Batch data tagging method and device, equipment and storage medium

Abstract

According to the batch data tagging method and device, the equipment and the storage medium provided by the invention, the to-be-processed data is acquired, and the to-be-processed data comprises the unstructured data, the relational network data and the structured data; inputting the unstructured data into a feature extraction model to obtain deep features; inputting the relation network data into a relation feature model to obtain relation features; inputting the structured data into a label prediction model to obtain a label prediction result; and correcting the label prediction result according to the deep features and the relation features to obtain a target label result. According to the method, the deep features are extracted from the unstructured data, the relation features are extracted from the relation network data, the label prediction result of the structured data is combined, the information of the multi-dimensional data is fully utilized, the enterprise label generation precision and diversity are improved, and the enterprise label generation efficiency is improved. And information loss or deviation possibly caused by only depending on a single data source is avoided.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a batch data labeling method, device, equipment and storage medium. Background Art

[0002] In modern data processing and management, labeling is a key step, especially when managing enterprises in government systems. Enterprise data is large and complex, including structured, unstructured, and relational network data. Existing labeling methods are mostly based on a single data source, often using only structured data for simple classification and label generation, which is difficult to handle complex, multi-dimensional data, resulting in insufficient accuracy and diversity in label generation. In addition, traditional data processing methods are inefficient and cannot meet the needs of government systems for real-time processing of large-scale data.

[0003] In summary, the problems existing in the prior art need to be solved urgently. Summary of the invention

[0004] The present invention provides a batch data labeling method, device, equipment and storage medium, which are used to solve the defects in the prior art and realize automatic data labeling.

[0005] The present invention provides a batch data labeling method, comprising:

[0006] Acquiring data to be processed, wherein the data to be processed includes unstructured data, relationship network data, and structured data;

[0007] Inputting the unstructured data into a feature extraction model to obtain deep features;

[0008] Inputting the relationship network data into a relationship feature model to obtain relationship features;

[0009] Inputting the structured data into a label prediction model to obtain a label prediction result;

[0010] The label prediction result is corrected according to the deep features and the relationship features to obtain a target label result.

[0011] According to a batch data labeling method provided by the present invention, the step of correcting the label prediction result according to the deep features and the relationship features to obtain the target label result specifically includes:

[0012] Inputting the deep features, the relationship features and the label prediction results into the first fully connected layer for nonlinear transformation;

[0013] The transformed deep features, relationship features and label prediction results are input into the second fully connected layer for feature fusion to obtain the target feature vector;

[0014] The target feature vector is input into the output layer to obtain the target label result.

[0015] According to a batch data labeling method provided by the present invention, the step of inputting the transformed deep features, relationship features and label prediction results into the second fully connected layer to perform feature fusion is specifically implemented in the following manner:

[0016] Z fused =FC2([FC1(Z self-supervised ),FC1(Z GNN ),FC1(Z MLP )])

[0017] Among them, FC1 represents the operation of the first fully connected layer, FC2 represents the operation of the second fully connected layer, and Z self-supervised is the deep feature, Z GNN is the relationship feature, Z MLP Predict the result for the label.

[0018] According to a batch data labeling method provided by the present invention, the feature extraction model is a self-supervised learning model, the relationship feature model is a GNN model, and the label prediction model is an MLP model.

[0019] According to a batch data labeling method provided by the present invention, the activation function of the output layer is Sigmoid.

[0020] According to a batch data labeling method provided by the present invention, before the step of obtaining the data to be processed, the method further includes:

[0021] The data to be processed is preprocessed, and the preprocessing includes missing value filling, outlier processing and normalization processing.

[0022] According to a batch data labeling method provided by the present invention, after the step of correcting the label prediction result according to the deep feature and the relationship feature to obtain the target label result, the method further includes:

[0023] Get label correction data;

[0024] The target label result is corrected according to the label correction data.

[0025] The present invention also provides a batch data labeling device, comprising:

[0026] A data acquisition module, used to acquire data to be processed, wherein the data to be processed includes unstructured data, relationship network data and structured data;

[0027] A deep feature module, used for inputting the unstructured data into a feature extraction model to obtain deep features;

[0028] A relationship feature module, used for inputting the relationship network data into a relationship feature model to obtain relationship features;

[0029] A label prediction module, used to input the structured data into a label prediction model to obtain a label prediction result;

[0030] The result correction module is used to correct the label prediction result according to the deep feature and the relationship feature to obtain the target label result.

[0031] The present invention also 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, a batch data labeling method as described in any one of the above is implemented.

[0032] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described batch data labeling methods.

[0033] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the batch data labeling method described in any one of the above methods is implemented.

[0034] The batch data labeling method, device, equipment and storage medium provided by the present invention obtain the data to be processed, wherein the data to be processed includes unstructured data, relational network data and structured data; input the unstructured data into a feature extraction model to obtain deep features; input the relational network data into a relational feature model to obtain relational features; input the structured data into a label prediction model to obtain a label prediction result; and according to the deep features and the relational features, the label prediction result is corrected to obtain a target label result. The method of the present invention makes full use of the information of multi-dimensional data by extracting deep features from unstructured data, extracting relational features from relational network data, and combining the label prediction results of structured data, thereby improving the accuracy and diversity of enterprise label generation and avoiding information loss or deviation that may be caused by relying solely on a single data source. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 It is a flow chart of the batch data labeling method provided by the present invention;

[0037] Figure 2 It is a structural schematic diagram of a batch data labeling device provided by the present invention;

[0038] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In order to solve the problems in the prior art, the present invention proposes a batch data labeling method to achieve automatic data labeling. The batch data labeling method is described below. Figure 1 As shown, including but not limited to the following steps:

[0041] Step 110: Obtain data to be processed, where the data to be processed includes unstructured data, relationship network data, and structured data.

[0042] In this step, firstly, the enterprise data to be processed is obtained from multiple data sources. The data includes:

[0043] Unstructured data: such as company profiles, news reports, audit reports, product descriptions, etc. This type of data cannot be directly used for structured analysis and needs to be further processed.

[0044] Relationship network data: such as equity relationships, cooperative relationships, upstream and downstream supply chain information between enterprises. This data can be represented by a graph structure to reflect the correlation between enterprises.

[0045] Structured data: such as corporate registration information, industry classification, operating conditions, financial statements, administrative approval records, etc. This type of data already has clear fields and formats, which are suitable for traditional statistical analysis and modeling.

[0046] These data can be obtained from multiple sources, such as databases in government systems, third-party data platforms, web crawlers, or the company's own official website.

[0047] Step 120: Input the unstructured data into a feature extraction model to obtain deep features.

[0048] In this step, the system inputs the acquired unstructured data into the feature extraction model. The feature extraction model can adopt a self-supervised learning model, such as a model based on the Transformer architecture (such as BERT, GPT, etc.), which can extract label-related data from the data, and then use encoding models such as BERT to convert the extracted content into deep features.

[0049] In specific implementation, unstructured data is first preprocessed, including removing irrelevant characters, filtering stop words, and segmenting words. The preprocessed text is input into the self-supervised learning model, and deep semantic features are extracted through multi-layer neural network operations. These deep features can reflect the specific behavior patterns or attributes of the enterprise in the text data, and the generated feature vectors can be used for subsequent labeling analysis.

[0050] Step 130: Input the relationship network data into a relationship feature model to obtain relationship features.

[0051] In this step, the system inputs the relationship network data into the relationship feature model. The relationship feature model is usually a graph neural network (GNN) model, which is used to analyze the complex relationship network between enterprises and generate a feature representation of each enterprise in the network.

[0052] Specifically, the system uses enterprises as nodes in the graph and the relationships between enterprises (such as equity relationships, cooperative relationships, upstream and downstream supply chains, etc.) as edges in the graph. Through the GNN model, the system can effectively capture the interactions between enterprises and extract the characteristics of each enterprise in the relationship network (i.e., relationship characteristics). These characteristics reflect the position of the enterprise in the entire economic or industry ecosystem and the strength of its relationship with other enterprises.

[0053] The GNN model propagates and aggregates the information of neighbor nodes layer by layer through multiple graph convolution operations, thereby generating a feature representation of each node (enterprise).

[0054] Step 140: Input the structured data into a label prediction model to obtain a label prediction result.

[0055] In this step, the system inputs the structured data into the label prediction model. The label prediction model can adopt a multi-layer perceptron (MLP) model, which is good at processing structured data and can generate preliminary label prediction results based on the basic information of the enterprise.

[0056] In specific implementation, structured data (such as registered capital, number of employees, business industry, administrative approval status, etc.) is input into the MLP model, and after calculation by a multi-layer neural network, one or more label prediction results are output. These prediction results can be the probability distribution of a certain enterprise belonging to a certain label category, or they can be multi-classified label indicators.

[0057] Step 150: According to the deep features and the relationship features, the label prediction result is corrected to obtain a target label result.

[0058] In this step, the system combines deep features, relationship features, and label prediction results to generate the final target label result.

[0059] Specifically, the deep features extracted by self-supervised learning, the relational features extracted by GNN, and the label prediction results of MLP are first input into a fusion module. The fusion module is processed through multiple fully connected layers (i.e., MLP), and the three sets of input features are first nonlinearly transformed in the first fully connected layer. The transformed features are further fused in the second fully connected layer to generate a unified feature representation.

[0060] The target feature vector after feature fusion is input to the output layer, and the final target label result is generated through an appropriate activation function (such as Sigmaid). The target label result not only combines the label prediction of structured data, but also incorporates the deep information in unstructured data and the upstream and downstream associations in the relationship network, so as to more accurately reflect the characteristics and behavior patterns of the enterprise.

[0061] This implementation acquires different types of data and uses self-supervised learning models, GNN models, and MLP models to extract prediction results of deep features, relational features, and structured data, and finally corrects label predictions through feature fusion to generate accurate enterprise labels. This method effectively improves the accuracy and diversity of label generation and is suitable for labeling large-scale enterprise data in government affairs, finance, and other fields.

[0062] As a further optional embodiment, the step of correcting the label prediction result according to the deep feature and the relationship feature to obtain the target label result specifically includes:

[0063] Inputting the deep features, the relationship features and the label prediction results into the first fully connected layer for nonlinear transformation;

[0064] The transformed deep features, relationship features and label prediction results are input into the second fully connected layer for feature fusion to obtain the target feature vector;

[0065] The target feature vector is input into the output layer to obtain the target label result.

[0066] As a further optional embodiment, the step of inputting the transformed deep features, relationship features and label prediction results into the second fully connected layer to perform feature fusion is specifically implemented in the following manner:

[0067] Z fused =FC2([FC1(Z self-supervised ),FC1(Z GNN ),FC1(Z MLP )])

[0068] Among them, FC1 represents the operation of the first fully connected layer, FC2 represents the operation of the second fully connected layer, and Z self-supervised is the deep feature, Z GNN is the relationship feature, Z MLP Predict the result for the label.

[0069] In this step, the deep features obtained from the self-supervised learning model, the relational features obtained from the GNN model, and the label prediction results obtained by the MLP model are first input into the first fully connected layer (FC1) for processing. The operation of the fully connected layer realizes the nonlinear mapping of features through linear transformation plus activation function. The purpose of this operation is to perform preliminary dimensionality reduction and nonlinear transformation on each type of feature to capture more complex relationships between features.

[0070] In this step, the three sets of features output by the first fully connected layer are concatenated to form a unified feature vector. Then, the feature vector is input into the second fully connected layer (FC2) for further feature fusion. Through the feature fusion of the fully connected layer, the system can capture the relationship between different types of features, thereby generating a more representative target feature vector.

[0071] The specific operations are as follows:

[0072] Z fused =FC2([FC1(Z self-supervised ),FC1(Z GNN ),FC1(Z MLP )])

[0073] Among them, FC1 represents the operation of the first fully connected layer, FC2 represents the operation of the second fully connected layer, and Z self-supervised is the deep feature, ZGNN is the relationship feature, Z MLP Predict the result for the label.

[0074] In the last step, the fused target feature vector Z fused Input to the output layer. The output layer can use appropriate activation functions for label prediction according to specific task requirements. Common activation functions include:

[0075] Sigmoid: used for binary classification label problems, the output result is the probability value of the label belonging to a certain category.

[0076] Softmax: Used for multi-classification labeling problems and outputs the probability distribution of each class of labels.

[0077] According to the application scenario, the output layer will generate the final target label result. This result has been comprehensively corrected by deep features, relationship features and MLP predictions, and more accurately reflects the multi-dimensional characteristics of the enterprise.

[0078] Through the above steps, this embodiment can effectively perform nonlinear transformation and feature fusion on the deep features extracted by self-supervised learning, the relational features extracted by GNN, and the prediction results of MLP. Through the processing of two fully connected layers, the system can realize the comprehensive extraction of meaningful features from multidimensional data and generate accurate labeling results. This method greatly improves the accuracy and diversity of label generation and is suitable for large-scale enterprise data labeling tasks.

[0079] As a further optional embodiment, the feature extraction model is a self-supervised learning model, the relationship feature model is a GNN model, and the label prediction model is an MLP model.

[0080] Self-supervised learning models are used to process unstructured data, such as text, images, or other complex enterprise-related information. The model extracts deep features from the data by pre-training without explicit labels. Self-supervised learning can extract representative high-dimensional feature vectors from large amounts of data by designing specific tasks (such as occluded word prediction, contrastive learning, etc.). Ultimately, these high-dimensional feature vectors are able to capture the complex semantic information of the enterprise, such as the business scope and development history of the enterprise. The role of this model is to generate deep features related to the enterprise and provide input for subsequent feature fusion.

[0081] The graph neural network (GNN) model is mainly used to process the relationship network data between enterprises, such as the equity structure, business cooperation relationship, management overlap, upstream and downstream supply chain, etc. These relationships form a graph structure, where nodes represent enterprises and edges represent certain connections between enterprises. The GNN model can effectively capture the relationship characteristics between enterprises by propagating and aggregating the information of node neighbors. Through multi-layer graph convolution operations, GNN can extract the global representation of nodes (enterprises) in the relationship network, namely, the relationship characteristics. These relationship characteristics reflect the mutual influence between enterprises and help take into account the business ecosystem in which the enterprise is located when generating labels.

[0082] The multi-layer perceptron (MLP) model is used to process structured data, such as corporate registration information, annual report data, financial data, social security number, number of patents, etc. These data are usually standardized numerical data. The MLP model can extract features from these structured data and generate preliminary label prediction results through a multi-layer fully connected network structure combined with nonlinear activation functions (such as ReLU).

[0083] By learning from these structured data, the MLP model can predict the label category to which the enterprise belongs, such as "high-tech enterprise" and "key regulatory target". This label prediction result will serve as the basis for subsequent feature fusion and correction.

[0084] As a further optional embodiment, the activation function of the output layer is Sigmoid.

[0085] The Sigmoid function is an activation function commonly used in binary classification tasks, and its output value ranges from 0 to 1. When the feature fused vector is input to the output layer, the Sigmoid function converts it into a probability value, which indicates the probability that the sample belongs to a specific category. The probability value usually sets a threshold (such as 0.5) to determine whether the label is valid. For example, if the output value is greater than 0.5, the label is 1; if it is less than 0.5, the label is 0.

[0086] As a further optional embodiment, before the step of obtaining the data to be processed, the method further includes:

[0087] The data to be processed is preprocessed, and the preprocessing includes missing value filling, outlier processing and normalization processing.

[0088] In practical applications, some fields in the data set may have missing values ​​(such as some financial indicators or administrative information of the enterprise). The existence of missing values ​​will affect the learning effect of the model, so these missing data need to be processed during preprocessing. Common processing methods include: mean filling, median filling, mode filling and prediction filling.

[0089] Outliers are extreme values ​​in a data set that are far away from other data points. They may be caused by data input errors, measurement errors, or other abnormal situations. If outliers are not handled, the model may overfit to these extreme values, thus affecting the accuracy of label prediction. Common outlier processing methods include: direct elimination, range restriction, and scaling.

[0090] Normalization is used to scale the data to a uniform scale to ensure that the performance of the model is not affected by differences in numerical ranges between different features.

[0091] By preprocessing the data to be processed, including missing value filling, outlier processing and normalization, the quality and consistency of the data can be greatly improved, ensuring that the model can better learn the characteristics of the data. The preprocessed data will be more suitable for input into subsequent self-supervised learning models, GNN models and MLP models, thereby improving the accuracy and robustness of label generation.

[0092] As a further optional embodiment, after the step of correcting the label prediction result according to the deep feature and the relationship feature to obtain the target label result, the method further includes:

[0093] Get label correction data;

[0094] The target label result is corrected according to the label correction data.

[0095] Label correction data refers to reference data for further adjustment of labels automatically generated by the model based on actual usage feedback, expert annotations, or manual intervention information provided by external data sources during system operation. Ways to obtain label correction data include:

[0096] User feedback: After enterprise labels are generated, the system can provide an interface for supervisors or other users to provide feedback on the accuracy of the labels, such as deleting inappropriate labels, adding missing labels, etc.

[0097] Expert labeling: Invite domain experts to manually review the labels of certain companies and record the revised labels recommended by the experts.

[0098] Update of external data sources: Further improve existing enterprise label data by integrating external trusted data sources (such as industry reports, updated policies from regulatory authorities, etc.).

[0099] By obtaining these label correction data, a more accurate reference can be provided for the label generation results, ensuring that the enterprise labels match the actual business scenarios.

[0100] After obtaining the label correction data, the system will adjust the originally generated label results based on this data. Specific operations may include:

[0101] Tag addition: If the correction data indicates that a certain tag should be added to a certain enterprise, the system will add the tag to the tag collection of the enterprise.

[0102] Tag removal: If feedback indicates that a tag for a business is incorrect or inappropriate, the tag will be removed from the business's tag collection.

[0103] Label replacement: When the revised data indicates that the original label is not accurate or the description is not comprehensive enough, the system will replace the label based on the revised data.

[0104] By introducing label correction data and its corresponding correction steps, the flexibility and accuracy of the label generation system can be effectively improved. Especially in the early stage of the system or during large-scale deployment, combined with user feedback, expert annotation and other means, the quality of enterprise labels can be gradually optimized to ensure that the labels generated by the system are more in line with actual needs.

[0105] This mechanism can also achieve self-improvement. As more correction data accumulates, the label generation system can continue to learn and the correction rules become more intelligent, thereby greatly improving the accuracy of model predictions and the business adaptability of enterprise labels.

[0106] The batch data labeling device provided by the present invention is described below. Figure 2 As shown, the batch data labeling device described below and the batch data labeling method described above can correspond to each other.

[0107] A batch data labeling device, comprising:

[0108] The data acquisition module 210 is used to acquire the data to be processed, wherein the data to be processed includes unstructured data, relationship network data and structured data;

[0109] A deep feature module 220, used for inputting the unstructured data into a feature extraction model to obtain deep features;

[0110] A relationship feature module 230, used for inputting the relationship network data into a relationship feature model to obtain relationship features;

[0111] The label prediction module 240 is used to input the structured data into the label prediction model to obtain a label prediction result;

[0112] The result correction module 250 is used to correct the label prediction result according to the deep features and the relationship features to obtain a target label result.

[0113] Figure 3 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 3 As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330 and a communication bus 340, wherein the processor 310, the communications interface 320 and the memory 330 communicate with each other through the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute the batch data labeling method, which includes:

[0114] Acquiring data to be processed, wherein the data to be processed includes unstructured data, relationship network data, and structured data;

[0115] Inputting the unstructured data into a feature extraction model to obtain deep features;

[0116] Inputting the relationship network data into a relationship feature model to obtain relationship features;

[0117] Inputting the structured data into a label prediction model to obtain a label prediction result;

[0118] The label prediction result is corrected according to the deep features and the relationship features to obtain a target label result.

[0119] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0120] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the batch data labeling method provided by the above methods, the method includes:

[0121] Acquiring data to be processed, wherein the data to be processed includes unstructured data, relationship network data, and structured data;

[0122] Inputting the unstructured data into a feature extraction model to obtain deep features;

[0123] Inputting the relationship network data into a relationship feature model to obtain relationship features;

[0124] Inputting the structured data into a label prediction model to obtain a label prediction result;

[0125] The label prediction result is corrected according to the deep features and the relationship features to obtain a target label result.

[0126] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform the batch data labeling method provided by the above methods, the method comprising:

[0127] Acquiring data to be processed, wherein the data to be processed includes unstructured data, relationship network data, and structured data;

[0128] Inputting the unstructured data into a feature extraction model to obtain deep features;

[0129] Inputting the relationship network data into a relationship feature model to obtain relationship features;

[0130] Inputting the structured data into a label prediction model to obtain a label prediction result;

[0131] The label prediction result is corrected according to the deep features and the relationship features to obtain a target label result.

[0132] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0133] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A batch data labeling method, characterized in that: include: Acquiring data to be processed, wherein the data to be processed includes unstructured data, relationship network data, and structured data; Inputting the unstructured data into a feature extraction model to obtain deep features; Inputting the relationship network data into a relationship feature model to obtain relationship features; Inputting the structured data into a label prediction model to obtain a label prediction result; The label prediction result is corrected according to the deep features and the relationship features to obtain a target label result.

2. The batch data labeling method according to claim 1, characterized in that: The step of correcting the label prediction result according to the deep feature and the relationship feature to obtain the target label result specifically includes: Inputting the deep features, the relationship features and the label prediction results into the first fully connected layer for nonlinear transformation; The transformed deep features, relationship features and label prediction results are input into the second fully connected layer for feature fusion to obtain the target feature vector; The target feature vector is input into the output layer to obtain the target label result.

3. The batch data labeling method according to claim 2, characterized in that: The step of inputting the transformed deep features, relationship features and label prediction results into the second fully connected layer to perform feature fusion is specifically implemented in the following way: From fused =FC2([FC1(Z self-supervised ),FC1(Z GNN ),FC1(Z MLP )]) Among them, FC1 represents the operation of the first fully connected layer, FC2 represents the operation of the second fully connected layer, and Z self-supervised is the deep feature, Z GNN is the relationship feature, Z MLP Predict the result for the label.

4. The batch data labeling method according to claim 3, characterized in that: The feature extraction model is a self-supervised learning model, the relationship feature model is a GNN model, and the label prediction model is an MLP model.

5. The batch data labeling method according to claim 4, characterized in that: The activation function of the output layer is Sigmoid.

6. The batch data labeling method according to claim 1, characterized in that: Before the step of obtaining the data to be processed, the method further includes: The data to be processed is preprocessed, and the preprocessing includes missing value filling, outlier processing and normalization processing.

7. The batch data labeling method according to claim 1, characterized in that: After the step of correcting the label prediction result according to the deep features and the relationship features to obtain the target label result, the method further includes: Get label correction data; The target label result is corrected according to the label correction data.

8. A batch data labeling device, characterized in that: include: A data acquisition module, used to acquire data to be processed, wherein the data to be processed includes unstructured data, relationship network data and structured data; A deep feature module, used for inputting the unstructured data into a feature extraction model to obtain deep features; A relationship feature module, used for inputting the relationship network data into a relationship feature model to obtain relationship features; A label prediction module, used to input the structured data into a label prediction model to obtain a label prediction result; The result correction module is used to correct the label prediction result according to the deep feature and the relationship feature to obtain the target label result.

9. 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 batch data labeling method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the batch data labeling method according to any one of claims 1 to 7 is implemented.

Images