Resource pushing data processing method and device, computer device and storage medium

Abstract

The application relates to a resource pushing data processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: extracting features from acquired resource association information corresponding to an object set; classifying each object in the object set through a decision tree, acquiring historical resource data corresponding to a group category, and generating candidate item sets at different levels; based on a preset occurrence parameter and a comparison result of historical occurrence parameters of the candidate item sets at different levels, screening the candidate item sets to obtain a target frequent item set corresponding to the group category, and performing permutation and combination on each non-empty true subset corresponding to the target frequent item set to form a corresponding item association relationship; based on a comparison result of a confidence parameter corresponding to the item association relationship and a preset confidence parameter, obtaining and storing the group category and the corresponding target item association relationship; and finally, pushing the associated resource data to a terminal. The method can improve the utilization rate of resource data.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a resource push data processing method, apparatus, computer equipment, storage medium, and computer program product. Background Technology

[0002] In today's rapidly developing information age, everyone has different habits of using and combining resource data, creating correlations between different resource data usages. Analyzing these correlations to predict resource combinations can improve the utilization rate of people's resource data, thereby increasing their benefits.

[0003] However, the current method of distributing resource data is not targeted and does not take into account the correlation between different resource data uses. This results in no restrictions on the server sending resource data, leading to high pressure on the server to send resource data and low resource data utilization. Summary of the Invention

[0004] Therefore, it is necessary to address the aforementioned technical problems by providing a resource push data processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product capable of obtaining the correlation between different resource data, thereby improving the utilization rate of resource data.

[0005] A method for processing resource push data. The method includes:

[0006] Obtain the resource association information corresponding to the object set, and perform feature extraction to obtain the corresponding resource association features;

[0007] Based on the resource association features and resource association information, each object in the object set is divided into a corresponding group category using a decision tree;

[0008] Obtain historical resource data corresponding to the group category, generate multiple candidate item sets at different levels based on the historical resource data, and filter the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemset corresponding to the group category.

[0009] For each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category, the corresponding item association relationship is formed by permuting and combining them. Based on the comparison results of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained.

[0010] The association between the group categories and the corresponding target items is stored, and the association between the target items is used to push the associated resource data to the corresponding group categories.

[0011] In one instance, based on the resource association features and resource association information, a decision tree is used to classify each object in the object set into corresponding group categories, including:

[0012] The resource association information is divided into a training set and a validation set;

[0013] Calculate the information gain ratio corresponding to each of the resource association features;

[0014] The resource association feature with the largest information gain ratio corresponding to each of the resource association features is selected as the root node partitioning feature. The training set is then branched based on the root node partitioning feature to generate branch nodes, and each branch node has samples from the training set.

[0015] Calculate the information gain ratio corresponding to the resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node partitioning feature of each branch node, and perform branching operation on the samples in each branch node based on the node partitioning feature to generate new branch nodes.

[0016] Repeatedly calculate the information gain ratio corresponding to the resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node splitting feature of each branch node, perform branching operation on the samples in each branch node based on the node splitting feature, and generate new branch nodes until each sample has a clear classification, and obtain the decision tree set.

[0017] The decision tree set is predicted using the validation set to obtain the target decision tree, and the group category corresponding to each object in the object set is obtained based on the target decision tree.

[0018] In one embodiment, historical resource data corresponding to the group category is obtained, and multiple candidate item sets at different levels are generated based on the historical resource data. Based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item set, the candidate item sets are filtered to obtain the target frequent itemsets corresponding to the group category, including:

[0019] Obtain preset occurrence parameters;

[0020] Read the historical resource data, treat each historical resource data as an item, and treat each item as a candidate set at one level, and calculate the historical occurrence parameters of each candidate set at one level;

[0021] Select itemsets from the candidate itemsets at level 1 whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as frequent itemsets at the current level.

[0022] Combine the itemsets in the current layer frequent itemsets to generate candidate itemsets for the next layer, calculate the historical occurrence parameters of each candidate itemset for the next layer, and perform pre-pruning operations;

[0023] After the pre-pruning operation, select the itemsets in the next-level candidate itemset whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as the next-level frequent itemsets;

[0024] Returning to the steps of combining the itemsets in the current layer frequent itemset to generate the next layer candidate itemset, calculating the historical occurrence parameter of each next layer candidate itemset, and performing pre-pruning operations, until no itemset greater than or equal to the preset occurrence parameter can be selected, the target frequent itemset is obtained.

[0025] In one embodiment, each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category is permuted and combined to form the corresponding item association relationship. Based on the comparison result between the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained as follows:

[0026] Retrieve preset parameters;

[0027] Calculate the confidence parameters for the association between the items;

[0028] Item associations whose confidence parameters are greater than or equal to a preset confidence parameter are selected as the target item associations corresponding to the group category.

[0029] In one embodiment, after storing the association between the group category and the corresponding target item, the method further includes:

[0030] Obtain resource association information of the target object, and receive the basis conditions for the terminal corresponding to the target object to display the resource data;

[0031] Based on the resource association information of the target object, feature extraction is performed to obtain the resource association features of the target object;

[0032] Based on the resource association characteristics and resource association information of the target object, the target object is classified into groups to obtain the target group category of the target object;

[0033] Based on the target group category of the target object, obtain the target item association relationship corresponding to the target group category of the target object;

[0034] Based on the target item association relationship and the conditions, obtain the target resource data associated with the target object, and send the target resource data to the terminal corresponding to the target object.

[0035] In one embodiment, obtaining the target resource data associated with the target object based on the target item association relationship and the conditions includes:

[0036] Based on the association relationship of the target items, obtain the initial resource data corresponding to the target object;

[0037] When the basis condition is resource conversion rate, calculate the conversion rate of each initial resource data in the initial resource data, sort the initial resource data from largest to smallest based on the conversion rate of each initial resource data, and take the sorted initial resource data as the target resource data.

[0038] When the basis condition is the frequency of resource use, the confidence parameters of the target item association relationship corresponding to each initial resource data in the initial resource data are sorted from largest to smallest, and the sorted initial resource data is used as the target resource data.

[0039] A resource push data processing device, the device comprising:

[0040] The feature extraction module is used to obtain resource association information corresponding to the object set and perform feature extraction to obtain the corresponding resource association features;

[0041] The object classification module is used to classify each object in the object set into a corresponding group category based on the resource association features and resource association information using a decision tree.

[0042] The target frequent itemset acquisition module is used to acquire historical resource data corresponding to the group category, generate multiple candidate itemsets at different levels based on the historical resource data, and filter the candidate itemsets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate itemsets to obtain the target frequent itemsets corresponding to the group category.

[0043] The target item association module is used to arrange and combine each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category to form the corresponding item association. Based on the comparison result between the confidence parameter corresponding to the item association and the preset confidence parameter, the target item association corresponding to the group category is obtained.

[0044] The data storage module is used to store the association between the group categories and the corresponding target items. The association between the target items is used to push the associated resource data to the corresponding group categories.

[0045] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0046] Obtain the resource association information corresponding to the object set, and perform feature extraction to obtain the corresponding resource association features;

[0047] Based on resource association features and resource association information, each object in the object set is divided into corresponding group categories using a decision tree;

[0048] Obtain historical resource data corresponding to the group category, generate multiple candidate item sets at different levels based on the historical resource data, and filter the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemset corresponding to the group category.

[0049] For each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category, the corresponding item association relationship is formed by permuting and combining them. Based on the comparison results of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained.

[0050] Store the association between group categories and corresponding target items. The target item association is used to push associated resource data to the corresponding group category.

[0051] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following operations when executed by a processor:

[0052] Obtain the resource association information corresponding to the object set, and perform feature extraction to obtain the corresponding resource association features;

[0053] Based on resource association features and resource association information, each object in the object set is divided into corresponding group categories using a decision tree;

[0054] Obtain historical resource data corresponding to the group category, generate multiple candidate item sets at different levels based on the historical resource data, and filter the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemset corresponding to the group category.

[0055] For each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category, the corresponding item association relationship is formed by permuting and combining them. Based on the comparison results of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained.

[0056] Store the association between group categories and corresponding target items. The target item association is used to push associated resource data to the corresponding group category.

[0057] The aforementioned resource push data processing method, apparatus, computer equipment, and storage medium obtain resource association information corresponding to an object set, extract features to obtain corresponding resource association features, classify each object in the object set into corresponding group categories using a decision tree based on the resource association features and resource association information, obtain historical resource data corresponding to the group categories, generate multiple candidate item sets at different levels based on the historical resource data, filter the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain target frequent itemsets corresponding to the group categories, arrange and combine each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group categories to form corresponding item association relationships, obtain target item association relationships corresponding to the group categories based on the comparison results of confidence parameters corresponding to the item association relationships and preset confidence parameters, store the group categories and corresponding target item association relationships, and use the target item association relationships to push associated resource data to the corresponding group categories. By obtaining resource association information corresponding to the object set from the server, extracting the resource association features of the resource association information, and classifying each object in the object set into a group category based on the resource association features, and finding the relationship between historical resource data according to the historical resource data corresponding to the group category, the server can send resource data to the terminal in a targeted manner according to the relationship between different resource data usage, thereby reducing the data sending pressure on the server and improving resource utilization.

[0058] A method for receiving resource data, the method comprising:

[0059] The system acquires resource data pushed by the server, wherein the resource data is associated with a group category corresponding to an object identifier; the resource data is determined by the server through the association relationship between the group category and the corresponding target item; the association relationship between the target item corresponding to the group category is obtained by the server by acquiring resource association information corresponding to the object set, performing feature extraction to obtain the corresponding resource association features, and based on the resource association features, classifying each object in the object set into the corresponding group category through a decision tree, acquiring historical resource data corresponding to the group category, generating multiple candidate item sets at different levels based on the historical resource data, filtering the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemsets corresponding to the group category, arranging and combining each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group category to form the corresponding item association relationship, and obtaining the relationship based on the comparison results of the confidence parameters corresponding to the item association relationship and the preset confidence parameters.

[0060] Display the resource data.

[0061] A resource data receiving device, the device comprising:

[0062] The resource data acquisition module is used to acquire resource data pushed by the server. The resource data is associated with a group category corresponding to an object identifier. The resource data is determined by the server through the association relationship between the group category and the corresponding target item. The association relationship between the target item corresponding to the group category is obtained by the server by acquiring the resource association information corresponding to the object set, performing feature extraction to obtain the corresponding resource association features, and based on the resource association features and resource association information, classifying each object in the object set into the corresponding group category through a decision tree. The module acquires historical resource data corresponding to the group category, generates multiple candidate item sets at different levels based on the historical resource data, and filters the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item set to obtain the target frequent itemset corresponding to the group category. The module arranges and combines each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category to form the corresponding item association relationship, which is obtained based on the comparison results of the confidence parameter corresponding to the item association relationship and the preset confidence parameter.

[0063] The resource data display module is used to display the resource data.

[0064] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0065] The system acquires resource data pushed by the server, wherein the resource data is associated with a group category corresponding to an object identifier; the resource data is determined by the server through the association relationship between the group category and the corresponding target item; the association relationship between the target item corresponding to the group category is obtained by the server by acquiring resource association information corresponding to the object set, performing feature extraction to obtain the corresponding resource association features, and based on the resource association features, classifying each object in the object set into the corresponding group category through a decision tree, acquiring historical resource data corresponding to the group category, generating multiple candidate item sets at different levels based on the historical resource data, filtering the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemsets corresponding to the group category, arranging and combining each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group category to form the corresponding item association relationship, and obtaining the relationship based on the comparison results of the confidence parameters corresponding to the item association relationship and the preset confidence parameters.

[0066] Display the resource data.

[0067] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0068] The process involves acquiring resource data pushed by the server, where the resource data is associated with a group category corresponding to an object identifier. The resource data is determined by the server through the association between the group category and the corresponding target item. The association between the target item and the group category is obtained by the server acquiring resource association information corresponding to the object set, performing feature extraction to obtain corresponding resource association features, and using a decision tree to classify each object in the object set into a corresponding group category based on the resource association features and resource association information. Historical resource data corresponding to the group category is acquired, and multiple candidate item sets at different levels are generated based on the historical resource data. The candidate item sets are then filtered based on a comparison between preset occurrence parameters and historical occurrence parameters for each different level of candidate item set to obtain the target frequent itemset corresponding to the group category. Each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category is arranged and combined to form the corresponding item association, which is obtained based on a comparison between the confidence parameter corresponding to the item association and a preset confidence parameter.

[0069] Display the resource data.

[0070] The aforementioned resource data receiving method, apparatus, computer equipment, and storage medium acquire resource data pushed by a server. This resource data is associated with a group category corresponding to an object identifier. The resource data is determined by the server through the association relationship between the group category and the corresponding target item. The association relationship between the target item and the group category is obtained by the server acquiring resource association information corresponding to the object set, performing feature extraction to obtain corresponding resource association features, and using a decision tree to classify each object in the object set into a corresponding group category based on the resource association features and resource association information. Historical resource data corresponding to the group category is acquired, and multiple candidate item sets at different levels are generated based on the historical resource data. The candidate item sets are then filtered based on a comparison of preset occurrence parameters and historical occurrence parameters for each different level of candidate item set to obtain the target frequent itemset corresponding to the group category. Each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category is arranged and combined to form a corresponding item association relationship. The resource data is then displayed based on a comparison of the confidence parameter corresponding to the item association relationship and a preset confidence parameter. By obtaining pushed resource data from the server based on the correlation between different resource uses, and displaying the resource data to the terminal according to the criteria, the utilization rate of resources is effectively improved and the pressure on the server to send resource data is reduced. Attached Figure Description

[0071] Figure 1 This is an application environment diagram of a resource push data processing method in one embodiment;

[0072] Figure 2 This is a flowchart illustrating a resource push data processing method in one embodiment;

[0073] Figure 3 This is a flowchart illustrating the object classification process in one embodiment;

[0074] Figure 4 This is a schematic diagram of the process for obtaining the target frequent itemset in one embodiment;

[0075] Figure 5 This is a schematic diagram of the algorithm flow for obtaining the target frequent itemset in one embodiment;

[0076] Figure 6 This is a schematic diagram of the process for obtaining the association relationship of target items in one embodiment;

[0077] Figure 7 This is a schematic diagram of the data storage process in one embodiment;

[0078] Figure 8 This is a flowchart illustrating the condition judgment process in one embodiment;

[0079] Figure 9 This is a flowchart illustrating a resource data receiving method in one embodiment;

[0080] Figure 10 This is a structural block diagram of a resource push data processing device in one embodiment;

[0081] Figure 11 This is a structural block diagram of a resource data receiving device in one embodiment;

[0082] Figure 12 This is an internal structural diagram of a computer device in one embodiment;

[0083] Figure 13 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0084] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0085] The resource push data processing method provided in this application embodiment can be applied to, for example, Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on a cloud or other network server. Terminal 102 is used to acquire and display resource data. Server 104 is used to acquire resource association information, process the resource association information, and analyze the relationships between various resources in the resource association information corresponding to various groups. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0086] In one embodiment, such as Figure 2 As shown, a resource push data processing method is provided, which is applied to... Figure 1 Taking the server in the example, the following steps are included:

[0087] Step S200: Obtain the resource association information corresponding to the object set, and perform feature extraction to obtain the corresponding resource association features.

[0088] Here, the object set refers to the collection of various social groups using resources. Resource association information refers to the data information of each group within these social groups. Resource association features refer to the selection of representative features after representing the resource association information, and these representative features are used as resource association features. Resource association features include the general features of the object set and the specific features of the resources corresponding to the object set.

[0089] Specifically, before implementing this resource recommendation data method, massive amounts of data related to various objects need to be stored in a database. Leveraging Hadoop's suitability for big data storage and analysis, its ability to support clusters of thousands to tens of thousands of servers, and its support for petabyte-level storage capacity, this resource recommendation data method utilizes Hadoop for data storage and computation. When the method enters the deployment and implementation phase, it retrieves resource association information corresponding to the object sets from the database, performs feature representation on this information, and then selects representative resource association features as the data for subsequent processes.

[0090] Step S202: Based on resource association features and resource association information, each object in the object set is divided into corresponding group categories using a decision tree.

[0091] Decision trees refer to a type of supervised learning. Supervised learning involves providing a set of resource-related feature samples, each with a set of attributes and a pre-determined category. Through learning, a classifier is developed that can correctly classify newly encountered objects. Classifiers are a general term in data mining for classifying samples, encompassing algorithms such as decision trees, logistic regression, Naive Bayes, and neural networks.

[0092] Specifically, resource association features extracted from the resource association information corresponding to the object set on the server are used as the attribute features for constructing the decision tree, and the resource association information is used as the input data for the decision tree. In addition, before each branching operation on the decision tree node, it is necessary to calculate the information gain ratio of each resource association feature in each node, and select the resource association feature with the largest information gain ratio in each node as the node splitting feature of that node, until the sample data in each node has a clear classification.

[0093] Step S204: Obtain historical resource data corresponding to the group category, generate multiple candidate item sets at different levels based on the historical resource data, and filter the candidate item sets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemset corresponding to the group category.

[0094] Historical resource data refers to the data records of resource usage for each object in the object set. Candidate itemsets refer to the set of various resource combinations for each object in the object set. Preset occurrence parameters refer to fixed probability values ​​defined by the server for resource usage in any manner, used as reference values ​​for the probabilities of various resource usages to select resource usage scenarios with probabilities greater than or equal to the preset parameters. Historical occurrence parameters of candidate itemsets refer to the probabilities of various resource combinations for each object in the object set. Frequent itemsets refer to the set in the candidate itemset whose historical occurrence parameters are greater than or equal to the preset occurrence parameters. These frequent itemsets are used to discover association rules between resource usages, i.e., correlations between resource usages.

[0095] Specifically, the server obtains historical resource data corresponding to the group category, distinguishes each object in the object set into group categories through a decision tree, selects resource usage data records from the resource association information of the group category corresponding to each object as historical resource data corresponding to the group category, treats each historical resource data as an item, uses each item to form candidate itemsets at different levels, calculates the historical occurrence parameter of each candidate itemset at each level, and selects the itemsets whose historical occurrence parameter is greater than or equal to the preset occurrence parameter as the frequent itemsets of the corresponding level.

[0096] Step S206: Arrange and combine each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category to form the corresponding item association relationship. Based on the comparison result of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained.

[0097] Here, the target frequent itemset refers to the set of frequently used historical resource data within the historical resource data corresponding to the group category. Item association refers to the correlation between the historical resource data within the target frequent itemset. The confidence parameter corresponding to the item association refers to the conditional probability that one or more other historical resource data will also be used given that one or more historical resource data are used. The preset confidence parameter refers to a fixed value defined by the server indicating the conditional probability that one or more other resources will also be used given that one or more resources are used. The target item association refers to item associations where the confidence parameter is greater than or equal to the preset confidence parameter.

[0098] Specifically, the server retrieves each non-empty proper subset corresponding to the target frequent itemset of the obtained group category. At this point, the non-empty proper subsets are filtered, removing sets that were already removed before the target frequent itemset retrieval step but still appear in the non-empty proper subsets corresponding to the target frequent itemset. The removed non-empty proper subsets are then permuted and combined, and the confidence parameter for each combination is calculated. The confidence parameter for each combination is compared with a preset confidence parameter, and combinations whose confidence parameter is greater than or equal to the preset confidence parameter are selected as the target item association relationships corresponding to the respective group.

[0099] Step S208: Store the association between the group category and the corresponding target item. The association between the target item is used to push associated resource data to the corresponding group category.

[0100] Specifically, the server stores the group categories and corresponding target item associations of the resulting object set in the database and performs calculations and analyses on the data.

[0101] In the above-mentioned resource push data processing method, resource association information corresponding to the object set is obtained, and feature extraction is performed to obtain corresponding resource association features. Based on the resource association features and resource association information, each object in the object set is divided into corresponding group categories through a decision tree. Historical resource data corresponding to the group category is obtained, and multiple candidate item sets at different levels are generated based on the historical resource data. Based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate item sets, the candidate item sets are filtered to obtain the target frequent itemsets corresponding to the group category. Each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group category is arranged and combined to form the corresponding item association relationship. Based on the comparison results of the confidence parameters corresponding to the item association relationship and preset confidence parameters, the target item association relationship corresponding to the group category is obtained. The group category and the corresponding target item association relationship are stored. The target item association relationship is used to push associated resource data to the corresponding group category. By obtaining resource association information corresponding to the object set from the server, extracting the resource association features of the resource association information, and classifying each object in the object set into a group category based on the resource association features, and finding the relationship between historical resource data according to the historical resource data corresponding to the group category, the server can send resource data to the terminal in a targeted manner according to the relationship between different resource data usage, thereby reducing the data sending pressure on the server and improving resource utilization.

[0102] In one embodiment, the method requires multiple accesses to historical resource data, leading to low algorithm efficiency. Furthermore, the method is a serial algorithm, generating a large number of candidate itemsets, resulting in slow execution. To further address these issues, an improved AprioriTip algorithm can be employed. A Tip table is constructed to store subsets of candidate itemsets in the database. By scanning this Tip table, the support of candidate itemsets can be statistically analyzed, thereby identifying frequent itemsets. As the number of iterations increases, the number of itemsets requiring scanning decreases accordingly, thus improving algorithm efficiency. Alternatively, Hadoop can be used to parallelize the algorithm, splitting large-scale datasets into shards and setting up multiple nodes on different virtual machines within a single computer. Each node executes Map and Reduce tasks, thereby reducing program execution time.

[0103] In this context, a serial algorithm refers to an algorithm whose steps must be executed one after another, and cannot be executed simultaneously. Support refers to historical occurrence parameters. Parallelism refers to the simultaneous execution of multiple tasks at the same time. Map tasks are mainly used to filter some raw data, while Reduce tasks are mainly used to process the data filtered by Map tasks.

[0104] In one embodiment, such as Figure 3 As shown, step S202 includes:

[0105] Step S300: Divide the resource association information into a training set and a validation set.

[0106] Specifically, the resource-related information is divided into a training set and a validation set according to a custom ratio. The training set is used to train a classification model and fit multiple classifiers. Then, the validation set is used to predict using multiple classifiers to obtain the classification model with the best classification ability.

[0107] Step S302: Calculate the information gain ratio corresponding to each of the resource association features.

[0108] Information gain ratio is a feature selection criterion used to measure how a decision tree continuously divides the dataset into datasets with higher purity during the decision tree generation process.

[0109] Specifically, before determining the root node of the decision tree, the information gain ratio corresponding to each resource association feature is calculated to provide data for the initial partitioning of the training set in subsequent processes.

[0110] Step S304: Select the resource association feature with the largest information gain ratio corresponding to each of the resource association features as the root node partitioning feature, perform a branching operation on the training set based on the root node partitioning feature to generate branch nodes, and each branch node has samples in the training set.

[0111] In this context, the root node's partitioning feature refers to the resource association feature with the strongest classification ability. The branching operation refers to the process of further partitioning the samples within a decision tree node. A branch node is a node regenerated from a node. A sample refers to the data information corresponding to an object in a set of objects.

[0112] Specifically, the resource association feature with the largest information gain ratio corresponding to each resource association feature in each branch node is taken as the root node partitioning feature. After performing branching operations on the training set, the initial partitioning of the training set is completed.

[0113] Step S306: Calculate the information gain ratio corresponding to each resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node partitioning feature of each branch node, and perform branching operation on the samples in each branch node based on the node partitioning feature to generate new branch nodes.

[0114] Among them, node partitioning features refer to the resource association features that are most capable of classifying a branch node before it performs a branching operation.

[0115] Specifically, before each sample is clearly classified, each branch node in the decision tree needs to calculate the information gain ratio of each resource association feature in each branch node before performing a branch operation. Then, the resource association feature with the largest information gain ratio in each branch node is selected as the node splitting feature of the corresponding branch node. Then, the samples on the corresponding branch node are split based on the node splitting feature and a new branch node is generated.

[0116] Step S308: Repeatedly calculate the information gain ratio corresponding to each resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node splitting feature of each branch node, perform branching operation on the samples in each branch node based on the node splitting feature, and generate new branch nodes until each sample has a clear classification, and obtain the decision tree set.

[0117] The decision tree set refers to the collection of decision trees constructed during the training process.

[0118] Step S400: Predict the decision tree set using the validation set to obtain the target decision tree, and obtain the group category corresponding to each object in the object set based on the target decision tree.

[0119] Among them, the target decision tree refers to the decision tree with the strongest classification ability among all the decision trees constructed during the training process.

[0120] Specifically, the validation set is used to evaluate the various decision trees with different parameters obtained during training. Then, based on the evaluation results of the decision trees on the validation set, the decision trees are selected and their parameters are tuned so that the classification performance of the decision trees reaches the optimal level on the validation set. Finally, this decision tree is used as the target decision tree.

[0121] In this embodiment, dividing the resource association information into a training set and a validation set ensures that a decision tree model with higher accuracy can be trained. The training set can be used to train the parameters of the decision tree model, while the validation set can evaluate the performance of various trained decision tree models or the same decision tree model with different parameters. Through training based on the training set and prediction based on the validation set, the classification performance of the decision tree is improved, resulting in a better-performing target decision tree.

[0122] In one embodiment, such as Figure 4 As shown, step S204 includes:

[0123] Step S402: Obtain preset generation parameters.

[0124] Specifically, the values ​​of preset occurrence parameters can be customized based on resource data.

[0125] Step S404: Read the historical resource data, treat each historical resource data as an item, and treat each item as a candidate set of a layer, and calculate the historical occurrence parameters of each candidate set of a layer.

[0126] Specifically, each historical resource data is a record of the resources used by the group category corresponding to that historical resource data. By treating each record as an item, the probability of each record appearing is analyzed in a mathematically quantifiable way. This probability is used as the historical occurrence parameter of the historical resource data to provide data for the execution of subsequent processes.

[0127] Step S406: Select itemsets from the candidate itemsets of the first layer whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as frequent itemsets of the current layer.

[0128] Specifically, there are one or more candidate item sets in a level. Based on the idea of ​​obtaining frequently occurring historical resource data in the candidate item sets, the historical occurrence parameters of each item set in the candidate item sets under this level are compared with the preset occurrence parameters. The item sets obtained based on the comparison results are used as the level frequent item sets to provide data for the execution of subsequent processes.

[0129] Step S408: Combine the itemsets in the current layer frequent itemsets to generate candidate itemsets for the next layer, calculate the historical occurrence parameters of each candidate itemset for the next layer, and perform pre-pruning operations.

[0130] Pre-pruning refers to checking each item set in the candidate item set before filtering out frequent itemsets from the candidate item set. If a subset of the candidate item set at this level contains an item set that was removed during the generation of the candidate item set at the previous level, then the item set that was previously removed but reappears here will be removed.

[0131] Specifically, the itemsets in the frequent itemsets of the current level are arranged and combined to generate candidate itemsets for the next level. Each candidate itemset in the candidate itemsets is checked and evaluated. If an itemset that was eliminated during the generation of the candidate itemsets of the previous level exists in a subset of the candidate itemsets of this level, then that candidate itemset under the current level is eliminated.

[0132] Step S500: Select itemsets in the next-level candidate itemset after the pre-pruning operation whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as frequent itemsets in the next level.

[0133] Specifically, the candidate itemsets after the pre-pruning operation are filtered again to select the frequent itemsets that meet the requirements of that level.

[0134] Step S502: Return to the steps of combining the itemsets in the current layer frequent itemset to generate the next layer candidate itemset, calculating the historical occurrence parameter of each next layer candidate itemset, and performing pre-pruning operation, until no itemset greater than or equal to the preset occurrence parameter can be selected, and the target frequent itemset is obtained.

[0135] Specifically, before obtaining the target frequent itemset, it is necessary to continuously increase the number of itemset levels and calculate the historical occurrence parameters of candidate itemsets at each level until the target itemset is found. The specific process of the method is as follows: Figure 5 As shown, data D represents historical resource data, C_k represents candidate itemsets, L_k represents frequent itemsets, and k is a parameter used for counting. The value of k corresponds to the number of candidate itemsets or frequent itemsets, and the output rule is the target frequent itemset.

[0136] In this embodiment, the preset occurrence parameter can be used as a criterion for measuring the frequency of itemset occurrence. By progressively filtering candidate itemsets to identify frequent itemsets, it is ensured that each item is combined with others. Furthermore, during the combination and filtering process, itemsets with a condition greater than or equal to the preset occurrence parameter are selected, ensuring that the final target frequent itemsets all meet the criteria, thus laying the foundation for improving resource conversion rate.

[0137] In one embodiment, such as Figure 6 As shown, step S206 includes:

[0138] Step S504: Obtain the preset information parameters.

[0139] Specifically, the values ​​of preset information parameters can be customized based on the resource data.

[0140] Step S506: Calculate the confidence parameter of the association relationship between the items.

[0141] Specifically, each type of item association has a corresponding confidence parameter. The confidence parameter for each item association is calculated based on the probability of occurrence of each type of item association, providing a basis for the execution of subsequent processes.

[0142] Step S508: Select item associations whose confidence parameters are greater than or equal to a preset confidence parameter as target item associations corresponding to the group category.

[0143] Specifically, the target item association relationship needs to meet the condition that the confidence parameter of the item association relationship is greater than or equal to the preset confidence parameter. The confidence parameter corresponding to each item association relationship is calculated and compared with the preset confidence parameter. Item association relationships that meet the conditions of the target item association relationship are selected as the target item association relationships.

[0144] In this embodiment, the preset confidence parameter can be used as a measure of the frequency of item associations. By calculating and comparing the confidence parameter of each item association and filtering out item associations that are greater than or equal to the preset confidence parameter, the server can more intuitively understand the relationships between items, that is, the resource usage habits frequently made by each group. This can greatly improve the resource conversion rate and thus reduce the data analysis and calculation pressure on the server when sending resource data to a certain extent.

[0145] In one embodiment, such as Figure 7 As shown, after step S208, the following is also included:

[0146] Step S600: Obtain resource association information of the target object and receive the basis conditions for the terminal corresponding to the target object to display the resource data.

[0147] The target audience refers to users who log into the terminal interface, including both new and existing users. The criteria for determining where each type of resource data is displayed on the terminal after receiving resource data from the server include resource conversion rate and resource usage frequency.

[0148] Step S602: Based on the resource association information of the target object, perform feature extraction to obtain the resource association features of the target object.

[0149] Specifically, when the target object is operated on the terminal, the resource data record corresponding to the target object may change. The server needs to reprocess the resource data of the target object, extract the features of the resource association information of the target object, and use the extracted resource association features as data for subsequent process execution.

[0150] Step S604: Based on the resource association characteristics and resource association information of the target object, classify the target object into groups to obtain the target group category of the target object.

[0151] Specifically, the server categorizes the target objects into groups and uses the obtained group categories as data for further data analysis in subsequent processes.

[0152] Step S606: Based on the target group category of the target object, obtain the target item association relationship corresponding to the target group category of the target object.

[0153] Specifically, the server itself already stores the target item association relationships corresponding to various group categories. When the group category of the target object is obtained, the server can obtain the target item association relationship corresponding to the target object based on the stored data information.

[0154] Step S608: Based on the target item association relationship and the conditions, obtain the target resource data associated with the target object, and send the target resource data to the terminal corresponding to the target object.

[0155] Among them, target resource data refers to resource data that is selected from the resources owned by the target object itself, meets the preset occurrence parameters and preset confidence level, and is calculated, analyzed and sorted according to the basis conditions.

[0156] The criteria are already stored on the server. When the terminal sends the selected criteria to the server, the server further processes the resource data according to the criteria and then sends the processed resource data to the terminal.

[0157] In this embodiment, since the server already stores the target resource data corresponding to the group category, it can directly push the target resource data to the target object based on the target object's group category after obtaining the relevant data. This greatly reduces the pressure on the server from analyzing and calculating massive amounts of data. Furthermore, the server's analysis and calculation of the target resource data based on conditions makes it easier for the target object to make choices regarding the use of the target resource data, thereby improving the utilization rate of the target resource data.

[0158] In one embodiment, such as Figure 8 As shown, step S608 includes:

[0159] Step S700: Obtain the initial resource data corresponding to the target object based on the target item association relationship.

[0160] Initial resource data refers to the resource data corresponding to the target item association relationship of the target object that has not yet undergone other processing.

[0161] Step S702: When the basis condition is resource conversion rate, calculate the conversion rate of each initial resource data in the initial resource data, sort the initial resource data from largest to smallest based on the conversion rate of each initial resource data, and take the sorted initial resource data as the target resource data.

[0162] Among them, the resource conversion rate refers to the ratio of resource data to other types of resources.

[0163] Step S704: When the basis condition is the frequency of resource use, sort the confidence parameters of the target item association relationship corresponding to each initial resource data in the initial resource data from largest to smallest, and use the sorted initial resource data as the target resource data.

[0164] Among them, resource usage frequency refers to the number of times resource data is used.

[0165] In this embodiment, the non-singularity of the criteria provides the target object with more choices, improving the user experience. Furthermore, the server analyzes and calculates the initial resource data based on the criteria, recommending target resource data that meets the criteria to the target object. This facilitates the target object's selection of target resources, saves selection time, and significantly improves resource data utilization, thereby increasing the efficiency of server-side resource data delivery and reducing the pressure on the server.

[0166] In one embodiment, such as Figure 9 As shown, a resource data receiving method is provided, which is applied to... Figure 1 Taking the terminal in the example, the explanation includes the following steps:

[0167] Step S706: Obtain resource data pushed by the server. The resource data is associated with the group category corresponding to the object identifier. The resource data is determined by the server through the association relationship between the group category and the corresponding target item. The association relationship between the target item corresponding to the group category is obtained by the server by obtaining the resource association information corresponding to the object set, performing feature extraction to obtain the corresponding resource association features, and based on the resource association features, classifying each object in the object set into the corresponding group category through a decision tree. Obtain historical resource data corresponding to the group category, generate multiple candidate item sets at different levels based on the historical resource data, and filter the candidate item sets based on the comparison results of the preset occurrence parameters and the historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemsets corresponding to the group category. Arrange and combine each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group category to form the corresponding item association relationship, which is obtained based on the comparison results of the confidence parameters corresponding to the item association relationship and the preset confidence parameters.

[0168] Step S708: Display the resource data.

[0169] Specifically, the terminal receives resource data sent by the server, and this resource data is displayed on the terminal.

[0170] The aforementioned resource data receiving method, apparatus, computer equipment, and storage medium acquire resource data pushed by a server. This resource data is associated with a group category corresponding to an object identifier. The resource data is determined by the server through the association relationship between the group category and the corresponding target item. The association relationship between the target item and the group category is obtained by the server acquiring resource association information corresponding to the object set, performing feature extraction to obtain corresponding resource association features, and using a decision tree to classify each object in the object set into a corresponding group category based on the resource association features and resource association information. Historical resource data corresponding to the group category is acquired, and multiple candidate item sets at different levels are generated based on the historical resource data. The candidate item sets are then filtered based on a comparison of preset occurrence parameters and historical occurrence parameters for each different level of candidate item set to obtain the target frequent itemset corresponding to the group category. Each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category is arranged and combined to form a corresponding item association relationship. The resource data is then displayed based on a comparison of the confidence parameter corresponding to the item association relationship and a preset confidence parameter. By obtaining pushed resource data from the server based on the correlation between different resource uses, and displaying the resource data to the terminal according to the criteria, the utilization rate of resources is effectively improved, thereby reducing the pressure on the server to send resource data.

[0171] In one embodiment, the merchant's internal customer data is used as resource-related data, and the corresponding customers are considered as a set of objects. This internal customer data includes general features and specific features highly correlated with customer creditworthiness. General features include basic customer information, product holdings, transaction information, and asset-liability information. Specific features highly correlated with customer creditworthiness include borrowing frequency, borrowing amount, and installment information. This embodiment first performs feature extraction on the resource-related data to obtain resource-related features, which are the general features and specific features highly correlated with customer creditworthiness. Based on these resource-related features and the resource-related data, a decision tree algorithm is used to classify customers into groups, ultimately categorizing them into groups such as residents of emerging towns, high-net-worth clients, and students. Next, data related to the coupon usage of each customer type is selected as the customer's historical resource data. Then, based on the historical resource data, the correlation between the historical resource data (i.e., item correlation) is mined. According to the correlation, reasonable coupon combinations are issued to different groups. Based on the groups corresponding to the correlation, suitable resource data (i.e., credit coupons) are recommended to the target audience (i.e., new and old customers). This reduces the time customers spend choosing coupons, improves their utilization rate of coupons, and increases the merchant's profits.

[0172] Based on the same inventive concept, this application also provides a resource push data processing apparatus for implementing the resource push data processing method described above. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in one or more resource push data processing apparatus embodiments provided below can be found in the limitations of the resource push data processing method described above, and will not be repeated here.

[0173] In one embodiment, such as Figure 10 As shown, a resource push data processing device is provided, including: a feature extraction module 800, an object classification module 802, a target frequent itemset acquisition module 804, a target item association relationship acquisition module 806, and a data storage module 808, wherein:

[0174] The feature extraction module 800 is used to obtain resource association information corresponding to the object set and perform feature extraction to obtain the corresponding resource association features.

[0175] The object classification module 802 is used to classify each object in the object set into a corresponding group category based on the resource association features and resource association information through a decision tree.

[0176] The target frequent itemset acquisition module 804 is used to acquire historical resource data corresponding to the group category, generate multiple candidate itemsets at different levels based on the historical resource data, and filter the candidate itemsets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate itemsets to obtain the target frequent itemsets corresponding to the group category.

[0177] The target item association module 806 is used to arrange and combine each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category to form the corresponding item association. Based on the comparison result of the confidence parameter corresponding to the item association and the preset confidence parameter, the target item association corresponding to the group category is obtained.

[0178] The data storage module 808 is used to store the association between the group category and the corresponding target item, and the target item association is used to push the associated resource data to the corresponding group category.

[0179] In one example, the object classification module 802 includes: dividing the resource association features into a training set and a validation set; calculating the information gain ratio (IVR) of each resource association feature in the training set; selecting the resource association feature with the largest IVR as the root node, and performing branching operations based on the root node to generate branch nodes; calculating the IVR of each resource association feature in each branch node, selecting the resource association feature with the largest IVR as the node splitting feature of each branch node, and performing branching operations based on the node splitting feature to generate new branch nodes; repeating the process of calculating the IVR of each resource association feature in each branch node, selecting the resource association feature with the largest IVR as the node splitting feature of each branch node, and performing branching operations based on the node splitting feature to generate new branch nodes, until each resource association feature has a clear classification, thus obtaining a decision tree set; predicting the decision tree set using the validation set to obtain a target decision tree, and obtaining the group category corresponding to each object in the object set based on the target decision tree.

[0180] In one example, the target frequent itemset acquisition module 804 includes: acquiring preset occurrence parameters; reading the historical resource data, treating each historical resource data as an item, and each item as a first-level candidate itemset, calculating the historical occurrence parameters of each first-level candidate itemset; selecting itemsets from the first-level candidate itemsets whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as the current-level frequent itemsets; combining the itemsets in the current-level frequent itemsets to generate the next-level candidate itemsets, calculating the historical occurrence parameters of each next-level candidate itemset, and performing a pre-pruning operation; selecting itemsets in the next-level candidate itemsets after the pre-pruning operation whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as the next-level frequent itemsets; returning to the steps of combining the itemsets in the current-level frequent itemsets to generate the next-level candidate itemsets, calculating the historical occurrence parameters of each next-level candidate itemset, and performing a pre-pruning operation, until no itemsets greater than or equal to the preset occurrence parameters can be selected, thus obtaining the target frequent itemset.

[0181] In one instance, the target item association acquisition module 806 includes: acquiring a preset confidence parameter; calculating the confidence parameter of the item association; and filtering item associations whose confidence parameter is greater than or equal to the preset confidence parameter as target item associations corresponding to the group category.

[0182] In one example, after the data storage module 808, the system further includes: acquiring resource association information of the target object; receiving the basis conditions for the terminal corresponding to the target object to display the resource data; performing feature extraction based on the resource association information of the target object to obtain resource association features of the target object; classifying the target object into groups based on the resource association features and resource association information of the target object to obtain the target group category of the target object; acquiring the target item association relationship corresponding to the target group category of the target object based on the target item association relationship and the basis conditions; acquiring the target resource data associated with the target object based on the target item association relationship and the basis conditions, and sending the target resource data to the terminal corresponding to the target object.

[0183] In one embodiment, such as Figure 11 As shown, a resource push data processing device is provided, including: a resource data acquisition module 900 and a resource data display module 902, wherein:

[0184] The resource data acquisition module 900 is used to acquire resource data pushed by the server. The resource data is associated with a group category corresponding to an object identifier. The resource data is determined by the server through the association relationship between the group category and the corresponding target item. The association relationship between the target item corresponding to the group category is obtained by the server by acquiring the resource association information corresponding to the object set, performing feature extraction to obtain the corresponding resource association features, and based on the resource association features and resource association information, classifying each object in the object set into the corresponding group category through a decision tree, acquiring the historical resource data corresponding to the group category, generating multiple candidate item sets at different levels based on the historical resource data, filtering the candidate item sets based on the comparison results of the preset occurrence parameters and the historical occurrence parameters of each different level of candidate item sets to obtain the target frequent itemsets corresponding to the group category, and arranging and combining each non-empty proper subset corresponding to the target frequent itemsets corresponding to the group category to form the corresponding item association relationship, which is obtained based on the comparison results of the confidence parameters corresponding to the item association relationship and the preset confidence parameters.

[0185] The resource data display module 902 is used to display the resource data.

[0186] Each module in the aforementioned resource push data processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0187] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 12As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores resource data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a resource push data processing method.

[0188] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 13 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a resource push data processing method. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0189] Those skilled in the art will understand that Figure 12 and Figure 13The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0190] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method examples.

[0191] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0192] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps in the above method embodiments.

[0193] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0194] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0195] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0196] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for processing resource push data, characterized in that, The method includes: The resource association information corresponding to the object set is obtained, and the corresponding resource association features are extracted. The object set refers to the set of various social groups that use resources. The resource association information refers to the data information of each group in various social groups. The resource association features refer to the representative features selected after the resource association information is represented by features. Based on the resource association features and resource association information, each object in the object set is divided into a corresponding group category using a decision tree; The historical resource data corresponding to the group category is obtained. Based on the historical resource data, multiple candidate item sets at different levels are generated. Based on the comparison results of the preset occurrence parameters and the historical occurrence parameters of each different level of candidate item sets, the candidate item sets are filtered to obtain the target frequent itemsets corresponding to the group category. The historical resource data refers to the data records of the resources used by each object in the object set. The candidate item set refers to the set of various resource combination methods in the resources used by each object in the object set. For each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category, the corresponding item association relationship is formed by permuting and combining them. Based on the comparison results of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, the target item association relationship corresponding to the group category is obtained. The association between the group categories and the corresponding target items is stored, and the association between the target items is used to push the associated resource data to the corresponding group categories.

2. The method according to claim 1, characterized in that, The step of classifying each object in the object set into corresponding group categories using a decision tree based on the resource association features and resource association information includes: The resource association information is divided into a training set and a validation set; Calculate the information gain ratio corresponding to each of the resource association features; The resource association feature with the largest information gain ratio corresponding to each of the resource association features is selected as the root node partitioning feature. The training set is then branched based on the root node partitioning feature to generate branch nodes, and each branch node has samples from the training set. Calculate the information gain ratio corresponding to each resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node partitioning feature of each branch node, and perform branching operation on the samples in each branch node based on the node partitioning feature to generate new branch nodes. Repeatedly calculate the information gain ratio corresponding to each resource association feature in each branch node, select the resource association feature with the largest information gain ratio in each branch node as the node splitting feature of each branch node, perform branching operation on the samples in each branch node based on the node splitting feature, and generate new branch nodes until each sample has a clear classification, and obtain the decision tree set. The decision tree set is predicted using the validation set to obtain the target decision tree, and the group category corresponding to each object in the object set is obtained based on the target decision tree.

3. The method according to claim 1, characterized in that, The process involves acquiring historical resource data corresponding to the group category, generating multiple candidate item sets at different levels based on the historical resource data, and filtering the candidate item sets based on the comparison results between preset occurrence parameters and historical occurrence parameters of each different level of candidate item set to obtain the target frequent itemset corresponding to the group category, including: Obtain preset occurrence parameters; Read the historical resource data, treat each historical resource data as an item, and treat each item as a candidate set at one level, and calculate the historical occurrence parameters of each candidate set at one level; Select itemsets from the candidate itemsets at level 1 whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as frequent itemsets at the current level. Combine the itemsets in the current layer frequent itemsets to generate candidate itemsets for the next layer, calculate the historical occurrence parameters of each candidate itemset for the next layer, and perform pre-pruning operations; After the pre-pruning operation, select the itemsets in the next-level candidate itemset whose historical occurrence parameters are greater than or equal to the preset occurrence parameters as the next-level frequent itemsets; Returning to the steps of combining the itemsets in the current layer frequent itemset to generate the next layer candidate itemset, calculating the historical occurrence parameter of each next layer candidate itemset, and performing pre-pruning operations, until no itemset greater than or equal to the preset occurrence parameter can be selected, the target frequent itemset is obtained.

4. The method according to claim 1, characterized in that, The step of permuting and combining each non-empty proper subset corresponding to the target frequent itemset of the group category to form the corresponding item association relationship, and obtaining the target item association relationship corresponding to the group category based on the comparison result of the confidence parameter corresponding to the item association relationship and the preset confidence parameter, includes: Retrieve preset parameters; Calculate the confidence parameters for the association between the items; Item associations whose confidence parameters are greater than or equal to a preset confidence parameter are selected as the target item associations corresponding to the group category.

5. The method according to claim 1, characterized in that, After storing the association between the group category and the corresponding target item, the method further includes: Obtain resource association information of the target object, and receive the basis conditions for the terminal corresponding to the target object to display the resource data; Based on the resource association information of the target object, feature extraction is performed to obtain the resource association features of the target object; Based on the resource association characteristics and resource association information of the target object, the target object is classified into groups to obtain the target group category of the target object; Based on the target group category of the target object, obtain the target item association relationship corresponding to the target group category of the target object; Based on the target item association relationship and the conditions, obtain the target resource data associated with the target object, and send the target resource data to the terminal corresponding to the target object.

6. The method according to claim 5, characterized in that, The step of obtaining the target resource data associated with the target object based on the target item association relationship and the conditions includes: Based on the association relationship of the target items, obtain the initial resource data corresponding to the target object; When the basis condition is resource conversion rate, calculate the conversion rate of each initial resource data in the initial resource data, sort the initial resource data from largest to smallest based on the conversion rate of each initial resource data, and take the sorted initial resource data as the target resource data. When the basis condition is the frequency of resource use, the confidence parameters of the target item association relationship corresponding to each initial resource data in the initial resource data are sorted from largest to smallest, and the sorted initial resource data is used as the target resource data.

7. A method for receiving resource data, characterized in that, The method includes: The system acquires resource data pushed by the server, whereby the resource data is associated with a group category corresponding to an object identifier. This resource data is determined by the server through the association between the group category and the corresponding target item. The association between the group category and the target item is obtained by the server through feature extraction of resource association information corresponding to the object set. Based on these resource association features and information, a decision tree is used to categorize each object in the object set into its corresponding group category. Historical resource data corresponding to each group category is acquired. Multiple candidate option sets at different levels are generated based on this historical resource data. The candidate option sets are then evaluated based on a comparison of preset occurrence parameters and historical occurrence parameters for each different level of candidate option set. The process involves filtering to obtain the target frequent itemsets corresponding to the group categories. Each non-empty proper subset of the target frequent itemsets corresponding to the group categories is then arranged and combined to form corresponding item association relationships. This is based on the comparison between the confidence parameters corresponding to the item association relationships and the preset confidence parameters. The object set refers to the collection of various social groups using resources. The resource association information refers to the data information of each group within each social group. The resource association features refer to the representative features selected after representing the resource association information. The historical resource data refers to the data records of resource usage corresponding to each object in the object set. The candidate itemset refers to the set of various resource combinations among the resources used by each object in the object set. Display the resource data.

8. A resource push data processing device, characterized in that, The device includes: The feature extraction module is used to obtain resource association information corresponding to the object set, and to extract the corresponding resource association features. The object set refers to the set of various social groups that use resources. The resource association information refers to the data information of each group in the various social groups. The resource association features refer to the representative features selected after the resource association information is represented by features. The object classification module is used to classify each object in the object set into a corresponding group category based on the resource association features and resource association information using a decision tree. The target frequent itemset acquisition module is used to acquire historical resource data corresponding to the group category, generate multiple candidate itemsets at different levels based on the historical resource data, and filter the candidate itemsets based on the comparison results of preset occurrence parameters and historical occurrence parameters of each different level of candidate itemsets to obtain the target frequent itemsets corresponding to the group category. The historical resource data refers to the data records of the resources used by each object in the object set, and the candidate itemset refers to the set of various resource combinations in the resources used by each object in the object set. The target item association module is used to arrange and combine each non-empty proper subset corresponding to the target frequent itemset corresponding to the group category to form the corresponding item association. Based on the comparison result between the confidence parameter corresponding to the item association and the preset confidence parameter, the target item association corresponding to the group category is obtained. The data storage module is used to store the association between the group categories and the corresponding target items. The association between the target items is used to push the associated resource data to the corresponding group categories.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images