corresponding weights to each indicator according to its importance and impact.

[0118] Next, design the judgment logic of multi-dimensional indicators, clarify the correlation, judgment order and judgment conditions between the indicators. Through logical design, ensure that the multi-dimensional indicators can work together and act together in the judgment process of data classification results.

[0119] Finally, the designed decision logic is encoded into a computer-executable rule set. The rule set should contain information such as the complete decision process, decision conditions, and decision result output. The classification and grading engine is constructed through the extracted multi-dimensional classification and grading indicators, the importance weights assigned to the indicators, and the decision logic.

[0120] It can be understood that the above steps build a classification and grading engine to classify and grade the received data, and the rule engine can ensure the accuracy of the classification and grading.

[0121] Optionally, build a classification and grading agent. In order to improve the accuracy and efficiency of data classification and grading, the prediction results of the large model engine and the decision results of the rule engine are usually integrated. Integrate the preset classification and grading engine with the model to ensure that the engine can call the model for classification and grading operations and perform logical processing according to business needs.

[0122] Among them, the big model engine is a prediction model based on deep learning, which can automatically extract features from a large amount of data and realize automatic classification and grading of data. The advantage of the big model engine lies in its powerful learning and generalization capabilities, and its ability to handle complex data patterns.

[0123] Among them, the rule engine is a decision support system based on business logic, which can classify and grade data according to preset rules. The advantage of the rule engine lies in its interpretability and flexibility, and its ability to quickly respond to changes in business needs.

[0124] It is understandable that the integration of rule engines and big model engines can combine the advantages of both to achieve more efficient and accurate data classification and grading. Specifically, the rule engine can be used to process simple and clear business rules, while the big model engine can be used to process complex and ambiguous data patterns. At the same time, on the other hand, the rule engine increases the accuracy of the results, while the big model engine increases the versatility of the results.

[0125] Step S103: Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0126] Optional, this step is to build a classification and grading intelligent agent, perform classification and grading according to the intelligent agent, obtain real-time enterprise data, obtain corresponding classification and grading data through the classification and grading model and classification and grading engine respectively, and intelligently integrate the results of the large model engine and the results based on the rule engine to complement each other and further improve the initial accuracy. Among them, the priority of the rule-based engine is higher than the large model engine.

[0127] That is to say, when the output of the rule engine is inconsistent with the big model engine, the output of the rule engine shall prevail, because the output of the rule engine is more accurate. Subsequently, the big model engine can be re-adjusted and trained based on the output of the rule engine, and the big model engine can be iteratively tuned to increase the accuracy of the big model engine.

[0128] When there is data that is not included in the rule engine, the prediction of the large model engine is output. After manual review, it can be written into the rule engine by defining rules. In this way, not only the accuracy of data classification and grading is increased, but also the versatility of data classification and grading is increased, realizing customized classification not only for one enterprise and one business, but also providing a broad database.

[0129] For example, taking the production process data of a manufacturing enterprise as an example, the specific implementation steps are as follows:

[0130] Data acquisition: Obtain production process data in real time from the production system, including data on each link such as raw material warehousing, production and processing, and finished product delivery.

[0131] Large model engine classification and grading: Use the trained classification and grading model to automatically classify and grade the pre-processed production process data, and preliminarily identify the data types and data levels of different production links.

[0132] Classification and grading of rule engines: Based on the production process data classification and grading rules formulated by the enterprise (such as data type, business logic, data sensitivity, etc.), the results of the large model engine are verified and adjusted in multiple dimensions. If the final results are inconsistent, the rule engine will take priority.

[0133] For example, data involving sensitive information (such as customer information and financial data) needs to be managed and protected at a higher security level regardless of how the big model classifies it.

[0134] Result fusion and output: Intelligently integrate the results of the large model engine and the adjustment results based on the multi-dimensional indicator judgment engine to ensure that the final classification and grading results are consistent with both business logic and compliance requirements.

[0135] Through the above embodiments, the company has successfully achieved accurate classification and grading management of production process data, improving the efficiency and security of data management.

[0136] Optionally, after obtaining the classified and graded data, this embodiment constructs a knowledge graph through the classified and graded data, which is used to launch a data classification knowledge base based on the knowledge graph through abstract knowledge graph construction process and related technologies, and provide a platform for quickly building a digital asset library for different industries and enterprises, enabling large, medium and small enterprises to build their own data asset pools.

[0137] Specifically, the construction of the knowledge graph first performs entity recognition and relationship extraction on the classified and graded data, identifies entities in the text, such as names, organizations, places, etc., and links the identified entities to corresponding entries in the known knowledge base. Entity pairs that may contain relationships are extracted from the text, and each entity pair is classified to determine whether it has a certain type of relationship.

[0138] Then, the knowledge graph is constructed through the identified entities and the relationships between entities, and a graph structure used to represent the classified and hierarchical data and the relationships between the data is obtained, namely the classified and hierarchical knowledge graph.

[0139] This example demonstrates how this embodiment can achieve automatic classification and grading of data, and construct a knowledge graph based on the classified and graded data for use in building a data asset pool.

[0140] From the above description, it can be seen that the knowledge graph construction method based on data classification and grading provided in the embodiment of the present application can obtain historical enterprise data, determine the corresponding enterprise classification data according to preset industry field classification rules and business attribute rules within the preset industry field; perform a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, input the classification and grading features into the preset initial model for model training, and determine the corresponding classification and grading model; perform an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; perform classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and construct a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0141] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 2 , and can also include the following:

[0142] Step S201: performing data classification operation on the historical enterprise data according to preset industry field classification rules to determine corresponding industry field classification data, wherein the industry field classification data is obtained by classifying data according to industry fields;

[0143] Step S202: The industry field classification data is classified according to the business attribute rules in the preset industry field to determine the corresponding enterprise classification data, wherein the enterprise classification data is obtained by further classifying the business attributes in the industry field to which it belongs based on the industry field classification.

[0144] Optionally, in this embodiment, industry field classification rules are preset, and data is divided into industrial data, telecommunications data, financial data, energy data, transportation data, natural resource data, health data, education data, scientific data, etc. according to the industry field.

[0145] Optionally, in this embodiment, business attribute rules within the industry field are preset, and the data in the industry field is classified in detail according to the business attributes of the industry field.

[0146] Common business attributes include but are not limited to:

[0147] 1) Business field: classified according to business scope, business type or business function;

[0148] 2) Responsible department: classified according to data management department or division of responsibilities;

[0149] 3) Description object: Classify data according to the object it describes;

[0150] Note 1: According to the description object, it is divided into user data, business data, business management data, and system operation and maintenance data.

[0151] 4) Process links: detailed classification according to business processes and industrial chain links;

[0152] Note 2: Energy data is divided into exploration, mining, production, processing, sales, and usage data according to the process links.

[0153] 5) Data subject: Detailed classification according to the data subject or owner;

[0154] Note 3: According to the data subject, it is divided into public data, organizational data and personal information.

[0155] 6) Content theme: Detailed classification according to the content theme described by the data;

[0156] 7) Data usage: Detailed classification according to data processing purpose and usage;

[0157] 8) Data processing: classified according to data processing activities or degree of data processing;

[0158] 9) Data source: Detailed classification according to data source and collection method.

[0159] Specifically, the data classification method is:

[0160] Data classification can be based on data management and usage requirements, combined with the existing data classification basis, and flexibly select business attributes to refine data classification. Please refer to the following steps to carry out industry data classification.

[0161] a) Clarify the scope of data: According to the responsibilities of the competent (regulatory) departments in the industry field, clarify the scope of data managed in this industry and field.

[0162] b) Refine business classification: Refine the business of this industry and field, including:

[0163] 1) Clarify the classification of industry fields or business lines in combination with the division of departmental responsibilities;

[0164] Note 1: Industrial sector data are divided into categories such as raw materials, equipment manufacturing, consumer goods, electronic information manufacturing, software and information technology services according to departmental responsibilities.

[0165] 2) According to business scope, operating model, business process, etc., refine the industry fields or clarify the key business classification of each business line.

[0166] c) Business attribute classification: Select appropriate business attributes and classify key business data in detail.

[0167] d) Determine classification rules: Sort out and analyze the data classification results of each key business, and determine the classification rules according to the data management and usage requirements of the industry.

[0168] Determine industry data classification rules, such as:

[0169] 1) Data classification rules can be given in the form of "business line - key business - business attribute classification";

[0170] 2) The data classification results of key businesses can also be classified and analyzed, and data subcategories with similar themes can be classified.

[0171] Through step S202, this embodiment classifies historical enterprise data, laying a solid data foundation for the subsequent construction of a knowledge graph for data retrieval and management.

[0172] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 3 , and can also include the following:

[0173] Step S301: performing a feature importance evaluation operation on the enterprise classification data according to a preset feature evaluation rule to determine corresponding important classification features, and performing a feature extraction operation on the important classification features to determine corresponding classification features;

[0174] Step S302: performing a classification label adding operation on the classification feature to determine the corresponding classification and classification feature.

[0175] Optionally, in this embodiment, the feature evaluation rule uses statistical tests to select the features most relevant to the target variable. Preferably, for complex feature variables, a machine learning model is used to evaluate the importance of features, such as a random forest, and the feature set is iteratively improved by building a model and repeatedly selecting the most influential features.

[0176] More specifically, when using the model for importance assessment, feature engineering is performed based on the type of data:

[0177] For text data, the bag-of-words model is used to convert text into vectors of word occurrences, by evaluating the importance of a word for one of the documents in a text collection.

[0178] For numerical data, use binning to split continuous numerical data into discrete intervals and generate polynomial combinations of the original features.

[0179] For categorical data, one-hot encoding is used to convert categorical variables into binary (0 or 1) form. Label encoding maps categories to numerical values.

[0180] After the above feature evaluation, principal component analysis was performed on the features to transform the potentially correlated variables into a set of linearly uncorrelated variables with a smaller number of values ​​through orthogonal transformation.

[0181] It is understandable that after the above-mentioned importance evaluation of features with different characteristics, useful and important data can be accurately extracted in a targeted manner, improving the accuracy of subsequent model training. The subsequent main component analysis of the features is the process of feature dimensionality reduction. The main purpose is to visualize the data for easy observation and exploration. It also helps to simplify the training and prediction of subsequent machine learning models.

[0182] Then, feature extraction is performed on the important features obtained through the above evaluation, and feature extraction is performed on the classified data to extract key features that are helpful for subsequent processing. For example, for text data, features such as keywords and sentence structure can be extracted. Feature extraction involves different extraction methods for different data:

[0183] For text data, use pre-trained models such as the BERT model to extract embedded representations of sentences or paragraphs.

[0184] For image data, the Juanju neural network is used to extract local image features.

[0185] For audio data, Mel-frequency cepstral coefficients are used to extract audio signal features.

[0186] The different features extracted above are recombined into new features to discover the potential relationship between them and obtain the extracted classification features.

[0187] Then, hierarchical labels are added to the extracted classification features. According to the risk performance of historical data, labels (such as "low", "medium", "high") are added to each feature or feature combination to obtain classification and hierarchical features with hierarchical labels.

[0188] Through step S302, this embodiment obtains classification and grading features for training the model, laying a solid data foundation for subsequent classification and grading model training.

[0189] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 4 , and can also include the following:

[0190] Step S401: inputting the classification and grading features into a preset long short-term memory model for model training to determine a corresponding loss function;

[0191] Step S402: performing parameter tuning operations on the long short-term memory model according to the minimum loss function and the back propagation algorithm to determine the corresponding classification and grading model.

[0192] Optionally, in this embodiment, the obtained classification and grading features are input into a long short-term memory model for model training, and the model parameters are optimized through iterative training to improve the accuracy of classification and grading.

[0193] Specifically, the classification and grading features are divided into a training set, a validation set and a test set. The training set is used to train the model to obtain the predicted features of the model output, the predicted features are compared with the real features in the validation set, and the corresponding loss function is calculated. Through iterative training, the minimum loss function in a series of loss functions is used as the final training loss, which is used to tune and update the parameters of each functional layer of the model through back propagation according to the minimum training loss, so as to obtain an accurate model for classification and grading.

[0194] Through step S402, this embodiment successfully trains an accurate and efficient classification and grading model, laying a foundation for the subsequent construction of a classification and grading intelligent agent.

[0195] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 5 , and can also include the following:

[0196] Step S501: performing a multi-dimensional index extraction operation on the historical enterprise data to determine corresponding multi-dimensional classification and grading indexes;

[0197] Step S502: Determine the corresponding setting classification and grading engine according to the multi-dimensional classification and grading indicators, preset indicator weights and preset judgment rules.

[0198] Optionally, in this embodiment, multi-dimensional indicator extraction and multi-dimensional indicator determination is a complex decision-making system that integrates multiple dimensional determination indicators, such as data attributes, business logic, legal and regulatory requirements, etc., and performs multi-dimensional and multi-level verification and evaluation on data classification and grading results. Through this system, the classification results of data can be judged more comprehensively and accurately, and the quality and efficiency of data management can be improved.

[0199] For multi-dimensional indicator determination, first conduct demand analysis and indicator definition to clarify the purpose, application scenarios and compliance requirements of data classification. Based on the results of demand analysis, define a series of multi-dimensional indicators for data classification determination and multi-dimensional indicators for data grading determination. These indicators can include basic attributes of data (such as data type, data source), business logic indicators (such as business processes, business scenarios), legal and regulatory indicators (such as data protection regulations, personal information processing specifications), confidentiality levels (low, medium, high), etc.

[0200] After extracting the above-mentioned multi-dimensional classification and grading indicators, assign corresponding weights to each indicator according to its importance and impact.

[0201] Next, design the judgment logic of multi-dimensional indicators, clarify the correlation, judgment order and judgment conditions between the indicators. Through logical design, ensure that the multi-dimensional indicators can work together and act together in the judgment process of data classification results.

[0202] Finally, the designed decision logic is encoded into a computer-executable rule set. The rule set should contain information such as the complete decision process, decision conditions, and decision result output. The classification and grading engine is constructed through the extracted multi-dimensional classification and grading indicators, the importance weights assigned to the indicators, and the decision logic.

[0203] Through step S502, this embodiment successfully builds a classification and grading engine to classify and grade the received data, and the rule engine can ensure the accuracy of the classification and grading.

[0204] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 6 , and can also include the following:

[0205] Step S601: classify and grade the real-time enterprise data according to the classification and grading model in the classification and grading agent to determine the corresponding first enterprise classification and grading data;

[0206] Step S602: performing classification and grading operations on the real-time enterprise data according to the set classification and grading engine in the classification and grading agent to determine corresponding second enterprise classification and grading data;

[0207] Step S603: Determine corresponding enterprise classification and grading data according to the first enterprise classification and grading data, the second enterprise classification and grading data and a preset fusion rule.

[0208] Optionally, in the present embodiment, this step is to construct a classification and grading intelligent agent, perform classification and grading according to the intelligent agent, obtain real-time enterprise data, obtain corresponding classification and grading data through the classification and grading model and the classification and grading engine respectively, and intelligently integrate the results of the large model engine and the results based on the rule engine to complement each other and further improve the initial accuracy, wherein the priority of the rule-based engine is higher than that of the large model engine.

[0209] That is to say, when the output of the rule engine is inconsistent with the big model engine, the output of the rule engine shall prevail, because the output of the rule engine is more accurate. Subsequently, the big model engine can be re-adjusted and trained based on the output of the rule engine, and the big model engine can be iteratively tuned to increase the accuracy of the big model engine.

[0210] When there is data that is not included in the rule engine, the prediction of the large model engine is output. After manual review, it can be written into the rule engine by defining rules. In this way, not only the accuracy of data classification and grading is increased, but also the versatility of data classification and grading is increased, realizing customized classification not only for one enterprise and one business, but also providing a broad database.

[0211] For example, taking the production process data of a manufacturing enterprise as an example, the specific implementation steps are as follows:

[0212] Data acquisition: Obtain production process data in real time from the production system, including data on each link such as raw material warehousing, production and processing, and finished product delivery.

[0213] Large model engine classification and grading: Use the trained classification and grading model to automatically classify and grade the pre-processed production process data, and preliminarily identify the data types and data levels of different production links.

[0214] Classification and grading of rule engines: Based on the production process data classification and grading rules formulated by the enterprise (such as data type, business logic, data sensitivity, etc.), the results of the large model engine are verified and adjusted in multiple dimensions. If the final results are inconsistent, the rule engine will take priority.

[0215] For example, data involving sensitive information (such as customer information and financial data) needs to be managed and protected at a higher security level regardless of how the big model classifies it.

[0216] Result fusion and output: Intelligently integrate the results of the large model engine and the adjustment results based on the multi-dimensional indicator judgment engine to ensure that the final classification and grading results are consistent with both business logic and compliance requirements.

[0217] Through the above embodiments, the company has successfully achieved accurate classification and grading management of production process data, improving the efficiency and security of data management.

[0218] Through step S603, this embodiment successfully obtains accurate classification and grading data, laying the foundation for the subsequent construction of the knowledge graph.

[0219] In one embodiment of the knowledge graph construction method based on data classification and grading of the present application, see Figure 7 , and can also include the following:

[0220] Step S701: judging whether the second enterprise classification and grading data is consistent with the first enterprise classification and grading data according to the priority fusion rule;

[0221] Step S702: If they are inconsistent, the second enterprise classification and grading data is determined as the enterprise classification and grading data.

[0222] Optionally, in this embodiment, the priority fusion rule refers to a rule-based engine with a higher priority than a large model engine.

[0223] That is to say, when the output of the rule engine is inconsistent with the big model engine, the output of the rule engine shall prevail, because the output of the rule engine is more accurate. Subsequently, the big model engine can be re-adjusted and trained based on the output of the rule engine, and the big model engine can be iteratively tuned to increase the accuracy of the big model engine.

[0224] In this case, the output classification and grading data can not only ensure versatility but also accuracy.

[0225] Through step S702, this embodiment successfully obtains classified and graded data, laying the foundation for the subsequent construction of the knowledge graph.

[0226] In order to improve the efficiency and accuracy of data classification and grading, the present application provides an embodiment of a knowledge graph construction device based on data classification and grading for realizing all or part of the content of the knowledge graph construction method based on data classification and grading, see Figure 8 The knowledge graph construction device based on data classification and grading specifically includes the following contents:

[0227] The enterprise classification data determination module 10 is used to obtain historical enterprise data, classify the historical enterprise data according to the preset industry field classification rules and the business attribute rules within the preset industry field, and determine the corresponding enterprise classification data;

[0228] The classification and grading agent determination module 20 is used to perform feature extraction operations on the enterprise classification data, perform grading label addition operations on the classification features after the feature extraction, determine the corresponding classification and grading features, input the classification and grading features into a preset initial model for model training, and determine the corresponding classification and grading model; perform agent construction operations according to the classification and grading model and the set classification and grading engine, and determine the corresponding classification and grading agent;

[0229] The classification and grading knowledge graph construction module 30 is used to obtain real-time enterprise data, classify and grade the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct the corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0230] From the above description, it can be seen that the knowledge graph construction device based on data classification and grading provided in the embodiment of the present application can obtain historical enterprise data, determine the corresponding enterprise classification data according to preset industry field classification rules and business attribute rules within the preset industry field; perform a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, input the classification and grading features into a preset initial model for model training, and determine the corresponding classification and grading model; perform an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; perform classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and construct a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0231] From the hardware level, in order to improve the efficiency and accuracy of data classification and grading, the present application provides an embodiment of an electronic device for implementing all or part of the content of the knowledge graph construction method based on data classification and grading, and the electronic device specifically includes the following content:

[0232] Processor, memory, communication interface and bus; wherein the processor, memory and communication interface communicate with each other through the bus; the communication interface is used to realize the information transmission between the knowledge graph construction method based on data classification and grading and the core business system, user terminal and related databases and other related devices; the logic controller can be a desktop computer, a tablet computer and a mobile terminal, etc., but the present embodiment is not limited thereto. In the present embodiment, the logic controller can be implemented with reference to the embodiment of the knowledge graph construction method based on data classification and grading in the embodiment, and the embodiment of the knowledge graph construction method based on data classification and grading, and the contents thereof are incorporated herein, and the repeated parts are not repeated.

[0233] It is understandable that the user terminal may include a smart phone, a tablet electronic device, a network set-top box, a portable computer, a desktop computer, a personal digital assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Among them, the smart wearable device may include smart glasses, a smart watch, a smart bracelet, etc.

[0234] In practical applications, part of the method for constructing a knowledge graph based on data classification and grading can be performed on the electronic device side as described above, or all operations can be completed in the client device. The specific selection can be based on the processing capability of the client device and the limitations of the user's usage scenario. This application does not limit this. If all operations are completed in the client device, the client device may also include a processor.

[0235] The client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side, and other implementation scenarios may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, or a server cluster consisting of multiple servers, or a server structure of a distributed device.

[0236] Fig. 9 FIG. 9 is a schematic block diagram of the system structure of the electronic device 9600 according to an embodiment of the present application. Fig. 9 As shown, the electronic device 9600 may include a central processor 9100 and a memory 9140; the memory 9140 is coupled to the central processor 9100. It is worth noting that Fig. 9 is exemplary; other types of structures may also be used to supplement or replace this structure to implement telecommunication functions or other functions.

[0237] In one embodiment, the knowledge graph construction method based on data classification and grading can be integrated into the central processing unit 9100. The central processing unit 9100 can be configured to perform the following control:

[0238] Step S101: Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data;

[0239] Step S102: performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining the corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining the corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and the set classification and grading engine, and determining the corresponding classification and grading intelligent agent;

[0240] Step S103: Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0241] From the above description, it can be seen that the electronic device provided in the embodiment of the present application obtains historical enterprise data, determines the corresponding enterprise classification data according to the preset industry field classification rules and the business attribute rules within the preset industry field; performs a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, inputs the classification and grading features into the preset initial model for model training, and determines the corresponding classification and grading model; performs an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; performs classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and constructs a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0242] In another embodiment, the knowledge graph construction method based on data classification and grading can be configured separately from the central processing unit 9100. For example, the knowledge graph construction method based on data classification and grading can be configured as a chip connected to the central processing unit 9100, and the function of the knowledge graph construction method based on data classification and grading can be realized through the control of the central processing unit.

[0243] like Fig. 9As shown, the electronic device 9600 may also include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily have to include Fig. 9 In addition, the electronic device 9600 may also include Fig. 9 For components not shown, reference may be made to the prior art.

[0244] like Fig. 9 As shown, the central processing unit 9100 is sometimes also referred to as a controller or an operation control, and may include a microprocessor or other processor device and / or logic device, which receives input and controls the operation of various components of the electronic device 9600.

[0245] The memory 9140 may be, for example, one or more of a cache, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory or other suitable devices. The above-mentioned information related to the failure may be stored, and a program for executing the relevant information may also be stored. The CPU 9100 may execute the program stored in the memory 9140 to implement information storage or processing, etc.

[0246] The input unit 9120 provides input to the central processing unit 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used to display display objects such as images and texts. The display may be, for example, an LCD display, but is not limited thereto.

[0247] The memory 9140 may be a solid-state memory, such as a read-only memory (ROM), a random access memory (RAM), a SIM card, etc. It may also be a memory that saves information even when the power is off, can be selectively erased, and is provided with more data, examples of which are sometimes referred to as EPROMs, etc. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142, which is used to store application programs and function programs or processes for executing the operation of the electronic device 9600 through the central processor 9100.

[0248] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various drivers for communication functions of the electronic device and / or for executing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0249] The communication module 9110 is a transmitter / receiver that sends and receives signals via the antenna 9111. The communication module 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, which may be the same as the case of a conventional mobile communication terminal.

[0250] Based on different communication technologies, multiple communication modules 9110 may be provided in the same electronic device, such as a cellular network module, a Bluetooth module and / or a wireless LAN module, etc. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor 9130 is also coupled to the central processor 9100, so that the audio can be recorded on the local machine through the microphone 9132, and the audio stored on the local machine can be played through the speaker 9131.

[0251] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all the steps in the method for constructing a knowledge graph based on data classification and grading in the above-mentioned embodiment, where the execution subject is a server or a client. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, all the steps in the method for constructing a knowledge graph based on data classification and grading in the above-mentioned embodiment are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0252] Step S101: Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data;

[0253] Step S102: performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining the corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining the corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and the set classification and grading engine, and determining the corresponding classification and grading intelligent agent;

[0254] Step S103: Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0255] From the above description, it can be seen that the computer-readable storage medium provided in the embodiment of the present application obtains historical enterprise data, determines the corresponding enterprise classification data according to preset industry field classification rules and business attribute rules within the preset industry field; performs a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, inputs the classification and grading features into a preset initial model for model training, and determines the corresponding classification and grading model; performs an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; performs classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and constructs a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0256] The embodiments of the present application also provide a computer program product capable of implementing all the steps in the method for constructing a knowledge graph based on data classification and grading in the above-mentioned embodiment, where the execution subject is a server or a client. When the computer program / instruction is executed by a processor, the steps of the method for constructing a knowledge graph based on data classification and grading are implemented. For example, the computer program / instruction implements the following steps:

[0257] Step S101: Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data;

[0258] Step S102: performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining the corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining the corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and the set classification and grading engine, and determining the corresponding classification and grading intelligent agent;

[0259] Step S103: Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

[0260] From the above description, it can be seen that the computer program product provided in the embodiment of the present application obtains historical enterprise data, determines the corresponding enterprise classification data according to preset industry field classification rules and business attribute rules within the preset industry field; performs a grading label adding operation on the enterprise classification data to determine the corresponding classification and grading features, inputs the classification and grading features into a preset initial model for model training, and determines the corresponding classification and grading model; performs an intelligent body construction operation according to the classification and grading model and the set classification and grading engine to determine the corresponding classification and grading intelligent body; performs classification and grading operations on real-time enterprise data according to the classification and grading intelligent body to determine the corresponding enterprise classification and grading data, and constructs a corresponding classification and grading knowledge graph, thereby improving the efficiency and accuracy of data classification and grading.

[0261] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0262] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0263] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0264] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0265] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for constructing a knowledge graph based on data classification and grading, characterized in that: The method comprises: Acquire historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine corresponding enterprise classification data; Performing feature extraction on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, determining corresponding classification and grading features, inputting the classification and grading features into a preset initial model for model training, and determining a corresponding classification and grading model; performing an intelligent agent construction operation according to the classification and grading model and a set classification and grading engine, and determining a corresponding classification and grading intelligent agent; Acquire real-time enterprise data, perform classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

2. The method for constructing a knowledge graph based on data classification and grading according to claim 1 is characterized in that: The data classification of the historical enterprise data according to the preset industry field classification rules and the business attribute rules within the preset industry field to determine the corresponding enterprise classification data includes: Performing data classification operations on the historical enterprise data according to preset industry field classification rules to determine corresponding industry field classification data, wherein the industry field classification data is obtained by classifying data according to industry fields; The industry field classification data is refined and classified according to the business attribute rules in the preset industry field to determine the corresponding enterprise classification data, wherein the enterprise classification data is obtained by further classifying according to the business attributes in the industry field to which it belongs on the basis of the industry field classification.

3. The method for constructing a knowledge graph based on data classification and grading according to claim 1 is characterized in that: The step of performing a feature extraction operation on the enterprise classification data, performing a classification label adding operation on the classification features after the feature extraction, and determining corresponding classification and grading features includes: Performing a feature importance evaluation operation on the enterprise classification data according to a preset feature evaluation rule to determine corresponding important classification features, and performing a feature extraction operation on the important classification features to determine corresponding classification features; A classification label adding operation is performed on the classification feature to determine the corresponding classification and classification feature.

4. The method for constructing a knowledge graph based on data classification and grading according to claim 1 is characterized in that: The step of inputting the classification and grading features into a preset initial model for model training to determine a corresponding classification and grading model comprises: Input the classification and grading features into a preset long short-term memory model for model training, and determine the corresponding loss function; The long short-term memory model is subjected to parameter tuning operations according to the minimum loss function and the back propagation algorithm to determine the corresponding classification and grading model.

5. The method for constructing a knowledge graph based on data classification and grading according to claim 1 is characterized in that: Before performing the agent construction operation according to the classification and grading model and setting the classification and grading engine to determine the corresponding classification and grading agent, the method includes: Performing a multi-dimensional index extraction operation on the historical enterprise data to determine corresponding multi-dimensional classification and grading indicators; According to the multi-dimensional classification and grading indicators, preset indicator weights and preset judgment rules, a corresponding classification and grading engine is determined.

6. The method for constructing a knowledge graph based on data classification and grading according to claim 1, characterized in that: The step of performing classification and grading operations on the real-time enterprise data according to the classification and grading intelligent agent to determine corresponding enterprise classification and grading data includes: Classify and grade the real-time enterprise data according to the classification and grading model in the classification and grading agent to determine the corresponding first enterprise classification and grading data; Performing classification and grading operations on the real-time enterprise data according to the set classification and grading engine in the classification and grading intelligent agent to determine corresponding second enterprise classification and grading data; Corresponding enterprise classification and grading data are determined according to the first enterprise classification and grading data, the second enterprise classification and grading data, and a preset fusion rule.

7. The method for constructing a knowledge graph based on data classification and grading according to claim 6, characterized in that: The determining corresponding enterprise classification and grading data according to the first enterprise classification and grading data, the second enterprise classification and grading data and a preset fusion rule includes: Determining whether the second enterprise classification and grading data is consistent with the first enterprise classification and grading data according to a priority fusion rule; If they are inconsistent, the second enterprise classification and grading data is determined as the enterprise classification and grading data.

8. A knowledge graph construction device based on data classification and grading, characterized in that: The device comprises: An enterprise classification data determination module is used to obtain historical enterprise data, classify the historical enterprise data according to preset industry field classification rules and business attribute rules within the preset industry field, and determine the corresponding enterprise classification data; A classification and grading agent determination module is used to perform feature extraction operations on the enterprise classification data, perform grading label addition operations on the classification features after the feature extraction, determine the corresponding classification and grading features, input the classification and grading features into a preset initial model for model training, and determine the corresponding classification and grading model; perform agent construction operations according to the classification and grading model and the set classification and grading engine, and determine the corresponding classification and grading agent; A classification and grading knowledge graph construction module is used to obtain real-time enterprise data, classify and grade the real-time enterprise data according to the classification and grading intelligent agent, determine the corresponding enterprise classification and grading data, perform entity definition operations and relationship extraction operations on the enterprise classification and grading data, determine the corresponding enterprise classification and grading data entities and the relationships between entities, and construct a corresponding classification and grading knowledge graph based on the enterprise classification and grading data entities and the relationships between entities.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for constructing a knowledge graph based on data classification and grading as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for constructing a knowledge graph based on data classification and grading as described in any one of claims 1 to 7 are implemented.

Images