Product risk detection method, apparatus, medium, and program product
By acquiring data during the non-production phase for feature extraction and decision tree matching, the problem of low accuracy in risk monitoring of products in production was solved, resulting in more accurate risk assessment and robust testing results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- INDUSTRIAL AND COMMERCIAL BANK OF CHINA
- Filing Date
- 2025-08-18
- Publication Date
- 2026-06-02
AI Technical Summary
In existing technologies, risk monitoring of products in production relies on human experience or real-time production data, resulting in low monitoring accuracy and difficulty in achieving accurate and stable risk prediction, which affects product quality.
By acquiring data from non-production phases, feature extraction and decision tree construction are performed. Multiple decision trees are used to match features to determine risk detection results. By combining feature weights and information gain, risk level assessment is achieved.
It improves the accuracy and robustness of risk detection, enabling early identification of potential root causes of problems and enhancing the precision and interpretability of product quality control.
Smart Images
Figure CN122134090A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the financial field, artificial intelligence, intelligent operation and maintenance field or other fields, specifically to a product risk detection method, device, medium and program product. Background Technology
[0002] In the financial sector or other fields, quality control of products put into production is related to product reliability, production efficiency, and user satisfaction. Once a quality risk occurs, it can lead to production accidents.
[0003] In implementing the embodiments of this application, it was found that the risk monitoring method for products in production mainly relies on human experience or real-time production data. However, due to the subjectivity of experience and the limitations of data, it is difficult to achieve accurate and stable risk prediction, resulting in low monitoring accuracy and affecting product quality. Summary of the Invention
[0004] In view of the above problems, this application provides a product risk detection method, apparatus, equipment, medium and procedure.
[0005] According to the first aspect of this application, a product risk detection method is provided, comprising: acquiring non-production data associated with the product to be detected generated during the non-production stage; extracting features from different dimensions of the non-production data to obtain features of multiple dimensions; determining a target decision tree from multiple decision trees that matches the features of each dimension of the features of multiple dimensions, wherein each decision tree consists of a root node and leaf nodes, the root node being determined based on historical non-production data of historically produced products, and the leaf nodes being determined based on historical production data of historically produced products; and determining a risk detection result for the product to be detected based on the decision results obtained from making decisions on the features of each dimension.
[0006] According to embodiments of this application, any one of the multiple decision trees is obtained through the following operations: For any type of historically produced product among multiple different types of historically produced products: obtain historical non-production data and historical production data of the historically produced products; extract features from different dimensions of the historical production data to obtain historical production features of multiple dimensions; extract features from different dimensions of the historical non-production data to obtain historical features of multiple dimensions; for any one of the historical features of multiple dimensions: determine the historical feature as the root node; based on the information gain of each of the historical production features of multiple dimensions, recursively determine the leaf nodes from the root node until the recursion stopping condition is met, thus obtaining the decision tree.
[0007] According to an embodiment of this application, the operation further includes: determining a predetermined segmentation threshold from multiple candidate segmentation values for a root node or a leaf node, wherein the multiple candidate segmentation values are determined from historical non-production data or historical production data; and configuring weight values for the root node or leaf node based on the predetermined segmentation threshold.
[0008] According to an embodiment of this application, determining a target decision tree that matches the features of each dimension among multiple decision trees includes: for each dimension among multiple features: determining at least one initial root node that matches the features from the root nodes corresponding to the multiple decision trees; determining a target root node from the at least one initial root node based on the type of historically produced products corresponding to the decision tree corresponding to the at least one initial root node; and determining the decision tree corresponding to the target root node as the target decision tree.
[0009] According to an embodiment of this application, the product risk detection method further includes: obtaining the weight values of the features corresponding to each dimension and a predetermined segmentation threshold using a target decision tree that matches the features of each dimension; and determining the weight values of the features corresponding to each dimension as the decision results of the features of each dimension.
[0010] According to embodiments of this application, the risk detection result includes a risk level; based on the decision results obtained by making decisions on the features of each dimension, the risk detection result for the product to be tested is determined, including: determining the weighted value of features of multiple dimensions based on the weight values and weight coefficients corresponding to the features of each dimension, wherein the weight coefficients corresponding to the features of each dimension are obtained based on the information gain of the historical production features matched with the features of each dimension; and determining the risk level of the product to be tested based on the weighted value and the weight thresholds corresponding to different risk levels.
[0011] According to embodiments of this application, the non-production phase includes the R&D phase and the testing phase; the features across multiple dimensions include at least: R&D scale, testing scale, R&D progress, testing progress, R&D resources, and testing resources.
[0012] The second aspect of this application provides a product risk detection device, comprising: an acquisition module for acquiring non-production data associated with the product to be detected generated during the non-production stage; an extraction module for extracting features from the non-production data from different dimensions to obtain features of multiple dimensions; a matching module for determining a target decision tree from multiple decision trees that matches the features of each dimension of the features of multiple dimensions, wherein each decision tree consists of a root node and leaf nodes, the root node being determined based on historical non-production data of historically produced products, and the leaf nodes being determined based on historical production data of historically produced products; and a determination module for determining the risk detection result for the product to be detected based on the decision results obtained from making decisions on the features of each dimension.
[0013] A third aspect of this application provides an electronic device, comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the product risk detection method described above.
[0014] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions, when executed by a processor, implement the steps of the product risk detection method described above.
[0015] The fifth aspect of this application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps of the product risk detection method.
[0016] This application targets non-production data obtained from the non-production stage, extracts features across multiple dimensions, matches each feature with a corresponding decision tree, and determines the detection result based on the decision results of each decision tree. Since each feature is processed by its matching decision tree, the originally isolated, noisy, and low-value-density non-production data can be transformed into a structured chain of evidence. This not only preserves the interpretability of the decision trees but also reduces the risk of misjudgment from a single decision tree through multi-decision-tree collaboration, thereby improving the accuracy and robustness of risk detection in complex scenarios. Attached Figure Description
[0017] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:
[0018] Figure 1 The illustration shows an application scenario of the product risk detection method, apparatus, device, medium, and program products according to embodiments of this application;
[0019] Figure 2A flowchart illustrating a product risk detection method according to an embodiment of this application is shown schematically.
[0020] Figure 3 A schematic diagram of a decision tree according to an embodiment of this application is shown;
[0021] Figure 4 This illustration schematically shows a target determination decision tree according to an embodiment of the present application;
[0022] Figure 5 This illustration schematically depicts a diagram illustrating the determination of the risk level of a product to be tested according to an embodiment of this application;
[0023] Figure 6 This schematically illustrates a structural block diagram of a product risk detection device according to an embodiment of this application; and
[0024] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a product risk detection method according to an embodiment of this application. Detailed Implementation
[0025] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.
[0026] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.
[0027] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.
[0028] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).
[0029] It should be noted that the product risk detection method and device of this application can be used in the field of fintech, or in any field other than fintech, such as the field of testing, intelligent manufacturing, etc. This application does not limit the application field of the product risk detection method and device.
[0030] In the technical solution of this application, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.
[0031] In scenarios involving automated decision-making using personal information, the methods, devices, and systems provided in this application all offer users corresponding entry points for choosing to agree to or reject the automated decision-making results. If the user chooses to reject, the process proceeds to the expert decision-making stage. Here, "automated decision-making" refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic, health, and credit status through computer programs, and then making a decision. Here, "expert decision-making" refers to the activity of making decisions by personnel who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.
[0032] In implementing the embodiments of this application, it was found that the risk monitoring methods for products in production mainly rely on human experience or real-time production data. However, due to the subjectivity of experience and the limitations of data, it is difficult to achieve accurate and stable risk prediction, resulting in low monitoring accuracy and affecting product quality. For example, for products in the financial sector, when relying on human experience for quality control, the project management office mainly sets inspection indicators, and personnel manually check whether each indicator meets the production requirements. However, the setting of inspection indicators is relatively simple, mainly based on coverage rate and pass rate, i.e., the testing progress and pass status of each content in the test project. This inspection method is too formalized, and relying solely on the subjective feedback of the project management office makes it difficult to discover potential hidden dangers in the project. After actual production, it was found that even if the project management office subjectively reported that the coverage rate and pass rate met the standards, production accidents occurred after production. When relying on real-time production data for quality control, there is a risk that problems cannot be recovered after production.
[0033] In view of this, embodiments of this application provide a product risk detection method, comprising: acquiring non-production data associated with the product to be detected generated during the non-production stage; extracting features from different dimensions of the non-production data to obtain features of multiple dimensions; determining a target decision tree from multiple decision trees that matches the features of each dimension among the features of multiple dimensions, wherein each decision tree consists of a root node and leaf nodes, the root node being determined based on historical non-production data of historically produced products, and the leaf nodes being determined based on historical production data of historically produced products; and determining a risk detection result for the product to be detected based on the decision results obtained from making decisions on the features of each dimension.
[0034] Figure 1 The illustration shows an application scenario of the product risk detection method, apparatus, device, medium, and program product according to embodiments of this application.
[0035] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.
[0036] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).
[0037] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.
[0038] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.
[0039] It should be noted that the product risk detection method provided in this application embodiment can generally be executed by server 105. Correspondingly, the product risk detection device provided in this application embodiment can generally be located in server 105. The product risk detection method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the product risk detection device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.
[0040] It should be understood that Figure 1 The number of first terminal devices, second terminal devices, third terminal devices, networks, and servers shown in the diagram is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.
[0041] The following will be based on Figure 1 The described scene, through Figures 2-5 The product risk detection method according to the embodiments of this application will be described in detail.
[0042] Figure 2 A flowchart illustrating a product risk detection method according to an embodiment of this application is shown.
[0043] like Figure 2 As shown, the product risk detection method in this embodiment includes operations S210 to S240.
[0044] In operation S210, non-production data associated with the product to be tested, generated during the non-production phase, is acquired.
[0045] In operation S220, features are extracted from non-production data from different dimensions to obtain features in multiple dimensions.
[0046] In operation S230, a target decision tree is determined from multiple decision trees that matches the features of each dimension among the features of multiple dimensions.
[0047] In operation S240, based on the decision results obtained from making decisions on the features of each dimension, the risk detection results for the product to be tested are determined.
[0048] In this embodiment, the non-production stage may include, but is not limited to, the R&D stage and the testing stage. The product to be tested may refer to various systems or services that have been developed and are ready to be deployed to the production environment, such as, but not limited to, various applications.
[0049] Non-production data may include, but is not limited to, data related to the research and development of products to be tested, and data related to the testing of products to be tested. Data related to the research and development of products to be tested may include, but is not limited to, research and development scale, research and development cycle, number of researchers, research and development progress, research and development issues, and research and development environment. Data related to the testing of products to be tested may include, but is not limited to, testing scale, number of test cases, testing issues, number of issues, testing and fixing issues, testing cycle, number of testers, and testing environment.
[0050] Dimensions can include, but are not limited to, scale, process, and resource allocation. For example, non-production data can be divided into dimensions to obtain the corresponding non-production data for each dimension. Then, feature extraction can be performed on the non-production data corresponding to each dimension to obtain features for multiple dimensions. Features for multiple dimensions can include, but are not limited to, features based on scale, process, resources, and environment. Features of the scale dimension include test capacity and development scale. Features of the process dimension include case coverage, case pass rate, problem reproduction rate, number of new problems introduced after fixing, and characteristics of development or testing behaviors. Features of the resource dimension include manpower and computing power.
[0051] Any decision tree in a set of multiple decision trees consists of a root node and leaf nodes. The root node is determined based on historical non-production data of previously deployed products, while the leaf nodes are determined based on historical production data of previously deployed products. Previously deployed products can be understood as various systems or services already deployed in a production environment. Historical non-production data may include, but is not limited to, data related to the research and development of previously deployed products, and data related to the testing of previously deployed products. Historical production data may include, but is not limited to, data related to deployment and data on production issues. Data related to deployment includes, for example, deployment time, deployment strategy, and deployment impact.
[0052] For example, decision outcomes can include, but are not limited to, binary classification decisions based on low risk and high risk. If all decision outcomes based on the characteristics across all dimensions are low risk, the risk assessment result for the product under test can be determined as low risk, and production can proceed. If, however, some decision outcomes based on the characteristics across all dimensions contain high risk, the risk assessment result for the product under test can be determined as high risk, requiring manual inspection and repair before production can begin.
[0053] This application targets non-production data obtained from the non-production stage, extracts features across multiple dimensions, matches each feature with a corresponding decision tree, and determines the detection result based on the decision results of each decision tree. Since each feature is processed by its matching decision tree, the originally isolated, noisy, and low-value-density non-production data can be transformed into a structured chain of evidence. This not only preserves the interpretability of the decision trees but also reduces the risk of misjudgment from a single decision tree through multi-decision-tree collaboration, thereby improving the accuracy and robustness of risk detection in complex scenarios.
[0054] According to embodiments of this application, any one of the multiple decision trees is obtained through the following operations: For any type of historically produced product among multiple different types of historically produced products: obtain historical non-production data and historical production data of the historically produced products; extract features from different dimensions of the historical production data to obtain historical production features of multiple dimensions; extract features from different dimensions of the historical non-production data to obtain historical features of multiple dimensions; for any one of the historical features of multiple dimensions: determine the historical feature as the root node; based on the information gain of each of the historical production features of multiple dimensions, recursively determine the leaf nodes from the root node until the recursion stopping condition is met, thus obtaining the decision tree.
[0055] In this application embodiment, different types of historically produced products may include, but are not limited to: application systems, incremental functional modules, data products, components, etc.
[0056] Historical production data can be categorized into different dimensions, such as commissioning and issues. For example, historical production data can be divided according to these dimensions, and then features can be extracted from the resulting data to obtain features for dimensions such as commissioning and issues. Features for the commissioning dimension include commissioning time, commissioning strategy, and commissioning impact. Features for the issues dimension include the number of historical issues and the number of production accidents caused by similar modifications.
[0057] The information gain of each historical production characteristic can be determined based on the following equation (1):
[0058] G(D,A)=H(D)-H(D|A) (1)
[0059] Where G(D,A) represents the information gain of historical production feature A; H(D) represents the information entropy of all historical production feature sets D; and H(D|A) represents the information bar entropy of historical production feature A across all historical production feature sets D.
[0060] The recursion stopping condition can be used to characterize the condition for stopping recursion when the information gain is less than a preset gain threshold.
[0061] This application constructs a decision tree using features corresponding to historical non-production data as root nodes and features corresponding to historical production data as leaf nodes. Since the root node is located at the top of the decision path, any anomalies that might trigger subsequent production problems, such as design flaws, test failures, or parameter deviations, can immediately trigger the splitting rule and capture potential risks. Based on this, and combined with step-by-step verification using experience paths from the production stage, not only can the root causes of problems be identified in advance, but the calibration results from real production feedback can also be integrated, which helps improve the accuracy and interpretability of risk detection for products awaiting production.
[0062] According to embodiments of this application, the operation of constructing a decision tree may further include: determining a predetermined segmentation threshold from multiple candidate segmentation values for a root node or a leaf node; and configuring weight values for the root node or leaf node based on the predetermined segmentation threshold.
[0063] In the embodiments of this application, multiple candidate segmentation values are determined from historical non-production data or historical production data.
[0064] Figure 3 A schematic diagram of a decision tree according to an embodiment of this application is shown.
[0065] For example, for the root node, multiple candidate segmentation values can be determined by statistically analyzing the data corresponding to the root node in historical non-production data or the data corresponding to the leaf nodes in historical production data. Each candidate segmentation value can be iterated over, and the segmentation value with the highest information gain can be determined as the predetermined segmentation threshold. Weight values can be assigned to the root node or leaf node based on the impact of the features corresponding to the root node on risk or the confidence level of the features corresponding to the leaf nodes. Figure 3 As shown, historical features such as R&D scale 301 can be determined as the root node. Historical production features such as the number of historical problems 302 can be determined as leaf nodes, resulting in a decision tree. Regarding R&D scale 301, which can refer to the number of R&D modules, a predetermined segmentation threshold of 200 can be determined. When R&D scale < 200, the corresponding root node weight can be -1; when R&D scale ≥ 200 and the number of historical problems ≥ 10, the corresponding leaf node weight can be 2; and when R&D scale ≥ 200 and the number of historical problems < 10, the corresponding leaf node weight can be 1.
[0066] Since the weight value of the root node can reflect the risk level of the characteristics of the non-production stage, and the weight value of the leaf node can reflect the confidence level of the true results of the historical production stage, the combination of the two can amplify the sensitivity of anomalies in the non-production stage, and can also dynamically calibrate the sensitivity of anomalies in the non-production stage in combination with the historical production stage. Therefore, it can improve the robustness and interpretability of risk detection for products that have not yet been put into production.
[0067] According to embodiments of this application, regarding the above-mentioned... Figure 2 The operation S230 shown, which determines a target decision tree from multiple decision trees that matches the features of each dimension among the multiple dimensions of features, may include the following operations: for each dimension of features among the multiple dimensions of features: determine at least one initial root node that matches the feature from the root nodes corresponding to the multiple decision trees; determine a target root node from the at least one initial root node based on the type of historically produced products corresponding to the decision tree corresponding to the at least one initial root node; and determine the decision tree corresponding to the target root node as the target decision tree.
[0068] Figure 4 A schematic diagram of a target determination decision tree according to an embodiment of this application is shown.
[0069] like Figure 4 As shown, there can be N decision trees, where N is an integer greater than or equal to 2, such as decision tree 411, decision tree 412, ..., decision tree 41N. Decision trees 411, 412, ..., 41N correspond to root nodes 4111, 4122, ..., 41NN, respectively. Taking one feature from multiple dimensions as an example, the root nodes matching this feature can be root nodes 4111 and 4122. Therefore, root nodes 4111 and 4122 can be used as initial root nodes. Based on the type of the decision tree corresponding to the first root node 4111 and the second root node 4122, the root node that is the same as the product type to be detected can be selected from the first root node 4111 and the second root node 4122. For example, the first root node 4111 is selected as the target root node, and the first decision tree 411 is determined as the target decision tree.
[0070] Since the target root node is determined based on historically produced products of the same type as the product to be detected, it is possible to couple prior knowledge of products of the same type with the tree structure, thereby improving the accuracy of risk detection for products not yet in production.
[0071] According to embodiments of this application, the non-production phase may include a research and development phase and a testing phase. The characteristics across multiple dimensions may include at least: research and development scale, testing scale, research and development progress, testing progress, research and development resources, and testing resources.
[0072] For example, for an application, the scale of development can include, but is not limited to, the number of lines of code, the number of requirements, and the number of developers involved. The scale of testing can include, but is not limited to, the number of test cases, the number of automated scripts, and the number of test environments. The development process can include, but is not limited to, iteration completion rate and defect density trends. The testing process can include, but is not limited to, test case execution rate and the number of test rounds. Development resources can include, but are not limited to, personnel time and computing power consumption. Testing resources can include, but are not limited to, test duration and test equipment utilization.
[0073] By constructing a decision tree by these features, it is beneficial to lock the risk threshold of that dimension at the root node, and then calibrate it at the leaf node using the results of historical production stages.
[0074] According to embodiments of this application, the product risk detection method includes, in addition to the above... Figure 2 In addition to the operations S210 to S240 shown, the operation may also include: based on the feature values corresponding to the features of each dimension and a predetermined segmentation threshold, using a target decision tree that matches the features of each dimension to obtain the weight values corresponding to the features of each dimension; and determining the weight values corresponding to the features of each dimension as the decision results corresponding to the features of each dimension.
[0075] In this embodiment of the application, the feature values corresponding to the features of each dimension can be determined based on non-production data.
[0076] The feature values corresponding to each dimension can be compared with a predetermined segmentation threshold, and the comparison results can be matched with the weights in the target decision tree to obtain the weight values corresponding to each dimension.
[0077] By comparing the feature values corresponding to each dimension with the predetermined segmentation threshold, the weight values corresponding to each dimension of the feature can be quickly obtained from the target decision tree, which is beneficial for determining the subsequent detection results.
[0078] According to embodiments of this application, the risk detection result includes a risk level. Regarding the above... Figure 2 The operation S240 shown, based on the decision results obtained by making decisions on the features of each dimension, determines the risk detection result for the product to be tested, and may include the following operations: determining the weighted value of features of multiple dimensions based on the weight values and weight coefficients corresponding to the features of each dimension; and determining the risk level of the product to be tested based on the weighted value and the weight thresholds corresponding to different risk levels.
[0079] In this embodiment, the weight coefficients corresponding to the features of each dimension are obtained based on the information gain of the historical production features that match the features of each dimension. For example, if the information gain of a historical production feature is high, the weight coefficient of the feature that matches that historical production feature is also high.
[0080] Figure 5 Schematically shows a schematic diagram for determining the risk level of a product to be detected according to an embodiment of the present application.
[0081] As Figure 5 shown, the features of multiple dimensions can be, for example, N, the first feature 501, … the Nth feature 50N. For the first feature 501, the first weight value 511 can be determined from the target decision tree according to the comparison result between the feature value corresponding to the first feature 501 and a predetermined segmentation threshold. The product of the first weight value 511 and the first weight coefficient 512 can be calculated to obtain a part of the weighted value corresponding to the first feature 501, and so on until the Nth feature 50N. The product of the Nth weight value 5N1 and the Nth weight coefficient 5N2 is calculated to obtain a part of the weighted value corresponding to the Nth feature 50N. The partial weighted values corresponding to the first feature 501, … the Nth feature 50N are summed to obtain the weighted value of the features of multiple dimensions.
[0082] The weight thresholds corresponding to different risk levels can be different. For example, the risk levels can be divided into a low risk level, a medium risk level, and a high risk level. The low risk level can indicate that after the product to be detected is put into production, it is most likely that no production problems will occur. The medium risk level can indicate that after the product to be detected is put into production, production problems may occur, but the production problems can be repaired in time. The high risk level can indicate that the product to be detected cannot be smoothly put into production, or production accidents may occur after being put into production.
[0083] For example, the low risk level can correspond to the first weight threshold X1, the medium risk level corresponds to the second weight threshold X2, and the high risk level corresponds to the third weight threshold X3, where X1 < X2 < X3. When the weighted value ≤ X1, it is a low risk level; when X1 < weighted value < X2, it is a medium risk level; when the weighted value > X3, it is a high risk level.
[0084] Since the weight coefficients corresponding to the features of each dimension are determined based on the information gain of the historical production features matching the features of each dimension, the contribution degree of the features of different dimensions in risk detection can be reflected. The dimension with a higher information gain obtains a greater weight, suppressing the interference of low-value noise dimensions and achieving higher detection accuracy.
[0085] Based on the above product risk detection method, the present application also provides a product risk detection device. The following will be combined with Figure 6 to describe this device in detail.
[0086] Figure 6 Schematically shows a structural block diagram of a product risk detection device according to an embodiment of the present application.
[0087] As Figure 6 As shown, the product risk detection device 600 of this embodiment includes an acquisition module 610, an extraction module 620, a matching module 630, and a determination module 640.
[0088] The acquisition module 610 is used to acquire non-production data associated with the product to be tested, generated during the non-production phase. In one embodiment, the acquisition module 610 can be used to perform the operation S210 described above, which will not be repeated here.
[0089] The extraction module 620 is used to extract features from non-production data from different dimensions to obtain features in multiple dimensions. In one embodiment, the extraction module 620 can be used to perform the operation S220 described above, which will not be repeated here.
[0090] The matching module 630 is used to determine a target decision tree from multiple decision trees that matches the features of each dimension among the multiple dimensions of features. Each decision tree consists of a root node and leaf nodes. The root node is determined based on historical non-production data of historically produced products, and the leaf nodes are determined based on historical production data of historically produced products. In one embodiment, the matching module 630 can be used to perform the operation S230 described above, which will not be repeated here.
[0091] The determination module 640 is used to determine the risk detection result for the product to be tested based on the decision results obtained from making decisions on the features of each dimension. In one embodiment, the determination module 640 can be used to perform the operation S240 described above, which will not be repeated here.
[0092] According to embodiments of this application, any one of the multiple decision trees is obtained through the following operations: For any type of historically produced product among multiple different types of historically produced products: obtain historical non-production data and historical production data of the historically produced products; extract features from different dimensions of the historical production data to obtain historical production features of multiple dimensions; extract features from different dimensions of the historical non-production data to obtain historical features of multiple dimensions; for any one of the historical features of multiple dimensions: determine the historical feature as the root node; based on the information gain of each of the historical production features of multiple dimensions, recursively determine the leaf nodes from the root node until the recursion stopping condition is met, thus obtaining the decision tree.
[0093] According to an embodiment of this application, the operation further includes: determining a predetermined segmentation threshold from multiple candidate segmentation values for a root node or a leaf node, wherein the multiple candidate segmentation values are determined from historical non-production data or historical production data; and configuring weight values for the root node or leaf node based on the predetermined segmentation threshold.
[0094] According to an embodiment of this application, the matching module 630 includes: an initial root node determination unit, a target root node determination unit, and a target decision tree determination unit. The initial root node determination unit is used to determine at least one initial root node that matches the feature from among the root nodes corresponding to multiple decision trees. The target root node determination unit is used to determine a target root node from among the at least one initial root node based on the type of historically produced products corresponding to the decision trees corresponding to the at least one initial root node. The target decision tree determination unit is used to determine the decision tree corresponding to the target root node as the target decision tree.
[0095] According to an embodiment of this application, the product risk detection device 600 further includes a weight determination module and a result determination module. The weight determination module is used to obtain the weight values corresponding to the features of each dimension based on the feature values corresponding to the features of each dimension and a predetermined segmentation threshold, using a target decision tree that matches the features of each dimension. The result determination module is used to determine the weight values corresponding to the features of each dimension as the decision results corresponding to the features of each dimension.
[0096] According to embodiments of this application, the risk detection result includes a risk level. The determination module 640 includes a weight determination unit and a level determination unit. The weight determination unit is used to determine the weighted value of features across multiple dimensions based on the weight values and weight coefficients corresponding to the features in each dimension. The weight coefficients corresponding to the features in each dimension are obtained based on the information gain of historical production features matched with those features. The level determination unit is used to determine the risk level of the product to be tested based on the weighted values and weight thresholds corresponding to different risk levels.
[0097] According to embodiments of this application, the non-production phase includes a research and development phase and a testing phase. The characteristics across multiple dimensions include at least: research and development scale, testing scale, research and development progress, testing progress, research and development resources, and testing resources.
[0098] According to embodiments of this application, any multiple modules among the acquisition module 610, extraction module 620, matching module 630, and determination module 640 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of this application, at least one of the acquisition module 610, extraction module 620, matching module 630, and determination module 640 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any appropriate combination of any of these three implementation methods. Alternatively, at least one of the acquisition module 610, extraction module 620, matching module 630, and determination module 640 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.
[0099] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a product risk detection method according to an embodiment of this application.
[0100] like Figure 7 As shown, an electronic device 700 according to an embodiment of this application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.
[0101] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 702 and / or RAM 703. It should be noted that programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.
[0102] According to embodiments of this application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the input / output (I / O) interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.
[0103] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.
[0104] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.
[0105] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this application.
[0106] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.
[0107] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 709, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.
[0108] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.
[0109] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages include, but are not limited to, languages such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).
[0110] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0111] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.
Claims
1. A product risk detection method, characterized in that, The method includes: Acquire non-production data generated during non-production phases that are associated with the product to be tested; Feature extraction is performed on the non-production data from different dimensions to obtain features in multiple dimensions; A target decision tree is determined from multiple decision trees that matches the features of each dimension among the multiple dimensions. Each decision tree consists of a root node and leaf nodes. The root node is determined based on the historical non-production data of the historical products that have been put into production, and the leaf nodes are determined based on the historical production data of the historical products that have been put into production. Based on the decision results obtained from making decisions on the features of each dimension, the risk detection result for the product to be tested is determined.
2. The method according to claim 1, characterized in that, Any of the multiple decision trees is obtained through the following operations: For any type of historically produced product among multiple different types of historically produced products: Obtain the historical non-production data and the historical production data of the historically put-in-production products; Feature extraction is performed from different dimensions of the historical production data to obtain historical production features in multiple dimensions. Feature extraction is performed from different dimensions of the historical non-production data to obtain historical features in multiple dimensions; For any one of the historical features among the multiple dimensions: The historical features are determined as the root node; Based on the information gain of the historical production features of the multiple dimensions, the leaf nodes are recursively determined from the root node until the recursion stopping condition is met, thus obtaining the decision tree.
3. The method according to claim 2, characterized in that, The operation also includes: For the root node or the leaf node, a predetermined segmentation threshold is determined from a plurality of candidate segmentation values, wherein the plurality of candidate segmentation values are determined from the historical non-production data or the historical production data; Based on the predetermined segmentation threshold, a weight value is configured for the root node or the leaf node.
4. The method according to claim 2, characterized in that, The step of determining a target decision tree that matches the features of each dimension among the multiple dimensions of features from multiple decision trees includes: For each of the multiple dimensions of features: Determine at least one initial root node that matches the feature from the root nodes corresponding to multiple decision trees; Based on the type of the historically produced products corresponding to the decision tree corresponding to the at least one initial root node, a target root node is determined from the at least one initial root node; The decision tree corresponding to the target root node is determined as the target decision tree.
5. The method according to claim 3, characterized in that, The method further includes: Based on the feature values corresponding to the features of each dimension and the predetermined segmentation threshold, the weight values corresponding to the features of each dimension are obtained by using the target decision tree that matches the features of each dimension. The weight values corresponding to the features of each dimension are determined as the decision results corresponding to the features of each dimension.
6. The method according to claim 5, characterized in that, The risk detection results include the risk level; The decision-making results obtained based on the features of each dimension determine the risk detection results for the product to be tested, including: Based on the weight values and weight coefficients corresponding to the features of each dimension, the weighted values of the features of the multiple dimensions are determined. The weight coefficients corresponding to the features of each dimension are obtained based on the information gain of the historical production features that match the features of each dimension. The risk level of the product to be tested is determined based on the weighted value and the weight threshold corresponding to different risk levels.
7. The method according to any one of claims 1 to 6, characterized in that, The non-production phase includes the R&D phase and the testing phase; the features of the multiple dimensions include at least: R&D scale, testing scale, R&D process, testing process, R&D resources, and testing resources.
8. A product risk detection device, characterized in that, The device includes: The acquisition module is used to acquire non-production data associated with the product to be tested, generated during the non-production phase. The extraction module is used to extract features from the non-production data from different dimensions to obtain features in multiple dimensions. A matching module is used to determine a target decision tree that matches the features of each dimension among the multiple dimensions of features from multiple decision trees. Each decision tree consists of a root node and leaf nodes. The root node is determined based on the historical non-production data of the historical products that have been put into production, and the leaf nodes are determined based on the historical production data of the historical products that have been put into production. The determination module is used to determine the risk detection result for the product to be tested based on the decision results obtained from making decisions on the features of the various dimensions.
9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.
10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.