Industrial portrait analysis method, system and cloud platform combined with enterprise chain
By performing data feature replacement hidden and non-related feature loading in the enterprise hanging chain, and combining with the industrial portrait data restoration model for analysis, the portrait interference problem is solved, the accuracy and credibility of industrial portrait analysis is improved, and better Restore effect and data quality.
Patent Information
- Application Number
- CN202111495098.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-09
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-12-09
AI Technical Summary
In the enterprise chain, portrait analysis may have portrait interference problems, resulting in hidden and redundant data feature replacement, affecting the accuracy and credibility of portrait analysis.
By obtaining the complete source data of reference industry portraits, data feature replacement hidden and non-related features are loaded, reference redundant industrial portrait source data is generated, and industrial portrait data restoration model is used for analysis, update weight information, and form a target industrial portrait that completes training. The data restoration model is used to analyze industrial portraits based on the enterprise hanging chain for restoring industrial portrait source data.
The quality and completeness of the restoration visualization of industrial portrait source data has been improved, and the accuracy and credibility of the industrial portrait source data to be restored are improved, thereby improving the accuracy and display quality of the industrial portrait source data.
Smart Images

Figure CN114298737B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image analysis technology, and more specifically, to an industry image analysis method, system and cloud platform combined with enterprise chain. Background Art
[0002] A portrait refers to a labeled user model that is abstracted based on user attributes, user preferences, living habits, user behaviors, and other information. By labeling, some highly generalized and easy-to-understand features can be used to describe users, which makes it easier for people to understand users and facilitates computer processing. When portrait analysis is specifically applied to enterprise chain management, there may be problems such as portrait interference. Therefore, a portrait analysis method is urgently needed to solve the above technical problems. Summary of the invention
[0003] In view of this, the present application provides an industry portrait analysis method, system and cloud platform combined with enterprise chain.
[0004] In a first aspect, an industry portrait analysis method combined with an enterprise chain is provided, which is applied to an industry portrait analysis system, and the method at least includes:
[0005] Obtain complete source data of a reference industry portrait, and replace and hide data features of the complete source data of the reference industry portrait, to obtain missing source data of a reference industry portrait related to the complete source data of the reference industry portrait, wherein the missing source data of the reference industry portrait is industry portrait source data with data features replaced and hidden;
[0006] And, non-related features are loaded on the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait, wherein the reference redundant industry portrait source data is industry portrait source data with data feature replacement, hidden and redundant data;
[0007] The reference redundant industry portrait source data is passed to the first industry portrait data restoration model, and the first industry portrait data restoration model performs industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain, to obtain reference restored industry portrait source data related to the reference redundant industry portrait source data;
[0008] Determine the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data, update the weight information of the first industry portrait data restoration model according to the target data cost value, and obtain the second industry portrait data restoration model;
[0009] If it is determined that the second industry portrait data restoration model has matched the training termination requirement, the second industry portrait data restoration model is used as the target industry portrait data restoration model for completing the training;
[0010] Otherwise, the second industry portrait data restoration model is used as the first industry portrait data restoration model, and the step of passing the reference redundant industry portrait source data to the first industry portrait data restoration model is returned to execute; wherein, the target industry portrait data restoration model is used to perform industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain.
[0011] In a possible implementation example, the step of replacing and hiding the data features of the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait related to the complete source data of the reference industry portrait includes:
[0012] The complete source data of the reference industry portrait is feature processed by replacing the hidden template with data features set in advance to obtain the missing source data of the reference industry portrait; wherein, the hidden template with data features set in advance includes several feature processing quantization results, and the several feature processing quantization results match the feature processing rules set in advance.
[0013] In a possible implementation example, the step of performing non-related feature loading on the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait includes:
[0014] Non-relevant features are loaded for the missing source data of the reference industry portrait to obtain the reference redundant industry portrait source data.
[0015] In a possible implementation example, the first industry portrait data restoration model includes a first portrait data feature screening unit, a feature dimension reduction structure, a feature mapping structure, a feature dimension increase structure, a feature fusion structure, and a second portrait data feature screening unit;
[0016] The step of transferring the reference redundant industry portrait source data to the first industry portrait data restoration model, and using the first industry portrait data restoration model to perform industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain, to obtain reference restored industry portrait source data related to the reference redundant industry portrait source data, includes:
[0017] The reference redundant industry portrait source data is transmitted to a first portrait data feature screening unit, and the first portrait data feature screening unit performs portrait data feature screening on the reference redundant industry portrait source data to obtain a first portrait data feature;
[0018] The first portrait data feature is passed to a feature dimension reduction structure and a feature fusion structure, and the feature dimension reduction structure performs feature dimension reduction on the first portrait data feature to obtain a second portrait data feature;
[0019] The second portrait data feature is transferred to a feature mapping structure, and the feature mapping structure performs feature mapping processing on the second portrait data feature to obtain a third portrait data feature;
[0020] The third portrait data feature is transmitted to a feature dimension upgrading structure, and the feature dimension upgrading structure performs feature dimension upgrading on the third portrait data feature to obtain a fourth portrait data feature;
[0021] The fourth portrait data feature is transmitted to a feature fusion structure, and the feature fusion structure performs superposition processing on the fourth portrait data feature and the first portrait data feature to obtain a fifth portrait data feature;
[0022] The fifth portrait data feature is transmitted to the second portrait data feature screening unit, so that the second portrait data feature screening unit performs portrait data feature screening on the fifth portrait data feature to obtain the reference restored industrial portrait source data.
[0023] In a possible implementation example, the number of the feature dimension reduction structures is several, the number of the feature dimension upgrade structures is several, the number of the feature fusion structures is several, and the number of the feature mapping structures is several;
[0024] The number of the feature dimension-raising structures is the same as the number of the feature dimension-reducing structures, and the number of the feature fusion structures is the same as the number of the feature dimension-raising structures;
[0025] For each feature dimensionality reduction structure, after the image data feature passes through the feature dimensionality reduction structure, the number of feature categories of the image data feature increases, the feature vertical range of the image data feature decreases, and the feature horizontal range of the image data feature decreases;
[0026] For each feature dimensionality-upgrading structure, after the portrait data features pass through the feature dimensionality-upgrading structure, the number of feature categories of the portrait data features is reduced, the vertical range of the features of the portrait data features is expanded, and the horizontal range of the features of the portrait data features is expanded.
[0027] In a possible implementation example, the calculating of the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data includes:
[0028] Calculate the structured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data;
[0029] Calculate the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data;
[0030] Calculate the frequent item data cost between the reference restored industry portrait source data and the reference industry portrait complete source data;
[0031] Based on the structured data cost value, the unstructured data cost value and the frequent item data cost value, the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data is calculated.
[0032] In a possible implementation example, the calculating of the structured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data includes:
[0033] Calculate the mean square value based on the knowledge member data variable values of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable values of each knowledge entity in the reference industry portrait complete source data;
[0034] Calculating the structured data cost value according to the mean square cost value and the industry portrait source data;
[0035] The calculating of the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data includes:
[0036] According to the knowledge member data variable values of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable values of each knowledge entity in the reference industry portrait complete source data, calculate the first average value and the first standard deviation value corresponding to the reference restored industry portrait source data, the second average value and the second standard deviation value corresponding to the reference industry portrait complete source data, and the average standard deviation value between the reference restored industry portrait source data and the reference industry portrait complete source data;
[0037] Calculating the unstructured data cost value according to the first average value, the second average value, the first standard deviation value, the second standard deviation value, and the average standard deviation value;
[0038] The calculating of the frequent item data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data includes:
[0039] The reference restored industry portrait source data is transmitted to a variable mining unit to obtain a first data variable value corresponding to the reference restored industry portrait source data;
[0040] The complete source data of the reference industry portrait is transferred to the variable mining unit to obtain the second data variable value corresponding to the complete source data of the reference industry portrait;
[0041] The frequent item data cost value is calculated according to the first data variable value and the second data variable value.
[0042] In a possible implementation example, after using the second industry portrait data restoration model as the target industry portrait data restoration model that has completed training, the method further includes:
[0043] Obtaining the industry portrait source data to be restored, wherein the industry portrait source data to be restored is the industry portrait source data with data feature replacement, hiding, and redundancy;
[0044] The industry portrait source data to be restored is passed to the target industry portrait data restoration model, and the target industry portrait data restoration model performs industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain to obtain complete industry portrait source data related to the industry portrait source data to be restored.
[0045] On the second aspect, an industry portrait analysis system combined with enterprise chain is provided, including a processor and a memory that communicate with each other, the processor is used to call a computer program from the memory, and implement the above method by running the computer program.
[0046] According to a third aspect, a cloud platform is provided, comprising a processor and a memory communicating with each other, wherein the processor is used to retrieve a computer program from the memory and implement the method described in any one of claims 1 to 8 by running the computer program.
[0047] The industrial portrait analysis method, system and cloud platform combined with the enterprise chain provided in the embodiments of the present application can replace and hide data features and load non-relevant features for the reference industrial portrait complete source data to obtain reference redundant industrial portrait source data, that is, industrial portrait source data with data feature replacement, hiding and redundancy. The first industrial portrait data restoration model is trained based on the reference industrial portrait complete source data and the reference redundant industrial portrait source data to obtain the trained target industrial portrait data restoration model, and the industrial portrait analysis combined with the enterprise chain is performed on the industrial portrait source data to be restored based on the target industrial portrait data restoration model, which has good restoration visualization quality and completeness, can remove data feature replacement and hiding and interference for the industrial portrait source data to be restored, improve the precision and credibility of the industrial portrait source data to be restored, thereby improving the accuracy of the industrial portrait source data, thereby improving the visualization quality and completeness of the industrial portrait source data display. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0049] Figure 1 A flow chart of an industry portrait analysis method combined with enterprise chain provided in an embodiment of the present application.
[0050] Figure 2 A block diagram of an industry portrait analysis device combined with an enterprise chain provided in an embodiment of the present application.
[0051] Figure 3 An architectural diagram of an industry portrait analysis system combined with enterprise chain provided in an embodiment of the present application. DETAILED DESCRIPTION
[0052] In order to better understand the above technical scheme, the technical scheme of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, rather than limitations on the technical scheme of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.
[0053] See also Figure 1 , showing an industry portrait analysis method combined with enterprise chain, the method may include the technical solutions described in the following steps 201-207.
[0054] Step 201, obtain the complete source data of the reference industry portrait, and replace and hide the data features of the complete source data of the reference industry portrait, to obtain the reference industry portrait missing source data related to the complete source data of the reference industry portrait. For example, the complete source data of the reference industry portrait can be the industry portrait source data without redundancy and data feature replacement and hiding, and the missing source data of the reference industry portrait can be the industry portrait source data with data feature replacement and hiding.
[0055] For example, several complete source data of reference industry portraits can be obtained. The processing method of each complete source data of reference industry portraits is the same. Take a complete source data of reference industry portrait as an example for introduction. After obtaining the complete source data of the reference industry portrait, the data features of the complete source data of the reference industry portrait can be replaced and hidden to obtain the missing source data of the reference industry portrait. For example, the following steps can be used to replace and hide the data features of the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait.
[0056] (1) Using a pre-set data feature to replace a hidden template to generate missing source data of a reference industry portrait related to the complete source data of the reference industry portrait. For example, by using a pre-set data feature to replace a hidden template to perform feature processing on the complete source data of the reference industry portrait, the missing source data of the reference industry portrait is obtained.
[0057] In (1), it is necessary to first obtain a pre-set data feature replacement hidden template, wherein the pre-set data feature replacement hidden template includes a plurality of feature processing quantization results, and the plurality of feature processing quantization results match a pre-set feature processing rule.
[0058] In (1), after obtaining the pre-set data feature replacement hidden template, the pre-set data feature replacement hidden template can be used to perform feature processing on the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait.
[0059] (2) Using real-time data features to replace hidden sets to generate missing source data of the reference industry portrait related to the complete source data of the reference industry portrait. For example, by replacing hidden sets with real-time data features, feature processing is performed on the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait.
[0060] In (2), after obtaining the real-time data feature replacement hidden set, the real-time data feature replacement hidden set can be used to perform feature processing on the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait. For example, feature processing is performed on the real-time data feature replacement hidden set and the complete source data of the reference industry portrait. Furthermore, by referring to all the local contents of the complete source data of the reference industry portrait, the data feature replacement hidden results of all knowledge entities can be obtained, and the data feature replacement hidden results of all knowledge entities are used to form the missing source data of the reference industry portrait.
[0061] (3) Using an AI model to generate missing source data of the reference industry portrait related to the complete source data of the reference industry portrait, for example, using a pre-set AI model to perform feature dimensionality reduction on the complete source data of the reference industry portrait, and using a pre-set AI model to perform feature dimensionality increase on the complete source data of the reference industry portrait after the feature dimensionality reduction, so as to obtain the missing source data of the reference industry portrait.
[0062] For example, the AI model can be used to first perform feature dimensionality reduction on the complete source data of the reference industry portrait to obtain the complete source data of the reference industry portrait after feature dimensionality reduction, and the positioning of the complete source data of the reference industry portrait after feature dimensionality reduction is less than the positioning of the complete source data of the reference industry portrait. The AI model can be used to perform feature dimensionality increase on the complete source data of the reference industry portrait after feature dimensionality reduction to obtain the missing source data of the reference industry portrait, and the positioning of the missing source data of the reference industry portrait is equal to the positioning of the complete source data of the reference industry portrait.
[0063] For example, the AI model can be used to reduce the feature dimension of the complete source data of the reference industry portrait to obtain the complete source data of the reference industry portrait after the feature dimension reduction. Then, the AI model can be used to increase the feature dimension of the complete source data of the reference industry portrait after the feature dimension reduction to obtain the missing source data of the reference industry portrait.
[0064] Step 202, non-related features are loaded on the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait. The reference redundant industry portrait source data may be industry portrait source data with data feature replacement, hidden and redundant data.
[0065] For example, after obtaining the missing source data of the reference industry portrait, the following steps can be used to load non-relevant features of the missing source data of the reference industry portrait to obtain the reference redundant industry portrait source data (the reference redundant industry portrait source data can also be understood as the reference data feature replacing the hidden redundant industry portrait source data).
[0066] Load non-relevant features for the missing source data of the reference industry portrait to obtain the reference redundant industry portrait source data. The method for determining the non-relevant features may include the following: determine the non-relevant features based on the trained redundant standard deviation value.
[0067] For example, after obtaining the missing source data of the reference industry portrait, in order to improve the restoration effect of the industry portrait data restoration model, and considering that the industry portrait source data is often redundant, the reference industry portrait missing source data is loaded with non-relevant features to obtain the reference redundant industry portrait source data. The non-relevant feature can be: It can be understood that the non-relevant feature is loaded on the knowledge member data variable value of each knowledge entity in the reference industry portrait missing source data to obtain the reference redundant industry portrait source data.
[0068] Step 203, passing the reference redundant industry portrait source data to the first industry portrait data restoration model, and the first industry portrait data restoration model performs an industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain, to obtain reference restored industry portrait source data related to the reference redundant industry portrait source data.
[0069] In another possible implementation, a first industry portrait data restoration model may be set in advance, and the first industry portrait data restoration model may be an industry portrait data restoration model based on a neural network model.
[0070] For example, the first industry portrait data restoration model may include the following contents: a first portrait data feature screening unit, a feature dimension reduction structure, a feature mapping structure, a feature dimension increase structure, a feature fusion structure, and a second portrait data feature screening unit. In step 203, the reference redundant industry portrait source data may be passed to the first portrait data feature screening unit, and the first portrait data feature screening unit performs portrait data feature screening on the reference redundant industry portrait source data to obtain the first portrait data feature. The first portrait data feature is passed to the feature dimension reduction structure and the feature fusion structure, and the feature dimension reduction structure performs feature dimension reduction on the first portrait data feature to obtain the second portrait data feature. The second portrait data feature is passed to the feature mapping structure, and the feature mapping structure performs feature mapping processing on the second portrait data feature to obtain the third portrait data feature. The third portrait data feature is passed to the feature dimension increase structure, and the feature dimension increase structure performs feature dimension increase on the third portrait data feature to obtain the fourth portrait data feature. The fourth portrait data feature is passed to the feature fusion structure, and the feature fusion structure performs superposition processing on the fourth portrait data feature and the first portrait data feature to obtain the fifth portrait data feature. The fifth portrait data feature is transmitted to the second portrait data feature screening unit, so that the second portrait data feature screening unit performs portrait data feature screening on the fifth portrait data feature to obtain reference restoration of industrial portrait source data.
[0071] Step 204, determine the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data, and update the weight information of the first industry portrait data restoration model according to the target data cost value to obtain the second industry portrait data restoration model.
[0072] For example, after obtaining the target data cost value, the weight information of the first industry portrait data restoration model can be updated according to the target data cost value. For example, the weight information of the first industry portrait data restoration model is updated. There is no one-to-one limitation on this updating method, as long as the weight information of the first industry portrait data restoration model can be updated according to the target data cost value.
[0073] When the weight information of the first industry portrait data restoration model is updated, the preferred goal of updating the quantization result is to reduce the cost of the target data when step 203 is executed again according to the second industry portrait data restoration model.
[0074] In another possible implementation, based on the reference restored industry portrait source data and the reference industry portrait complete source data, the following steps can be used to determine the target data cost between the reference restored industry portrait source data and the reference industry portrait complete source data.
[0075] Step 2041, determine the structured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data. For example, determine the mean square cost based on the knowledge member data variable value of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable value of each knowledge entity in the reference industry portrait complete source data, and determine the structured data cost based on the mean square cost and the industry portrait source data.
[0076] Step 2042, determine the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data.
[0077] For example, based on the knowledge member data variable values of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable values of each knowledge entity in the reference restored industry portrait complete source data, determine the first average value (i.e., the average value of the knowledge member data variable values of all knowledge entities in the reference restored industry portrait source data) and the first standard deviation value (i.e., the standard deviation value of the knowledge member data variable values of all knowledge entities in the reference restored industry portrait source data), the second average value (i.e., the average value of the knowledge member data variable values of all knowledge entities in the reference restored industry portrait complete source data) and the second standard deviation value (i.e., the standard deviation value of the knowledge member data variable values of all knowledge entities in the reference restored industry portrait complete source data), and the average standard deviation value between the reference restored industry portrait source data and the reference restored industry portrait complete source data (i.e., the average standard deviation value between the knowledge member data variable values of all knowledge entities in the reference restored industry portrait source data and the knowledge member data variable values of all knowledge entities in the reference restored industry portrait complete source data). Determine the unstructured data cost based on the first average value, the second average value, the first standard deviation value, the second standard deviation value, and the average standard deviation value.
[0078] Step 2043, determine the frequent item data substitution value between the reference restored industry portrait source data and the reference industry portrait complete source data.
[0079] For example, the reference restored industry portrait source data is passed to the variable mining unit to obtain the first data variable value corresponding to the reference restored industry portrait source data, the reference industry portrait complete source data is passed to the variable mining unit to obtain the second data variable value corresponding to the reference industry portrait complete source data, and the frequent item data value is determined based on the first data variable value and the second data variable value.
[0080] For example, a variable mining unit may be set up in advance, and this variable mining unit is used to screen the data variable values of the industrial portrait source data (i.e., the portrait data variable values). According to the above content, the reference restored industrial portrait source data may be passed to the variable mining unit, and the variable mining unit screens the data variable values of the reference restored industrial portrait source data, which are subsequently recorded as the first data variable values. The reference industrial portrait complete source data is passed to the variable mining unit, and the variable mining unit screens the data variable values of the reference industrial portrait complete source data, which are subsequently recorded as the second data variable values.
[0081] Step 2044, based on the structured data cost value, the unstructured data cost value and the frequent item data cost value, determine the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data.
[0082] For example, determine the structured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data, and determine the target data cost based on the structured data cost. Alternatively, determine the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data, and determine the target data cost based on the unstructured data cost. Alternatively, determine the frequent item data cost between the reference restored industry portrait source data and the reference industry portrait complete source data, and determine the target data cost based on the frequent item data cost. Alternatively, determine the structured data cost and the unstructured data cost, and determine the target data cost based on the structured data cost and the unstructured data cost. Alternatively, determine the structured data cost and the frequent item data cost, and determine the target data cost based on the structured data cost and the frequent item data cost. Alternatively, determine the unstructured data cost and the frequent item data cost, and determine the target data cost based on the unstructured data cost and the frequent item data cost.
[0083] Step 205, determine whether the secondary industry portrait data restoration model has matched the training termination requirements.
[0084] If the determination is yes, step 206 may be executed, and if the determination is no, step 207 may be executed.
[0085] In another possible implementation, if it is determined that the target data cost value is less than a pre-set judgment value, it is determined that the second industry portrait data restoration model has matched the training termination requirement. If it is determined that the target data cost value is greater than or equal to the pre-set judgment value, it is determined that the second industry portrait data restoration model does not match the training termination requirement.
[0086] In another possible implementation, if it is determined that the optimization quantity of the first industry portrait data restoration model reaches the quantity determination value, then it is determined that the second industry portrait data restoration model has matched the training termination requirements; if it is determined that the optimization quantity of the first industry portrait data restoration model does not reach the quantity determination value, then it is determined that the second industry portrait data restoration model does not match the training termination requirements.
[0087] For example, after each optimization based on the primary industry portrait data restoration model, the optimization quantity of the primary industry portrait data restoration model is superimposed by 2, and it is determined whether the optimization quantity reaches the quantity determination value, so as to know whether the secondary industry portrait data restoration model matches the training termination requirements.
[0088] Furthermore, the above descriptions are only two examples of determining whether the secondary industry portrait data restoration model matches the training termination requirements. For example, if it is determined that the total optimization time of the primary industry portrait data restoration model has reached the time determination value, it can also be determined that the secondary industry portrait data restoration model has matched the training termination requirements.
[0089] Step 206, the second industry portrait data restoration model is used as the target industry portrait data restoration model that has completed training. At this point, the training process of the first industry portrait data restoration model is completed, and the target industry portrait data restoration model that has completed training is obtained. In another possible implementation, the target industry portrait data restoration model can be used to perform industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain.
[0090] Step 207, use the second industry portrait data restoration model as the first industry portrait data restoration model, return to execute the step of passing the reference redundant industry portrait source data to the first industry portrait data restoration model (i.e., the second industry portrait data restoration model), that is, return to step 203-step 204, update the weight information of the first industry portrait data restoration model again, and obtain the second industry portrait data restoration model.
[0091] In another possible implementation, based on the trained target industry portrait data restoration model and the identification steps of the target industry portrait data restoration model, in this embodiment, an industry portrait analysis method combined with enterprise chain is proposed. The industry portrait analysis method combined with enterprise chain may include the following contents.
[0092] Step 601, obtaining the industry portrait source data to be restored, which may be industry portrait source data with data feature replacement, hiding, and redundancy.
[0093] Step 602, transfer the industry portrait source data to be restored to the target industry portrait data restoration model, and the target industry portrait data restoration model performs industry portrait analysis on the industry portrait source data in combination with the enterprise chain to obtain the complete industry portrait source data related to the industry portrait source data to be restored.
[0094] For example, the structure of the target industry portrait data restoration model is consistent with the structure of the first industry portrait data restoration model. In this embodiment, the reference industry portrait complete source data can be subjected to data feature replacement hiding and non-relevant feature loading to obtain reference redundant industry portrait source data, that is, industry portrait source data with data feature replacement hiding and redundancy, and the first industry portrait data restoration model is trained based on the reference industry portrait complete source data and the reference redundant industry portrait source data to obtain the trained target industry portrait data restoration model, and the target industry portrait data restoration model is used to perform industry portrait analysis combined with the enterprise chain on the to-be-restored industry portrait source data, which has good restoration visualization quality and integrity, can perform data feature replacement hiding and interference removal on the to-be-restored industry portrait source data, and improve the accuracy and credibility of the to-be-restored industry portrait source data, thereby improving the accuracy of the industry portrait source data, and thus improving the visualization quality and integrity of the industry portrait source data display. The real-time data features can be filtered out to replace the hidden set based on the missing source data of the displayed content industry portrait, and then the reference industry portrait missing source data and the reference redundant industry portrait source data (the transmission to the first industry portrait data restoration model) can be obtained based on the reference industry portrait complete source data (the identification of the first industry portrait data restoration model).
[0095] Based on the above, please refer to Figure 2 , provides an industry portrait analysis device 200 combined with an enterprise chain, which is applied to an industry portrait analysis system combined with an enterprise chain, and the device includes:
[0096] The data replacement module 210 is used to obtain the complete source data of the reference industry portrait, and replace and hide the data features of the complete source data of the reference industry portrait, so as to obtain the missing source data of the reference industry portrait related to the complete source data of the reference industry portrait, wherein the missing source data of the reference industry portrait is the industry portrait source data with data features replaced and hidden;
[0097] The data acquisition module 220 is used to load non-related features on the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait, wherein the reference redundant industry portrait source data is industry portrait source data with data feature replacement, hidden and redundant data;
[0098] The image acquisition module 230 is used to transfer the reference redundant industry portrait source data to the first industry portrait data restoration model, and the first industry portrait data restoration model performs industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain to obtain reference restored industry portrait source data related to the reference redundant industry portrait source data;
[0099] The model acquisition module 240 is used to determine the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data, and update the weight information of the first industry portrait data restoration model according to the target data cost value to obtain the second industry portrait data restoration model;
[0100] A model restoration module 250 is used to use the second industry portrait data restoration model as a target industry portrait data restoration model for completing training if it is determined that the second industry portrait data restoration model has matched the training termination requirement;
[0101] The image analysis module 260 is used to, otherwise, use the second industry portrait data restoration model as the first industry portrait data restoration model, and return to execute the step of passing the reference redundant industry portrait source data to the first industry portrait data restoration model; wherein, the target industry portrait data restoration model is used to perform industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain.
[0102] Based on the above, please refer to Figure 3 , shows an industry portrait analysis system 300 combined with enterprise chain, including a processor 310 and a memory 320 that communicate with each other, and the processor 310 is used to read and execute a computer program from the memory 320 to implement the above method.
[0103] Based on the above, a computer-readable storage medium is also provided, on which a computer program stored implements the above method when running.
[0104] On the basis of the above, the present invention also provides a cloud platform, including a processor and a memory that communicate with each other, wherein the processor is used to call a computer program from the memory and implement the method described in any one of claims 1-8 by running the computer program.
[0105] In summary, based on the above scheme, the reference industry portrait complete source data can be used to replace and hide data features and load non-relevant features to obtain reference redundant industry portrait source data, that is, industry portrait source data with data feature replacement, hiding and redundancy. The first industry portrait data restoration model is trained based on the reference industry portrait complete source data and the reference redundant industry portrait source data to obtain the trained target industry portrait data restoration model, and the industry portrait analysis of the restored industry portrait source data combined with the enterprise chain is performed based on the target industry portrait data restoration model, which has good restoration visualization quality and completeness, and can remove data feature replacement, hiding and interference from the restored industry portrait source data, improve the precision and credibility of the restored industry portrait source data, thereby improving the accuracy of the industry portrait source data, thereby improving the visualization quality and completeness of the industry portrait source data display.
[0106] It should be understood that the system and its modules shown above can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or a dedicated design hardware. Those skilled in the art will understand that the above methods and systems can be implemented using computer executable instructions and / or included in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of the present application can not only be implemented by hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software such as executed by various types of processors, and can also be implemented by a combination of the above hardware circuits and software (e.g., firmware).
[0107] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced may be any one or a combination of the above, or any other beneficial effects that may be obtained.
[0108] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of the present application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements and amendments to the present application. Such modifications, improvements and amendments are suggested in the present application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of the present application.
[0109] At the same time, the present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.
[0110] In addition, it will be appreciated by those skilled in the art that various aspects of the present application may be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of the present application may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.
[0111] A computer storage medium may include a propagated data signal containing computer program code, for example, in baseband or as part of a carrier wave. The propagated signal may be in a variety of forms, including electromagnetic, optical, etc., or a suitable combination. A computer storage medium may be any computer-readable medium other than a computer-readable storage medium, which can be connected to an instruction execution system, device or apparatus to communicate, propagate or transmit the program for use. The program code on the computer storage medium may be transmitted via any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the above media.
[0112] The computer program codes required for the operation of each part of the present application can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., conventional procedural programming languages such as C language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages, etc. The program code can be run entirely on the user's computer, or run on the user's computer as an independent software package, or run partially on the user's computer and partially on a remote computer, or run entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).
[0113] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this application, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this application. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this application. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.
[0114] Similarly, it should be noted that in order to simplify the description of the disclosure of this application and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this application, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the object of this application are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.
[0115] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used for the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers allow adaptive changes. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which can be changed according to the required characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining the digits. Although the numerical domains and parameters used to confirm the breadth of the range in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values is as accurate as possible within the feasible range.
[0116] Each patent, patent application, patent application disclosure, and other materials, such as articles, books, instructions, publications, documents, etc., cited in this application are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the content of this application, documents that limit the broadest scope of the claims of this application (currently or later attached to this application) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the attached materials of this application are inconsistent or conflicting with the content described in this application, the descriptions, definitions, and / or use of terms in this application shall prevail.
[0117] Finally, it should be understood that the embodiments described in this application are only used to illustrate the principles of the embodiments of the present application. Other variations may also fall within the scope of the present application. Therefore, as an example and not a limitation, the alternative configurations of the embodiments of the present application may be considered to be consistent with the teachings of the present application. Accordingly, the embodiments of the present application are not limited to the embodiments explicitly introduced and described in the present application.
[0118] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. An industrial portrait analysis method combined with enterprise chain, It is characterized in that Applied to an industry portrait analysis system, the method at least includes: obtaining complete source data of a reference industry portrait, and performing data feature replacement and hiding on the complete source data of the reference industry portrait, and obtaining missing source data of a reference industry portrait related to the complete source data of the reference industry portrait, wherein the missing source data of the reference industry portrait is industry portrait source data with data feature replacement and hiding; And, non-related features are loaded on the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait, wherein the reference redundant industry portrait source data is industry portrait source data with data feature replacement, hidden and redundant data; The reference redundant industry portrait source data is passed to the first industry portrait data restoration model, and the first industry portrait data restoration model performs industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain, to obtain reference restored industry portrait source data related to the reference redundant industry portrait source data; Calculate the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data, update the weight information of the first industry portrait data restoration model according to the target data cost value, and obtain the second industry portrait data restoration model; If it is determined that the second industry portrait data restoration model has matched the training termination requirement, the second industry portrait data restoration model is used as the target industry portrait data restoration model for completing the training; Otherwise, the second industry portrait data restoration model is used as the first industry portrait data restoration model, and the step of transferring the reference redundant industry portrait source data to the first industry portrait data restoration model is returned; wherein the target industry portrait data restoration model is used to perform industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain; The first industry portrait data restoration model includes a first portrait data feature screening unit, a feature dimension reduction structure, a feature mapping structure, a feature dimension increase structure, a feature fusion structure, and a second portrait data feature screening unit; The method of passing the reference redundant industry portrait source data to the first industry portrait data restoration model, and having the first industry portrait data restoration model perform industry portrait analysis on the reference redundant industry portrait source data in combination with the enterprise chain, and obtaining reference restored industry portrait source data related to the reference redundant industry portrait source data, includes: passing the reference redundant industry portrait source data to the first portrait data feature screening unit, and having the first portrait data feature screening unit perform portrait data feature screening on the reference redundant industry portrait source data, and obtaining first portrait data features; The first portrait data feature is passed to a feature dimension reduction structure and a feature fusion structure, and the feature dimension reduction structure performs feature dimension reduction on the first portrait data feature to obtain a second portrait data feature; The second portrait data feature is transferred to a feature mapping structure, and the feature mapping structure performs feature mapping processing on the second portrait data feature to obtain a third portrait data feature; The third portrait data feature is transmitted to a feature dimension upgrading structure, and the feature dimension upgrading structure performs feature dimension upgrading on the third portrait data feature to obtain a fourth portrait data feature; The fourth portrait data feature is transmitted to a feature fusion structure, and the feature fusion structure performs superposition processing on the fourth portrait data feature and the first portrait data feature to obtain a fifth portrait data feature; The fifth portrait data feature is transmitted to the second portrait data feature screening unit, so that the second portrait data feature screening unit performs portrait data feature screening on the fifth portrait data feature to obtain the reference restored industry portrait source data; The number of the feature fusion structures is several, and the number of the feature mapping structures is several; The number of the feature dimension-raising structures is the same as the number of the feature dimension-reducing structures, and the number of the feature fusion structures is the same as the number of the feature dimension-raising structures; For each feature dimensionality reduction structure, after the image data feature passes through the feature dimensionality reduction structure, the number of feature categories of the image data feature increases, the feature vertical range of the image data feature decreases, and the feature horizontal range of the image data feature decreases; For each feature dimension-raising structure, after the image data feature passes through the feature dimension-raising structure, the number of feature categories of the image data feature is reduced, the vertical range of the feature of the image data feature is expanded, and the horizontal range of the feature of the image data feature is expanded; The calculating the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data includes: calculating the structured data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data; Calculate the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data; Calculate the frequent item data cost between the reference restored industry portrait source data and the reference industry portrait complete source data; Calculate the target data cost value between the reference restored industry portrait source data and the reference industry portrait complete source data according to the structured data cost value, the unstructured data cost value and the frequent item data cost value; The calculation of the structured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data includes: calculating the mean square cost based on the knowledge member data variable values of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable values of each knowledge entity in the reference industry portrait complete source data; Calculating the structured data cost value according to the mean square cost value and the industry portrait source data; The calculation of the unstructured data cost between the reference restored industry portrait source data and the reference industry portrait complete source data includes: calculating the first average value and the first standard deviation value corresponding to the reference restored industry portrait source data, the second average value and the second standard deviation value corresponding to the reference industry portrait complete source data, and the average standard deviation value between the reference restored industry portrait source data and the reference industry portrait complete source data based on the knowledge member data variable values of each knowledge entity in the reference restored industry portrait source data and the knowledge member data variable values of each knowledge entity in the reference industry portrait complete source data; Calculating the unstructured data cost value according to the first average value, the second average value, the first standard deviation value, the second standard deviation value, and the average standard deviation value; The calculating of the frequent item data substitution value between the reference restored industry portrait source data and the reference industry portrait complete source data includes: passing the reference restored industry portrait source data to a variable mining unit to obtain a first data variable value corresponding to the reference restored industry portrait source data; The complete source data of the reference industry portrait is transferred to the variable mining unit to obtain the second data variable value corresponding to the complete source data of the reference industry portrait; Calculate the frequent item data cost value according to the first data variable value and the second data variable value; After the second industry portrait data restoration model is used as the target industry portrait data restoration model for training, the method further includes: obtaining industry portrait source data to be restored, wherein the industry portrait source data to be restored is industry portrait source data with data feature replacement, hiding and redundancy; The industry portrait source data to be restored is passed to the target industry portrait data restoration model, and the target industry portrait data restoration model performs industry portrait analysis on the industry portrait source data to be restored in combination with the enterprise chain to obtain complete industry portrait source data related to the industry portrait source data to be restored.
2. The method according to claim 1, It is characterized in that The data feature replacement and hiding of the complete source data of the reference industry portrait to obtain the missing source data of the reference industry portrait related to the complete source data of the reference industry portrait includes: feature processing of the complete source data of the reference industry portrait through a pre-set data feature replacement and hiding template to obtain the missing source data of the reference industry portrait; wherein the pre-set data feature replacement and hiding template includes a number of feature processing quantization results, and the number of feature processing quantization results match the pre-set feature processing rules.
3. The method according to claim 1, It is characterized in that The method of loading non-relevant features for the missing source data of the reference industry portrait to obtain reference redundant industry portrait source data related to the missing source data of the reference industry portrait includes: loading non-relevant features for the missing source data of the reference industry portrait to obtain the reference redundant industry portrait source data.
4. An industry portrait analysis system combined with enterprise chain, It is characterized in that The method comprises a processor and a memory communicating with each other, wherein the processor is used to call a computer program from the memory and implement the method according to any one of claims 1 to 3 by running the computer program.
5. A cloud platform, It is characterized in that The method comprises a processor and a memory communicating with each other, wherein the processor is used to call a computer program from the memory and implement the method according to any one of claims 1 to 3 by running the computer program.
Citation Information
Patent Citations
Enterprise portrait drawing method, system and device based on industrial map and medium
CN112256762A
Customer portrait method, apparatus, computer-readable storage medium, and terminal device
WO2021147557A1
