Method, apparatus, device and medium for heterogeneous platform data mining retrieval and data analysis and processing

Abstract

The present invention provides a method, apparatus, device, and medium for data mining retrieval and data analysis processing on heterogeneous platforms. The method includes: generating a data mining strategy and a data analysis strategy based on multimodal live features; executing the data mining strategy to retrieve the mined target data; performing data analysis on the target data based on the data analysis strategy; and performing adaptive processing on the target data based on the data analysis result. The data mining retrieval and data analysis processing method on heterogeneous platforms of the present invention generates a data mining strategy and a data analysis strategy based on multimodal live features, executes the data mining strategy to mine the target data, performs data analysis on the target data based on the data analysis strategy, and performs adaptive processing on the target data based on the data analysis result. It is applicable to situations where the data sources, data types, and data usage purposes on heterogeneous platforms often change, improving the accuracy of mining and analysis and enhancing the business capabilities of heterogeneous platforms.

Description

Technical Field

[0001] The present invention relates to the technical field of computer data processing, and particularly to a method, device, equipment and medium for heterogeneous platform data mining retrieval and data analysis and processing. Background Art

[0002] Heterogeneous platforms have become important tools for processing large-scale data mining and analysis tasks. However, when traditional heterogeneous platforms perform data mining and analysis, they mostly adopt fixed mining and analysis strategies. For example, for the same type of data mining scenario, the same data mining method is used for data mining. However, the data sources, data types, data usage purposes, etc. of heterogeneous platforms often change. If a fixed mining and analysis strategy is still always adopted, it may lead to a significant reduction in the accuracy of mining and analysis, and further reduce the business capabilities of heterogeneous platforms.

[0003] Therefore, a solution is urgently needed. Summary of the Invention

[0004] One of the purposes of the present invention is to provide a method for heterogeneous platform data mining retrieval and data analysis and processing, which generates a data mining strategy and a data analysis strategy based on multi-modal live features, executes the data mining strategy to mine target data, performs data analysis on the target data based on the data analysis strategy, and performs adaptive processing on the target data based on the data analysis result. It is applicable to the situation where the data sources, data types, data usage purposes, etc. of heterogeneous platforms often change, improves the accuracy of mining and analysis, and improves the business capabilities of heterogeneous platforms.

[0005] The method for heterogeneous platform data mining retrieval and data analysis and processing provided by the embodiments of the present invention is applied to a heterogeneous platform and includes:

[0006] Generating a data mining strategy and a data analysis strategy based on multi-modal live features;

[0007] Executing the data mining strategy to retrieve the mined target data;

[0008] Performing data analysis on the target data based on the data analysis strategy;

[0009] Performing adaptive processing on the target data based on the data analysis result.

[0010] Preferably, the generating a data mining strategy and a data analysis strategy based on multi-modal live features includes:

[0011] Analyzing the feature type set of multi-modal live features;

[0012] Matching the feature type set with multiple standard feature type sets respectively to obtain multiple first matching degrees;

[0013] Obtain multiple levels of standard feature sets associated with the standard feature type set with the maximum first matching degree;

[0014] Traverse each standard feature set in ascending order of levels;

[0015] Each time when traversing, match the multi-modal live features with multiple standard features in the traversed standard feature set respectively. When the matched standard features meet the feature joint trigger condition, take the matched standard features as target features, and construct a description feature set based on the target features obtained during this traversal and the standard feature sets traversed historically; where, "matched and compliant" means that the second matching degree between the standard feature and the sub-feature of the multi-modal live feature is 100%;

[0016] Determine the policy generation knowledge corresponding to the description feature set from the policy generation knowledge base;

[0017] Based on the policy generation knowledge, determine the data mining policy and the data analysis policy;

[0018] Among them, the feature joint trigger condition includes one or more of the following combinations:

[0019] The sum of the trigger degrees of the matched and compliant standard features is greater than or equal to the trigger degree sum threshold;

[0020] The number of features of the matched and compliant standard features is greater than or equal to the quantity threshold corresponding to the level of the traversed standard feature set;

[0021] There is at least one feature association relationship among the matched and compliant standard features.

[0022] Preferably, after adaptively processing the target data based on the data analysis result, it further includes:

[0023] Obtain the processing records of adaptively processing different target data within the most recent first time period;

[0024] Analyze the visualization requirement value of the processing record;

[0025] When the visualization requirement value is greater than or equal to the requirement threshold, generate a visualization model based on the visualization model template according to the processing record;

[0026] Obtain the platform operation information of the user logging in to the heterogeneous platform within the most recent second time period;

[0027] Based on the platform operation information, perform targeted configuration on the visualization model;

[0028] Display the visually configured visualization model to the user.

[0029] Preferably, the visualization requirement value for parsing the processing record includes:

[0030] Determine whether there is at least one sub-record in the processing record whose object weight of the graph object corresponding to it in the requirement graph is greater than or equal to the weight threshold;

[0031] When it is yes, count the visualization requirement value of the processing record as the target value; otherwise, based on the quantization conversion table, perform quantization conversion on the sum of the object weights of the graph objects corresponding to the respective records of the processing record in the requirement graph to obtain the visualization requirement value of the processing record;

[0032] Wherein, the target value is greater than or equal to the requirement threshold.

[0033] Preferably, the targeted configuration of the visualization model based on the platform operation information includes:

[0034] Parse the information type set of the platform operation information;

[0035] Determine the area search rule corresponding to the information type set from the area search rule library;

[0036] Based on the area search rule, determine the feature extraction area from the visualization model;

[0037] Based on the first feature processing template, perform feature processing on the feature extraction area to obtain the area feature set;

[0038] Determine the second feature processing template corresponding to the area feature set from the feature processing template library;

[0039] Based on the second feature processing template, perform feature processing on the platform operation information to obtain the information feature set;

[0040] Obtain multiple groups of one-to-one corresponding standard feature combination pairs and configuration strategies;

[0041] Determine the target combination pair from the standard feature combination pairs; each of the area feature set and the information feature set contains one standard feature in the standard feature combination pairs;

[0042] Based on the configuration strategy corresponding to the target combination pair, perform targeted configuration on the visualization model.

[0043] The heterogeneous platform data mining retrieval and data analysis processing device provided by the embodiments of the present invention is applied to a heterogeneous platform and includes:

[0044] A generation module, configured to generate a data mining strategy and a data analysis strategy based on multi-modal live features;

[0045] A retrieval module, configured to execute the data mining strategy and retrieve the mined target data;

[0046] An analysis module for performing data analysis on target data based on a data analysis strategy;

[0047] A processing module for adaptively processing the target data based on the data analysis result.

[0048] Preferably, the generation module generates a data mining strategy and a data analysis strategy based on multi-modal live features, including:

[0049] Analyzing the feature type set of multi-modal live features;

[0050] Matching the feature type set with multiple standard feature type sets respectively to obtain multiple first matching degrees;

[0051] Obtaining multiple levels of standard feature sets associated with the standard feature type set with the maximum first matching degree;

[0052] Traversing each standard feature set in ascending order of levels;

[0053] Each time when traversing, matching the multi-modal live features with multiple standard features in the traversed standard feature set respectively. When the matching standard features meet the feature joint trigger condition, taking the matching standard features as target features, and constructing a description feature set based on the target features obtained during this traversal and the previously traversed standard feature sets; where, matching compliance means that the second matching degree between the standard feature and the sub-feature of the multi-modal live feature is 100%;

[0054] Determining the strategy generation knowledge corresponding to the description feature set from the strategy generation knowledge base;

[0055] Determining a data mining strategy and a data analysis strategy based on the strategy generation knowledge;

[0056] Among them, the feature joint trigger condition includes one or more of the following combinations:

[0057] The sum of the trigger degrees of the matching standard features is greater than or equal to the trigger degree sum threshold;

[0058] The number of the matching standard features is greater than or equal to the quantity threshold corresponding to the level of the traversed standard feature set;

[0059] There is at least one feature association relationship among the matching standard features.

[0060] Preferably, after the processing module adaptively processes the target data based on the data analysis result, it further includes:

[0061] A visualization module for:

[0062] Obtain the processing records of adaptive processing of different target data within the most recent first time period;

[0063] Analyze the visualization requirement values of the processing records;

[0064] When the visualization requirement value is greater than or equal to the requirement threshold, generate a visualization model based on the visualization model template according to the processing records;

[0065] Obtain the platform operation information performed by the user logging in to the heterogeneous platform within the most recent second time period;

[0066] Perform targeted configuration on the visualization model based on the platform operation information;

[0067] Display the visualization model after targeted configuration to the user.

[0068] An electronic device provided by an embodiment of the present invention includes a memory, a processor, and a computer program stored on the memory and executable. When the processor executes the program, the steps of the heterogeneous platform data mining retrieval and data analysis processing method described in any one of the above are implemented.

[0069] A computer-readable storage medium provided by an embodiment of the present invention stores a computer program. When the program is executed by a processor, the steps of the heterogeneous platform data mining retrieval and data analysis processing method described in any one of the above are implemented.

[0070] Other features and advantages of the present invention will be described in the following specification, and part of them will become obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in the written specification and the drawings.

[0071] Next, through the drawings and embodiments, the technical solutions of the present invention will be further described in detail. Description of the Drawings

[0072] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0073] Figure 1 It is a schematic diagram of the heterogeneous platform data mining retrieval and data analysis processing method in an embodiment of the present invention;

[0074] Figure 2 It is a schematic diagram of the heterogeneous platform data mining retrieval and data analysis processing device in an embodiment of the present invention. Detailed Embodiments

[0075] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for explaining and illustrating the present invention, and are not used to limit the present invention.

[0076] An embodiment of the present invention provides a method for heterogeneous platform data mining retrieval and data analysis processing, which is applied to a heterogeneous platform, such as Figure 1 as shown, including:

[0077] S1. Generate a data mining strategy and a data analysis strategy based on multimodal live features;

[0078] S2. Execute the data mining strategy to retrieve the mined target data;

[0079] S3. Perform data analysis on the target data based on the data analysis strategy;

[0080] S4. Perform adaptive processing on the target data based on the data analysis result.

[0081] In the above technical solution, the multimodal live features include: the type of scenario to be mined for data, the purpose of using the mined data, etc.; based on the multimodal live features, a data mining strategy and a data analysis strategy are generated. The data mining strategy is a strategy indicating how the system performs data mining, and the data analysis strategy is a strategy indicating how the system performs data analysis on the mined target data; by executing the data mining strategy, the target data can be mined; based on the data analysis strategy, data analysis is performed on the target data; finally, based on the data analysis result, adaptive processing is performed on the target data. For example, if the data analysis result indicates that the data quality of the target data is low, when performing adaptive processing on it, a voiding process is performed.

[0082] This application generates a data mining strategy and a data analysis strategy based on multimodal live features, executes the data mining strategy to mine the target data, performs data analysis on the target data based on the data analysis strategy, and performs adaptive processing on the target data based on the data analysis result. It is applicable to situations where the data sources, data types, and data usage purposes of heterogeneous platforms often change, improving the accuracy of mining and analysis and enhancing the business capabilities of heterogeneous platforms.

[0083] In one embodiment, the generating a data mining strategy and a data analysis strategy based on multimodal live features includes:

[0084] Analyze the feature type set of the multimodal live features;

[0085] Match the feature type set with multiple standard feature type sets respectively to obtain multiple first matching degrees;

[0086] Obtain multiple levels of standard feature sets associated with the standard feature type set with the maximum first matching degree;

[0087] Traverse each standard feature set in ascending order of levels;

[0088] Each time when traversing, match the multimodal live features with multiple standard features in the traversed standard feature set respectively. When the matched standard features meet the feature joint trigger condition, regard the matched standard features as target features, and construct a description feature set based on the target features obtained during this traversal and the standard feature sets traversed historically; where, being matched means that the second matching degree between the standard feature and the sub-features of the multimodal live features is 100%;

[0089] Determine the policy generation knowledge corresponding to the description feature set from the policy generation knowledge base;

[0090] Determine the data mining policy and data analysis policy based on the policy generation knowledge;

[0091] Among them, the feature joint trigger condition includes one or more of the following combinations:

[0092] The sum of the trigger degrees of the matched standard features is greater than or equal to the trigger degree sum threshold;

[0093] The number of features of the matched standard features is greater than or equal to the quantity threshold corresponding to the level of the traversed standard feature set;

[0094] There is at least one feature association relationship among the matched standard features.

[0095] The feature type set includes multiple feature types of multimodal live features; there is a preset standard feature type set, and different levels of standard feature sets are associated with the standard feature type set. The greater the first matching degree between the feature type set and the standard feature type set, the more suitable the multimodal live features are for using the different levels of standard feature sets associated with the corresponding standard feature type set to screen out the description feature set for generating knowledge of the determination strategy; therefore, obtain multiple levels of standard feature sets associated with the standard feature type set with the maximum first matching degree; the higher the level of the standard feature set, the more it needs to be traversed preferentially. Each time it is traversed, the multimodal live features are respectively matched with multiple standard features in the traversed standard feature set. When the matching standard features meet the feature joint trigger condition, it indicates that the determination of the strategy generation knowledge can be performed currently. The matching standard features that meet the condition are used as target features, and a description feature set is constructed based on the target features obtained during this traversal and the standard feature sets traversed historically; there is strategy generation knowledge corresponding to different description feature sets in the strategy generation knowledge base. Based on the strategy generation knowledge, determine the data mining strategy and the data analysis strategy; specifically, for example: the feature types in the standard feature type set are data source and data usage purpose, and the standard features in the different levels of standard feature sets associated with it are that the data source is a shopping platform and the data usage purpose is to analyze user shopping habits. Then, when the standard features meet the feature joint trigger condition, they are used as target features, and the corresponding strategy generation knowledge after forming the description feature set is that the data mining strategy is to crawl the frequency of users purchasing the same type of goods on the shopping platform, and the data analysis strategy is to analyze the average value, mode, etc. of this frequency.

[0096] In the feature joint trigger condition, the trigger degree represents the degree to which the standard features that can be represented after matching with the standard features can be used to determine the basis for generating the strategy generation knowledge, that is, the description feature set. It can be pre-set by technicians according to actual needs; a quantity threshold is preset corresponding to the level of the standard feature set. When the number of feature numbers of the matching standard features is greater than or equal to the quantity threshold corresponding to the level of the traversed standard feature set, it represents that the standard features can be used to determine the basis for generating the strategy generation knowledge, that is, the description feature set; the feature association relationship can be elements belonging to the same event, etc.

[0097] In the embodiment of the present invention, when generating the data mining strategy and the data analysis strategy based on the multimodal live features, the standard feature type set and the different levels of standard feature sets associated with the standard feature type set are introduced, which is convenient for quickly determining the basis for generating the strategy generation knowledge, that is, the description feature set, and greatly improves the accuracy and efficiency of generating the data mining strategy and the data analysis strategy.

[0098] In one embodiment, after adaptively processing the target data based on the data analysis result, it further includes:

[0099] Obtain the processing records of adaptively processing different target data within the most recent first time period; the most recent first time period can be the most recent 0.5 days;

[0100] Analyze the visualization requirement value of the processing records; the visualization requirement value represents the degree of requirement for visualizing the processing records;

[0101] When the visualization requirement value is greater than or equal to the requirement threshold, based on the visualization model template, generate a visualization model according to the processing records; the requirement threshold can be 8; when the visualization requirement value is greater than or equal to the requirement threshold, based on the visualization model template, generate a visualization model according to the processing records; the visualization model template is a template for generating a visualization model from the processing records, which can be set in advance by technicians;

[0102] Obtain the platform operation information of the users logging in to the heterogeneous platform within the most recent second time period; the most recent second time period can be 1000 seconds; the platform operation information includes the operation type, operation time, operation object, etc. of the users' operations on the heterogeneous platform;

[0103] Based on the platform operation information, perform targeted configuration on the visualization model; perform targeted configuration on the heterogeneous platform to make the visualization model more suitable for display to users;

[0104] Display the visualization model after targeted configuration to the users.

[0105] In one embodiment, the visualization requirement value for analyzing the processing records includes:

[0106] Determine whether there is at least one sub-record in the processing records whose object weight of the corresponding graph object in the requirement graph is greater than or equal to the weight threshold; there are graph objects corresponding to different sub-records in the requirement graph, and the object weight of the graph object represents the degree of requirement for visualizing the processing records. Technicians can pre-set the requirement graph according to actual needs; the weight threshold can be 8;

[0107] When it is yes, count the visualization requirement value of the processing records as the target value; otherwise, based on the quantization conversion table, perform quantization conversion on the sum of the object weights of the corresponding graph objects in the requirement graph of each record of the processing records to obtain the visualization requirement value of the processing records; when the object weight of the graph object corresponding to one sub-record in the requirement graph is greater than or equal to the weight threshold, it is sufficient to represent that the processing records need to be visualized, and count the visualization requirement value of the processing records as the target value; there are visualization requirement values corresponding to different sums of object weights in the quantization conversion table. Based on the quantization conversion table, perform quantization conversion on the sum of the object weights of the corresponding graph objects in the requirement graph of each record of the processing records to obtain the visualization requirement value of the processing records;

[0108] Among them, the target value is greater than or equal to the requirement threshold.

[0109] In one embodiment, the targeted configuration of the visualization model based on the platform operation information includes:

[0110] Analyze the information type set of the platform operation information; there are multiple information types of the platform operation information in the information type set;

[0111] Determine the area search rule corresponding to the information type set from the area search rule library; there are area search rules corresponding to different information type sets in the area search rule library. The information type set represents a situation where area search for feature extraction is required, and the area search rule is a rule indicating the system to search and determine the feature extraction area from the visualization model in this situation. For example, if the information type in the information type set is that the user selects to view the historical online user shopping habit data, the corresponding area search rule is to search out the data display area related to shopping habits in the visualization model;

[0112] Based on the area search rule, determine the feature extraction area from the visualization model;

[0113] Based on the first feature processing template, perform feature processing on the feature extraction area to obtain an area feature set; there are multiple area features of the feature extraction area in the area feature set;

[0114] Determine the second feature processing template corresponding to the area feature set from the feature processing template library; there are second feature processing templates corresponding to different area feature sets in the feature processing template library. The area features in the area feature set reflect which features are needed for the targeted configuration of the visualization model, so as to perform feature processing on the platform operation information to extract the corresponding information features and form an information feature set. For example, if the area feature in the area feature set is the shopping habit data of elderly online users, the corresponding feature processing template is to extract the features that may reflect the user's interest in the shopping habit data of elderly online users;

[0115] Based on the second feature processing template, perform feature processing on the platform operation information to obtain an information feature set;

[0116] Obtain multiple groups of one-to-one corresponding standard feature combination pairs and configuration strategies;

[0117] Determine the target combination pair from the standard feature combination pairs; each of the area feature set and the information feature set contains one standard feature in the standard feature combination pair; a preset standard feature combination pair, when; each of the area feature set and the information feature set contains one standard feature in the standard feature combination pair, it indicates that the corresponding configuration strategy can be used for the targeted configuration of the visualization model;

[0118] Target the visualization model for configuration according to the corresponding configuration strategy for the target combination.

[0119] This application targets the visualization model for configuration based on platform operation information, facilitating users to quickly, intuitively, and specifically understand the data mining and data analysis processes of the platform when using heterogeneous platforms, improving the applicability of the system, and also making it more intelligent and user-friendly.

[0120] The heterogeneous platform data mining retrieval and data analysis processing device provided by the embodiments of the present invention is applied to heterogeneous platforms, such as Figure 2 as shown, and includes:

[0121] A generation module 1, configured to generate a data mining strategy and a data analysis strategy based on multi-modal live features;

[0122] A retrieval module 2, configured to execute the data mining strategy and retrieve the target data mined;

[0123] An analysis module 3, configured to perform data analysis on the target data based on the data analysis strategy;

[0124] A processing module 4, configured to perform adaptive processing on the target data based on the data analysis result.

[0125] The generation module generates a data mining strategy and a data analysis strategy based on multi-modal live features, including:

[0126] Analyze the feature type set of the multi-modal live features;

[0127] Match the feature type set with multiple standard feature type sets respectively to obtain multiple first matching degrees;

[0128] Obtain multiple levels of standard feature sets associated with the standard feature type set with the maximum first matching degree;

[0129] Traverse each standard feature set in ascending order of levels;

[0130] Each time when traversing, match the multi-modal live features with multiple standard features in the traversed standard feature set respectively. When the matched standard features meet the feature joint trigger condition, use the matched standard features as target features, and construct a description feature set based on the target features obtained during this traversal and the previously traversed standard feature sets; where, "matched" means that the second matching degree between the standard feature and the sub-feature of the multi-modal live feature is 100%;

[0131] Determine the strategy generation knowledge corresponding to the description feature set from the strategy generation knowledge base;

[0132] Generate knowledge based on policies, and determine data mining policies and data analysis policies;

[0133] Among them, the feature combined trigger conditions include one or more of the following combinations:

[0134] The sum of the trigger degrees of the matching standard features is greater than or equal to the trigger degree sum threshold;

[0135] The number of features that match the standard features is greater than or equal to the corresponding quantity threshold of the level of the traversed standard feature set;

[0136] There is at least one feature association relationship among the matching standard features.

[0137] After the processing module performs adaptive processing on the target data based on the data analysis result, it further includes:

[0138] A visualization module, which is used for:

[0139] Obtain the processing records of adaptive processing of different target data in the most recent first time period;

[0140] Analyze the visualization requirement values of the processing records;

[0141] When the visualization requirement value is greater than or equal to the requirement threshold, based on the visualization model template, generate a visualization model according to the processing records;

[0142] Obtain the platform operation information of the user logging in to the heterogeneous platform in the most recent second time period;

[0143] Based on the platform operation information, perform targeted configuration on the visualization model;

[0144] Display the visualization model after targeted configuration to the user.

[0145] An electronic device provided by an embodiment of the present invention includes a memory, a processor, and a computer program stored on the memory and executable, and when the processor executes the program, the steps of the heterogeneous platform data mining retrieval and data analysis processing method as described in any one of the above are implemented.

[0146] A computer-readable storage medium provided by an embodiment of the present invention stores a computer program, and when the program is executed by a processor, the steps of the heterogeneous platform data mining retrieval and data analysis processing method as described in any one of the above are implemented.

[0147] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A method for data mining, retrieval and data analysis and processing on a heterogeneous platform, applied to a heterogeneous platform, characterized in that: include: Generate data mining and data analysis strategies based on multimodal real-time features; Execute data mining strategies and retrieve the mined target data; Based on the data analysis strategy, perform data analysis on the target data; Based on the data analysis results, adaptively process the target data; Obtaining a processing record of adaptively processing different target data within a recent first time period; Analyze and process the recorded visualization requirements; When the visualization requirement value is greater than or equal to the requirement threshold, a visualization model is generated based on the visualization model template and the processing records; Obtaining platform operation information performed by users who log in to the heterogeneous platform within the most recent second time period; Based on the platform operation information, the visualization model is configured in a targeted manner; Displaying a specially configured visualization model to the user; The method of generating data mining strategies and data analysis strategies based on multimodal real-time features includes: A feature type set for parsing multimodal ground truth features; Matching the feature type set with a plurality of standard feature type sets respectively to obtain a plurality of first matching degrees; Acquire a plurality of levels of standard feature sets associated with the standard feature type set having the maximum first matching degree; Traverse each standard feature set in order from small to large levels; Each time the multimodal actual situation feature is traversed, the multimodal actual situation feature is matched with multiple standard features in the traversed standard feature set. When the matched standard feature meets the feature joint trigger condition, the matched standard feature is used as the target feature, and the description feature set is constructed based on the target feature obtained in the current traversal and the historical traversal of the standard feature set; wherein the matching means that the second matching degree between the standard feature and the sub-feature of the multimodal actual situation feature is 100%; Determine the strategy generation knowledge corresponding to the description feature set from the strategy generation knowledge base; Based on the strategy-generated knowledge, determine the data mining strategy and data analysis strategy; The feature joint triggering conditions include one or more of the following combinations: The sum of the triggering degrees of the matching standard features is greater than or equal to the triggering degree and the threshold; The number of features that match the standard features is greater than or equal to the number threshold corresponding to the level of the traversed standard feature set; There is at least one feature association relationship between the matching standard features.

2. The heterogeneous platform data mining, retrieval and data analysis processing method according to claim 1, characterized in that: The visualization requirement value of the analysis and processing record includes: Determine whether there is at least one sub-record in the processing record whose object weight of the graph object corresponding to the demand graph is greater than or equal to a weight threshold; When the answer is yes, the visualization requirement value of the processing record is calculated as the target value; otherwise, based on the quantitative conversion table, the sum of the object weights of the graph objects corresponding to each record of the processing record in the demand graph is quantitatively converted to obtain the visualization requirement value of the processing record; Among them, the target value is greater than or equal to the demand threshold.

3. The heterogeneous platform data mining, retrieval and data analysis processing method according to claim 1, characterized in that: The targeted configuration of the visualization model based on the platform operation information includes: A set of information types for parsing platform operation information; Determine the regional search rule corresponding to the information type set from the regional search rule library; Based on the region search rule, determine the feature extraction region from the visualization model; Based on the first characterization processing template, the feature extraction region is subjected to characterization processing to obtain a regional feature set; Determine a second characterization processing template corresponding to the regional feature set from the characterization processing template library; Based on the second characterization processing template, the platform operation information is characterized to obtain an information feature set; Obtain multiple sets of one-to-one corresponding standard feature combination pairs and configuration strategies; Determine a target combination pair from the standard feature combination pairs; the regional feature set and the information feature set each contain a standard feature of the standard feature combination pair; Based on the configuration strategy corresponding to the target combination, the visualization model is configured in a targeted manner.

4. A heterogeneous platform data mining, retrieval and data analysis and processing device, applied to a heterogeneous platform, characterized in that: include: A generation module, used to generate data mining strategies and data analysis strategies based on multimodal real-time features; The retrieval module is used to execute the data mining strategy and retrieve the mined target data; An analysis module, used to perform data analysis on target data based on a data analysis strategy; A processing module, used for adaptively processing target data based on data analysis results; Visualization modules for: Obtaining a processing record of adaptively processing different target data within a recent first time period; Analyze and process the recorded visualization requirements; When the visualization requirement value is greater than or equal to the requirement threshold, a visualization model is generated based on the visualization model template and the processing records; Obtaining platform operation information performed by users who log in to the heterogeneous platform within the most recent second time period; Based on the platform operation information, the visualization model is configured in a targeted manner; Displaying a specially configured visualization model to the user; The generation module generates data mining strategies and data analysis strategies based on multimodal real-time features, including: A feature type set for parsing multimodal ground truth features; Matching the feature type set with a plurality of standard feature type sets respectively to obtain a plurality of first matching degrees; Acquire a plurality of levels of standard feature sets associated with the standard feature type set having the maximum first matching degree; Traverse each standard feature set in order from small to large levels; Each time the multimodal actual situation feature is traversed, the multimodal actual situation feature is matched with multiple standard features in the traversed standard feature set. When the matched standard feature meets the feature joint trigger condition, the matched standard feature is used as the target feature, and the description feature set is constructed based on the target feature obtained in the current traversal and the historical traversal of the standard feature set; wherein the matching means that the second matching degree between the standard feature and the sub-feature of the multimodal actual situation feature is 100%; Determine the strategy generation knowledge corresponding to the description feature set from the strategy generation knowledge base; Based on the strategy-generated knowledge, determine the data mining strategy and data analysis strategy; The feature joint triggering conditions include one or more of the following combinations: The sum of the triggering degrees of the matching standard features is greater than or equal to the triggering degree and the threshold; The number of features that match the standard features is greater than or equal to the number threshold corresponding to the level of the traversed standard feature set; There is at least one feature association relationship between the matching standard features.

5. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable, characterized in that: When the processor executes the program, the steps of the heterogeneous platform data mining retrieval and data analysis processing method as described in any one of claims 1-3 are implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the program is executed by a processor, the steps of the heterogeneous platform data mining, retrieval and data analysis processing method as described in any one of claims 1 to 3 are implemented.

Images