A risk detection method and related device for an industrial product

Abstract

The embodiments of the present application provide a risk detection method and related device for industrial products. Through a risk detection model, feature extraction is performed on industrial product data at multiple classification levels to obtain the feature attributes of multiple industrial product data; based on the correlation relationship between the feature attributes of multiple industrial product data, the correlation degree between the feature attributes is determined; according to the correlation degree, risk detection is performed on multiple industrial product data to obtain a detection result. In the embodiments of the present application, the risk detection model includes a bidirectional encoder representation model, an improved text convolutional neural network, and a bidirectional gated recurrent unit network, and the number of feature vectors output by the three is the same. Based on this risk detection model, a large number of features can be extracted. Based on the large number of feature attributes extracted by this risk detection model, risk detection is performed on multiple industrial product data, which can effectively improve the efficiency of risk detection.

Description

Technical Field

[0001] This application relates to the technical field of risk detection, and particularly to a risk detection method and related device for industrial products. Background Art

[0002] With the development of global industrialization, the safety of industrial products has attracted much attention. Detecting the risks of industrial products is an effective means to ensure the safety of industrial products.

[0003] Currently, the method of sampling inspection of industrial products can be used to predict the risks of industrial products. However, due to the wide variety and large quantity of industrial products, the method of detecting the risks of industrial products by sampling inspection has the problem of low detection efficiency. Summary of the Invention

[0004] Based on the above problems, this application provides a risk detection method and related device for industrial products.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] First aspect: The embodiments of this application provide a risk detection method for industrial products, including:

[0007] Based on the risk types of multiple industrial product data, perform label processing on the multiple industrial product data to obtain multiple labeled industrial product data;

[0008] Based on risk indicators, perform grade division on the multiple labeled industrial product data to obtain multiple industrial product data with divided grades;

[0009] Based on a risk detection model, perform feature extraction on the multiple industrial product data with divided grades to obtain characteristic attributes of the multiple industrial product data; the risk detection model includes a Bidirectional Encoder Representations from Transformers (Bert) model, an improved Text Convolutional Neural Network (Text-CNN), and a Bidirectional Gated Recurrent Unit (Bi-Gru) network, and the number of feature vectors output by the Bert model, the improved Text-CNN network, and the Bi-Gru network is the same;

[0010] Based on the association relationship between the characteristic attributes of the multiple industrial product data, determine the degree of association between the characteristic attributes;

[0011] According to the degree of association, perform risk detection on the multiple industrial product data to obtain a detection result.

[0012] In a possible implementation manner, the step of performing grade division on the multiple labeled industrial product data based on risk indicators to obtain multiple industrial product data with divided grades includes:

[0013] For each level of industrial product data among the multiple levels of industrial product data, based on the Bert model, feature extraction is performed on the first indicator parameter of the industrial product data at this level to obtain the feature vector corresponding to the first indicator parameter;

[0014] Based on the Bert model, the bidirectional gated recurrent unit Bi-Gru network, and the improved text convolutional neural network Text-CNN network, feature extraction is performed on the indicator parameters other than the first indicator parameter to obtain the feature vectors corresponding to the other indicator parameters;

[0015] The feature vectors corresponding to the first indicator parameter and the feature vectors corresponding to the other indicator parameters are fused to obtain the feature attributes of multiple industrial product data.

[0016] In a possible implementation manner, the step of performing feature extraction on the indicator parameters other than the first indicator parameter based on the Bert model, the bidirectional gated recurrent unit Bi-Gru network, and the improved text convolutional neural network Text-CNN network to obtain the feature vectors corresponding to the other indicator parameters includes:

[0017] Based on the Bert model, feature extraction is performed on the indicator parameters other than the first indicator parameter to obtain the feature vectors corresponding to the Bert model;

[0018] Based on the Bi-Gru network, feature extraction is performed on the feature vectors corresponding to the Bert model to obtain the feature vectors corresponding to the Bi-Gru network; the feature vectors corresponding to the Bi-Gru network obtained at the first moment are the linear interpolation between the feature vectors corresponding to the Bi-Gru network obtained at the second moment and the candidate feature vectors corresponding to the Bi-Gru network; the second moment is the moment before the first moment;

[0019] Based on the Text-CNN network, feature extraction is performed on the feature vectors corresponding to the Bi-Gru network to obtain the feature vectors corresponding to the other indicator parameters.

[0020] In a possible implementation manner, the step of determining the correlation degree between feature attributes based on the correlation relationship between the feature attributes of the multiple industrial product data includes:

[0021] Weight calculation is performed on the feature attributes of the multiple industrial product data to obtain the weights of each feature attribute;

[0022] Based on the correlation relationship between the feature attributes among the feature attributes of the multiple industrial product data whose weights are greater than or equal to the weight threshold, the correlation degree between the feature attributes is determined.

[0023] In a possible implementation, among the characteristic attributes of the multiple industrial product data, for the association relationship between the characteristic attributes with weights greater than or equal to the weight threshold, determining the association degree between the characteristic attributes includes:

[0024] Regarding the characteristic attributes of the multiple industrial product data with weights greater than or equal to the weight threshold as preferred characteristic attributes;

[0025] Based on the minimum support and confidence, calculating the association degree for the preferred characteristic attributes to determine the association degree between the characteristic attributes.

[0026] In a possible implementation, for determining the association degree between the characteristic attributes based on the association relationship between the characteristic attributes of the multiple industrial product data, it includes:

[0027] Based on the characteristic attributes of the multiple industrial product data, calculating the soft labels of the characteristic attributes of the multiple industrial product data;

[0028] Based on the soft labels of the characteristic attributes of the multiple industrial product data, determining the association degree between the characteristic attributes.

[0029] In a possible implementation, for performing label processing on the multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple labeled industrial product data, it includes:

[0030] Performing preprocessing on the multiple industrial product data to obtain the preprocessed multiple industrial product data;

[0031] Based on the risk types of the multiple industrial product data, performing label processing on the preprocessed multiple industrial product data to obtain multiple labeled industrial product data.

[0032] In a possible implementation, the risk indicators include a first indicator parameter, a second indicator parameter, a third indicator parameter, and a fourth indicator parameter. The first indicator parameter represents the possibility of a risk occurring, the second indicator parameter represents the severity of the risk occurring, the third indicator parameter represents the degree of risk diffusion, and the fourth indicator parameter represents the social impact factor of the risk. For grading the multiple labeled industrial product data based on the risk indicators to obtain multiple graded industrial product data, it includes:

[0033] For each labeled industrial product data among the multiple labeled industrial product data, calculating the risk value of the labeled industrial product data according to the first weight corresponding to the first indicator parameter of the labeled industrial product data, the second weight corresponding to the second indicator parameter, the third weight corresponding to the third indicator parameter, and the fourth weight corresponding to the fourth indicator parameter;

[0034] Classify the tagged industrial product data according to the risk values of multiple tagged industrial product data to obtain multiple classified industrial product data.

[0035] Second aspect: An embodiment of the present application provides a risk detection device for industrial products, including:

[0036] A label processing unit, a classification unit, a feature extraction unit, a determination unit, and a detection unit;

[0037] The label processing unit is configured to perform label processing on the multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple tagged industrial product data;

[0038] The classification unit is configured to classify the multiple tagged industrial product data based on risk indicators to obtain multiple classified industrial product data;

[0039] The feature extraction unit is configured to extract features from the multiple classified industrial product data based on a risk detection model to obtain feature attributes of the multiple industrial product data; the risk detection model includes a Bidirectional Encoder Representations from Transformers (Bert) model, an improved Text Convolutional Neural Network (Text-CNN) network, and a Bidirectional Gated Recurrent Unit (Bi-Gru) network, and the number of feature vectors output by the Bert model, the improved Text-CNN network, and the Bi-Gru network is the same;

[0040] The determination unit is configured to determine the degree of association between the feature attributes based on the association relationship between the feature attributes of the multiple industrial product data;

[0041] The detection unit is configured to perform risk detection on the multiple industrial product data according to the degree of association to obtain a detection result.

[0042] Third aspect: An embodiment of the present application provides a computer device, which includes: a processor and a memory;

[0043] The memory is used to store program code and transmit the program code to the processor;

[0044] The processor is configured to execute the steps of the risk detection method for an industrial product as described above according to the instructions in the program code.

[0045] Compared with the prior art, the present application has the following beneficial effects:

[0046] The embodiment of the present application provides a risk detection method for industrial products. Through a risk detection model, feature extraction can be performed on industrial product data at multiple classification levels to obtain the feature attributes of multiple industrial product data; based on the correlation relationship between the feature attributes of multiple industrial product data, the correlation degree between the feature attributes is determined; according to the correlation degree, risk detection is performed on multiple industrial product data to obtain a detection result. In the embodiment of the present application, the risk detection model includes a bidirectional encoder representation model, an improved text convolutional neural network, and a bidirectional gated recurrent unit network, and the number of output feature vectors of the bidirectional encoder representation model, the improved text convolutional neural network, and the bidirectional gated recurrent unit network is the same. Based on this risk detection model, a large number of features can be extracted, and based on the large number of feature attributes extracted by this risk detection model, risk detection is performed on multiple industrial product data, which can effectively improve the efficiency of risk detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a flowchart of a risk detection method for industrial products provided by an embodiment of the present application;

[0049] Figure 2 It is a schematic diagram of feature fusion provided by an embodiment of the present application;

[0050] Figure 3 It is a framework diagram of an improved Text-CNN network provided by an embodiment of the present application;

[0051] Figure 4 It is a schematic diagram of a risk detection method for industrial products provided by an embodiment of the present application;

[0052] Figure 5 It is a schematic structural diagram of a risk detection device for industrial products provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] Currently, the method of sampling industrial products can be used to predict the risks of industrial products. However, due to the wide variety and large quantity of industrial products, the method of detecting the risks of industrial products by sampling has the problem of low detection efficiency.

[0054] Based on this, the embodiments of the present application provide a risk detection method and related device for industrial products. Through a risk detection model, feature extraction can be performed on industrial product data at multiple classification levels to obtain the feature attributes of multiple industrial product data. Based on the correlation relationship between the feature attributes of multiple industrial product data, the correlation degree between the feature attributes is determined. According to the correlation degree, risk detection is performed on multiple industrial product data to obtain a detection result. In the embodiments of the present application, the risk detection model includes a bidirectional encoder representation model, an improved text convolutional neural network, and a bidirectional gated recurrent unit network. The number of feature vectors output by the bidirectional encoder representation model, the improved text convolutional neural network, and the bidirectional gated recurrent unit network is the same. Based on this risk detection model, a large number of features can be extracted. Based on the large number of feature attributes extracted by this risk detection model, risk detection is performed on multiple industrial product data, which can effectively improve the efficiency of risk detection.

[0055] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0056] See Figure 1 , which is a flowchart of a risk detection method for industrial products provided by an embodiment of the present application, including S101-S105.

[0057] S101. Perform label processing on the multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple labeled industrial product data.

[0058] In the embodiments of the present application, multiple industrial product data can be selected from the industrial product safety data set randomly inspected by the State Administration for Market Regulation. Since some data noise, missing values, and outliers may be included in the selected multiple industrial product data, which will affect the accuracy of risk prediction. Therefore, in the embodiments of the present application, the selected multiple industrial product data can be preprocessed to fill in blank values and process unbalanced data to obtain the preprocessed multiple industrial product data.

[0059] Exemplarily, the process of preprocessing the selected multiple industrial product data includes, but is not limited to, data preprocessing methods such as missing value filling and singular value regression to obtain multiple industrial product data with low noise and high availability.

[0060] For multiple industrial product data, based on their respective corresponding risk types, the preprocessed multiple industrial product data can be labeled to obtain multiple labeled industrial product data. The label information corresponding to each labeled industrial product data can represent the risk type of the industrial product data.

[0061] S102. Based on risk indicators, classify the multiple labeled industrial product data to obtain multiple classified industrial product data.

[0062] Exemplarily, the risk indicators may include a first indicator parameter, a second indicator parameter, a third indicator parameter, and a fourth indicator parameter. Among them, the first indicator parameter represents the likelihood of a risk occurring, the second indicator parameter represents the severity of the risk occurring, the third indicator parameter represents the degree of risk diffusion, and the fourth indicator parameter represents the social impact factor of the risk.

[0063] The first indicator parameter helps to determine which risks are most worthy of attention and what measures need to be taken to reduce the probability of these risks occurring. The higher the value of the first indicator parameter, the higher the likelihood of a risk occurring, and the greater the potential impact of a risk with a higher likelihood. The second indicator parameter helps to determine which risks may have a significant impact on the enterprise and the corresponding emergency response measures need to be taken or a long-term risk management plan needs to be developed. The third indicator parameter helps to determine which risks require more extensive attention and management and what response mechanisms need to be established to reduce the impact of diffusion. The fourth indicator parameter helps the management department to more comprehensively understand the possible consequences of a risk event and take corresponding measures to reduce these impacts.

[0064] In the embodiments of the present application, for each of the multiple labeled industrial product data, the risk value of the labeled industrial product data can be calculated according to the first weight corresponding to the first indicator parameter, the second weight corresponding to the second indicator parameter, the third weight corresponding to the third indicator parameter, and the fourth weight corresponding to the fourth indicator parameter of the labeled industrial product data; based on the risk values of the multiple labeled industrial product data, the labeled industrial product data is classified to obtain multiple classified industrial product data.

[0065] In the embodiments of the present application, the labeled industrial product data can be classified from four dimensions: the likelihood of a risk occurring, the severity of the risk occurring, the degree of risk diffusion, and the social impact factor of the risk.

[0066] Exemplarily, for the labeled industrial product data, the values corresponding to the first index parameter, the second index parameter, the third index parameter, and the fourth index parameter are A, B, C, and D respectively; the first weight corresponding to the first index parameter, the second weight corresponding to the second index parameter, the third weight corresponding to the third index parameter, and the fourth weight corresponding to the fourth index parameter are a, b, c, and d respectively. Then, the risk value of the labeled industrial product data can be expressed as (A×a + B×b + C×c + D×d). Based on the magnitude of the risk value, the labeled industrial product data can be classified into levels to obtain industrial product data at multiple classified levels.

[0067] S103. Based on the risk detection model, extract features from the industrial product data at the multiple classified levels to obtain the feature attributes of the multiple industrial product data.

[0068] Among them, the risk detection model includes a Bidirectional Encoder Representations from Transformers (Bert) model, an improved Text Convolutional Neural Networks (Text-CNN) network, and a Bidirectional Gated Recurrent Unit (Bi-Gru) network.

[0069] In the embodiment of the present application, the number of output feature vectors of the improved Text-CNN network is the same as the number of output feature vectors of the Bert model and the Bi-Gru network. Exemplarily, the number of output feature vectors of the improved Text-CNN network can be adjusted from 1 to 32.

[0070] In a possible implementation manner, for each piece of industrial product data at a classified level among the industrial product data at the multiple classified levels, based on the Bert model, extract features from the first index parameter of the industrial product data at the classified level to obtain the feature vector corresponding to the first index parameter. Based on the Bert model, the Bidirectional Gated Recurrent Unit Bi-Gru network, and the improved Text Convolutional Neural Text-CNN network, extract features from the other index parameters except the first index parameter to obtain the feature vectors corresponding to the other index parameters. As Figure 2 shown, this figure is a schematic diagram of feature fusion provided by the embodiment of the present application.

[0071] In the embodiments of the present application, a Bert model with enhanced semantic information extraction ability is adopted, and the encoder of the Transformer model (Transformer Encoder) is used as a feature extractor to extract features from the first metric parameter, improving the accuracy of feature extraction.

[0072] Based on the Bert model, the bidirectional gated recurrent unit Bi-Gru network, and the improved text convolutional neural network Text-CNN network, features are extracted from other metric parameters other than the first metric parameter to obtain feature vectors corresponding to the other metric parameters.

[0073] Exemplarily, when the first metric parameter represents the possibility of a risk occurring, the other metric parameters may include a second metric parameter representing the severity of the risk occurring, a third metric parameter representing the degree of risk diffusion, and a fourth metric parameter representing the social impact factor of the risk.

[0074] In a possible implementation manner, the step of extracting features from other metric parameters other than the first metric parameter based on the Bert model, the bidirectional gated recurrent unit Bi-Gru network, and the improved text convolutional neural network Text-CNN network to obtain feature vectors corresponding to the other metric parameters includes:

[0075] Extracting features from other metric parameters other than the first metric parameter based on the Bert model to obtain feature vectors corresponding to the Bert model;

[0076] Extracting features from the feature vectors corresponding to the Bert model based on the Bi-Gru network to obtain feature vectors corresponding to the Bi-Gru network; the feature vectors corresponding to the Bi-Gru network obtained at the first moment are a linear interpolation between the feature vectors corresponding to the Bi-Gru network obtained at the second moment and the candidate feature vectors corresponding to the Bi-Gru network; the second moment is the moment before the first moment.

[0077] Exemplarily, the feature vectors corresponding to the Bi-Gru network obtained at the first moment are shown in Equation (1):

[0078]

[0079] where t represents the first moment; h t is the output of the Bi-Gru network at the first moment; h t-1 is the output of the Bi-Gru network at the second moment; represents the candidate output corresponding to the Bi-Gru network; z t is the update gate, representing the degree of content update.

[0080] zt It can be expressed by Equation (2) as follows:

[0081] z t = σ(W z x t + U z h t-1 + b z ) (2)

[0082] It can be expressed by Equation (3) as follows:

[0083]

[0084] where r t is a set of reset gates and can be expressed by Equation (4) as follows:

[0085] r t = σ(W r x t + U r h t-1 + b r ) (4)

[0086] where W z , W h , W r are the weight matrices of the update gate, candidate output, and reset gate at the first moment of the neuron, respectively; U z , U h , U r are the weight matrices of the recurrent input of the update gate, candidate output, and reset gate at the first moment, respectively; b z , b h , b r are the bias vectors of the update gate, candidate output, and reset gate at the first moment, respectively.

[0087] After obtaining the feature vectors corresponding to the Bi-Gru network, feature extraction is performed on the feature vectors corresponding to the Bi-Gru network based on the improved Text-CNN network, and the feature vectors corresponding to the other index parameters can be obtained. As Figure 3 shown, this figure is a framework diagram of an improved Text-CNN network provided by an embodiment of the present application. The loss function of the improved Text-CNN network can be expressed by Equation (5) as follows:

[0088] L 1 (θ, θ 1 , θ 2 ) = L 1 (θ, θ 1 ) + L 1 (θ, θ 2 ) (5)

[0089] Fusing the feature vectors corresponding to the first index parameter and the feature vectors corresponding to the other index parameters can obtain the feature attributes of multiple industrial product data.

[0090] S104. Determine the degree of association between the feature attributes based on the association relationship between the feature attributes of the multiple industrial product data.

[0091] In a possible implementation, weight calculation can be performed on the feature attributes of the multiple industrial product data to obtain the weights of each feature attribute. Based on the association relationship between the feature attributes with weights greater than or equal to the weight threshold among the feature attributes of the multiple industrial product data, determine the degree of association between the feature attributes.

[0092] In the embodiments of the present application, the feature attributes with weights greater than or equal to the weight threshold among the feature attributes of the multiple industrial product data can be used as preferred feature attributes; based on the minimum support and confidence, calculate the degree of association of the preferred feature attributes to determine the degree of association between the feature attributes.

[0093] The influence of different index parameters on the industrial product risk detection process is different. In a possible implementation, based on the feature attributes of the multiple industrial product data, calculate the soft labels of the feature attributes of the multiple industrial product data; based on the soft labels of the feature attributes of the multiple industrial product data, determine the degree of association between the feature attributes.

[0094] Among them, frequent item set association rules can be determined based on the soft labels of the feature attributes of the multiple industrial product data to determine the degree of association between the feature attributes.

[0095] In the embodiments of the present application, risk analysis can be performed based on the frequent item set association rules, sort the data values corresponding to the feature attributes, and obtain the fluctuation range of the data values of multiple feature attributes. This fluctuation range can be expressed as [V min , V max , where V min and V max are respectively the minimum value and the maximum value among the data values of multiple feature attributes.

[0096] S105. Perform risk detection on the multiple industrial product data according to the degree of association to obtain a detection result.

[0097] In the embodiments of the present application, based on a risk detection model constructed by a Bert model, a Bi-Gru network, and an improved Text-CNN network, risk detection can be performed on industrial product data from multiple dimensions, the accuracy of risk prediction can be improved, and based on the detection result, downgrading or risk isolation processing can be performed on it.

[0098] For ease of understanding, the following will combine Figure 4 to provide an overall introduction to a risk detection method for industrial products provided by an embodiment of the present application.

[0099] In an embodiment of the present application, multiple industrial product data can be selected from an industrial product dataset randomly inspected by the State Administration for Market Regulation. Based on a data preprocessing module, the attributes of the multiple industrial product data are determined, and the multiple industrial product data are preprocessed to obtain the preprocessed multiple industrial product data. The preprocessed multiple industrial product data are subjected to label processing for service objects and security types to obtain multiple labeled industrial product data.

[0100] The data preprocessing module sends the multiple labeled industrial product data to a multi-dimensional risk data grading module, and the multi-dimensional risk data grading module can perform risk grading on the multiple labeled industrial product data based on multi-dimensional risk indicators to obtain multiple graded industrial product data.

[0101] The multi-dimensional risk data grading module sends the multiple graded industrial product data to an industrial product risk prediction module. The industrial product risk prediction module includes a feature extraction module, a feature fusion module based on soft labels, and a risk prediction module.

[0102] The feature extraction module can implement feature extraction for the multiple graded industrial product data based on the Bert model, the Bi-Gru network, and an improved Text-CNN network. The feature fusion module based on soft labels performs feature fusion on the extracted feature vectors to obtain fused feature vectors, and based on the fused feature vectors, frequent item set association rules are obtained. Based on the frequent item set association rules, the association degree can be determined.

[0103] The risk prediction module performs risk detection on the multiple industrial product data according to the association degree to obtain a detection result.

[0104] In summary, in an embodiment of the present application, the risk detection model includes a bidirectional encoder representation model, an improved text convolutional neural network, and a bidirectional gated recurrent unit network, and the number of output feature vectors of the bidirectional encoder representation model, the improved text convolutional neural network, and the bidirectional gated recurrent unit network is the same. Based on this risk detection model, a large number of features can be extracted, and based on the large number of feature attributes extracted by this risk detection model, risk detection is performed on multiple industrial product data, which can effectively improve the efficiency of risk detection.

[0105] The present application provides a risk detection device for industrial products. See Figure 5, This figure is a schematic structural diagram of a risk detection device for industrial products provided by an embodiment of the present application. Its specific implementation manner is consistent with the implementation manner and the achieved technical effects described in the embodiment of the above method, and some contents will not be elaborated again.

[0106] A risk detection device 1100 for industrial products includes:

[0107] A label processing unit 1101, a level division unit 1102, a feature extraction unit 1103, a determination unit 1104, and a detection unit 1105;

[0108] The label processing unit 1101 is configured to perform label processing on the multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple labeled industrial product data;

[0109] The level division unit 1102 is configured to perform level division on the multiple labeled industrial product data based on risk indicators to obtain multiple industrial product data with divided levels;

[0110] The feature extraction unit 1103 is configured to perform feature extraction on the multiple industrial product data with divided levels based on a risk detection model to obtain feature attributes of the multiple industrial product data; the risk detection model includes a Bidirectional Encoder Representations from Transformers (Bert) model, an improved Text Convolutional Neural Network (Text-CNN) network, and a Bidirectional Gated Recurrent Unit (Bi-Gru) network, and the number of feature vectors output by the Bert model, the improved Text-CNN network, and the Bi-Gru network is the same;

[0111] The determination unit 1104 is configured to determine the degree of association between the feature attributes based on the association relationship between the feature attributes of the multiple industrial product data;

[0112] The detection unit 1105 is configured to perform risk detection on the multiple industrial product data according to the degree of association to obtain a detection result.

[0113] In a possible implementation manner, the level division unit is specifically configured to:

[0114] For each industrial product data with a divided level among the multiple industrial product data with divided levels, perform feature extraction on the first index parameter of the industrial product data with the divided level based on the Bert model to obtain a feature vector corresponding to the first index parameter;

[0115] Based on the Bert model, the bidirectional gated recurrent unit Bi-Gru network, and the improved text convolutional neural network Text-CNN, feature extraction is performed on other metric parameters except the first metric parameter to obtain the feature vectors corresponding to the other metric parameters;

[0116] The feature vectors corresponding to the first metric parameter and the feature vectors corresponding to the other metric parameters are fused to obtain the feature attributes of multiple industrial product data.

[0117] In a possible implementation manner, the grading unit is specifically configured to:

[0118] Based on the Bert model, feature extraction is performed on other metric parameters except the first metric parameter to obtain the feature vectors corresponding to the Bert model;

[0119] Based on the Bi-Gru network, feature extraction is performed on the feature vectors corresponding to the Bert model to obtain the feature vectors corresponding to the Bi-Gru network; the feature vectors corresponding to the Bi-Gru network obtained at the first moment are the linear interpolation between the feature vectors corresponding to the Bi-Gru network obtained at the second moment and the candidate feature vectors corresponding to the Bi-Gru network; the second moment is the previous moment of the first moment;

[0120] Based on the Text-CNN network, feature extraction is performed on the feature vectors corresponding to the Bi-Gru network to obtain the feature vectors corresponding to the other metric parameters.

[0121] In a possible implementation manner, the determining unit is specifically configured to:

[0122] Calculate the weights of the feature attributes of the multiple industrial product data to obtain the weights of each feature attribute;

[0123] Based on the association relationship between the feature attributes with weights greater than or equal to the weight threshold among the feature attributes of the multiple industrial product data, determine the association degree between the feature attributes.

[0124] In a possible implementation manner, the determining unit is specifically configured to: use the feature attributes with weights greater than or equal to the weight threshold among the feature attributes of the multiple industrial product data as the preferred feature attributes;

[0125] Based on the minimum support and confidence, calculate the association degree of the preferred feature attributes to determine the association degree between the feature attributes.

[0126] In a possible implementation manner, the determining unit is specifically configured to:

[0127] Calculate soft labels for the characteristic attributes of multiple industrial product data based on the characteristic attributes of the multiple industrial product data;

[0128] Determine the degree of association between characteristic attributes based on the soft labels of the characteristic attributes of multiple industrial product data.

[0129] In a possible implementation manner, the label processing unit is specifically configured to:

[0130] Preprocess multiple industrial product data to obtain the preprocessed multiple industrial product data;

[0131] Perform label processing on the preprocessed multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple labeled industrial product data.

[0132] In a possible implementation manner, the risk indicators include a first indicator parameter, a second indicator parameter, a third indicator parameter, and a fourth indicator parameter. The first indicator parameter represents the possibility of a risk occurring, the second indicator parameter represents the severity of the risk occurring, the third indicator parameter represents the degree of risk diffusion, and the fourth indicator parameter represents the social impact factor of the risk. The grading unit is specifically configured to:

[0133] For each labeled industrial product data among the multiple labeled industrial product data, calculate the risk value of the labeled industrial product data according to the first weight corresponding to the first indicator parameter, the second weight corresponding to the second indicator parameter, the third weight corresponding to the third indicator parameter, and the fourth weight corresponding to the fourth indicator parameter of the labeled industrial product data;

[0134] Perform grading on the labeled industrial product data according to the risk values of the multiple labeled industrial product data to obtain multiple graded industrial product data.

[0135] In summary, in the embodiments of the present application, the risk detection model includes a bidirectional encoder representation model, an improved text convolutional neural network, and a bidirectional gated recurrent unit network. The number of output feature vectors of the bidirectional encoder representation model, the improved text convolutional neural network, and the bidirectional gated recurrent unit network is the same. Based on this risk detection model, a large number of features can be extracted. Based on the large number of characteristic attributes extracted by this risk detection model, risk detection is performed on multiple industrial product data, which can effectively improve the efficiency of risk detection.

[0136] The embodiments of the present application provide a computer device, and the computer device includes: a processor and a memory;

[0137] The memory is used to store program code and transmit the program code to the processor;

[0138] The processor is configured to execute the steps of the risk detection method for an industrial product as described above according to the instructions in the program code.

[0139] As described above, this is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A risk detection method for industrial products, characterized in that: include: Based on the risk types of the multiple industrial product data, label the multiple industrial product data to obtain multiple labeled industrial product data; Based on the risk index, the plurality of labeled industrial product data are graded to obtain a plurality of graded industrial product data; For each of the multiple classified industrial product data, extract the first index parameter of the classified industrial product data based on the bidirectional encoder representation method Bert model to obtain a feature vector corresponding to the first index parameter; Based on the Bert model, feature extraction is performed on other indicator parameters other than the first indicator parameter to obtain a feature vector corresponding to the Bert model; Based on the bidirectional gated recurrent unit Bi-Gru network, feature extraction is performed on the feature vector corresponding to the Bert model to obtain a feature vector corresponding to the Bi-Gru network; the feature vector corresponding to the Bi-Gru network obtained at the first moment is a linear interpolation between the feature vector corresponding to the Bi-Gru network obtained at the second moment and the candidate feature vector corresponding to the Bi-Gru network; the second moment is the moment before the first moment; Based on the improved text convolution neural network Text-CNN network, feature extraction is performed on the feature vector corresponding to the Bi-Gru network to obtain the feature vector corresponding to the other indicator parameters; The feature vector corresponding to the first indicator parameter is subjected to feature fusion with the feature vector corresponding to the other indicator parameters to obtain feature attributes of multiple industrial product data; the number of feature vectors output by the Bert model, the improved Text-CNN network, and the Bi-Gru network is the same; Determining the degree of association between the characteristic attributes based on the association relationship between the characteristic attributes of the plurality of industrial product data; According to the correlation degree, risk detection is performed on the plurality of industrial product data to obtain detection results.

2. The method according to claim 1, characterized in that The determining the degree of association between the characteristic attributes based on the association relationship between the characteristic attributes of the plurality of industrial product data includes: Performing weight calculation on the characteristic attributes of the plurality of industrial product data to obtain the weight of each characteristic attribute; Based on the association relationship between the feature attributes whose weights are greater than or equal to the weight threshold among the feature attributes of the plurality of industrial product data, the association degree between the feature attributes is determined.

3. The method according to claim 2, characterized in that The determining the degree of association between the characteristic attributes based on the association relationship between the characteristic attributes whose weights are greater than or equal to the weight threshold among the characteristic attributes of the plurality of industrial product data comprises: Among the characteristic attributes of the plurality of industrial product data, characteristic attributes whose weights are greater than or equal to the weight threshold are taken as preferred characteristic attributes; Based on the minimum support and confidence, the correlation degree is calculated for the preferred feature attributes to determine the correlation degree between the feature attributes.

4. The method according to claim 1, characterized in that: The determining the degree of association between the characteristic attributes based on the association relationship between the characteristic attributes of the plurality of industrial product data includes: Based on the characteristic attributes of the plurality of industrial product data, calculating soft labels of the characteristic attributes of the plurality of industrial product data; Based on soft labels of characteristic attributes of a plurality of industrial product data, the correlation between the characteristic attributes is determined.

5. The method according to claim 1, characterized in that The step of labeling the plurality of industrial product data based on the risk types of the plurality of industrial product data to obtain a plurality of labeled industrial product data includes: Preprocessing a plurality of industrial product data to obtain a plurality of preprocessed industrial product data; Based on the risk types of the plurality of industrial product data, label processing is performed on the plurality of preprocessed industrial product data to obtain a plurality of labeled industrial product data.

6. The method according to claim 1, characterized in that The risk index includes a first index parameter, a second index parameter, a third index parameter and a fourth index parameter, wherein the first index parameter indicates the possibility of risk occurrence, the second index parameter indicates the severity of risk occurrence, the third index parameter indicates the degree of risk diffusion, and the fourth index parameter indicates the social impact factor of risk. Based on the risk index, the plurality of labeled industrial product data are graded to obtain a plurality of graded industrial product data, including: For each labeled industrial product data among the multiple labeled industrial product data, calculating the risk value of the labeled industrial product data according to a first weight corresponding to a first indicator parameter of the labeled industrial product data, a second weight corresponding to a second indicator parameter, a third weight corresponding to a third indicator parameter, and a fourth weight corresponding to a fourth indicator parameter; According to the risk values ​​of a plurality of labeled industrial product data, the labeled industrial product data are graded to obtain a plurality of graded industrial product data.

7. A risk detection device for industrial products, characterized in that: include: A label processing unit, a grading unit, a feature extraction unit, a determination unit, and a detection unit; The label processing unit is used to perform label processing on the multiple industrial product data based on the risk types of the multiple industrial product data to obtain multiple labeled industrial product data; The grading unit is used to grade the plurality of labeled industrial product data based on the risk index to obtain a plurality of graded industrial product data; The feature extraction unit is used to extract features of a first indicator parameter of the industrial product data divided into multiple levels based on the Bert model of the bidirectional encoder representation method for each industrial product data divided into multiple levels, and obtain a feature vector corresponding to the first indicator parameter; extract features of other indicator parameters other than the first indicator parameter based on the Bert model to obtain a feature vector corresponding to the Bert model; extract features of the feature vector corresponding to the Bert model based on the bidirectional gated recurrent unit Bi-Gru network to obtain a feature vector corresponding to the Bi-Gru network; the feature vector corresponding to the Bi-Gru network obtained at the first moment is a linear interpolation between the feature vector corresponding to the Bi-Gru network obtained at the second moment and the candidate feature vector corresponding to the Bi-Gru network; the second moment is the moment before the first moment; extract features of the feature vector corresponding to the Bi-Gru network based on the improved text convolution neural Text-CNN network to obtain feature vectors corresponding to the other indicator parameters; feature fusion is performed on the feature vector corresponding to the first indicator parameter and the feature vector corresponding to the other indicator parameters to obtain feature attributes of multiple industrial product data, and the number of feature vectors output by the Bert model, the improved Text-CNN network and the Bi-Gru network is the same; The determining unit is used to determine the degree of association between the characteristic attributes based on the association relationship between the characteristic attributes of the plurality of industrial product data; The detection unit is used to perform risk detection on the multiple industrial product data according to the correlation degree to obtain a detection result.

8. A computer device, characterized in that: The computer device comprises: a processor and a memory; The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the steps of the risk detection method for industrial products as described in any one of claims 1 to 6 according to the instructions in the program code.

Images