Resource data processing method and device, computer equipment and readable storage medium

Abstract

The invention relates to a resource data processing method and device, computer equipment and a readable storage medium. The method comprises the following steps: acquiring user data corresponding to a user applying for a target resource in a resource period; obtaining service resource processing information corresponding to the target resource; processing the user data based on the service resource processing information to obtain application data corresponding to the target resource applied by the user in the resource period; inputting the application data into an abnormal behavior recognition model, analyzing trend change information corresponding to the application data based on the abnormal behavior recognition model, and recognizing application behavior data corresponding to the trend change information; and if the application behavior data corresponds to the abnormal behavior, obtaining adjustment information for the service resource processing information based on the trend change information, and updating the service resource processing information. By adopting the method, the service resource processing information can be updated in time to improve the accuracy of detecting the abnormal resource acquisition means, so that the security and reliability of resource data processing are improved.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and particularly to a method and apparatus for processing resource data, a computer device, a computer-readable storage medium, and a computer program product. Background Art

[0002] With the development of computer data processing technology, data is usually processed by a computer to improve the efficiency of data processing and optimize the process of manual data processing. Exemplarily, work related to resource applications can be processed by a computer.

[0003] In today's digital age, financial transactions are becoming increasingly frequent, and financial frauds are emerging in an endless stream. When using a computer to process work related to resource applications, there may be abnormal behavior data for resource applications. However, in the prior art, such abnormal behavior data cannot be identified in a timely manner, thereby reducing the security, reliability, etc. of the processing work for resource applications. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide a method and apparatus for processing resource data, a computer device, a computer-readable storage medium, and a computer program product that can update business resource processing information in a timely manner to improve the accuracy of detecting abnormal resource acquisition means, thereby improving the security and reliability of resource data processing.

[0005] In a first aspect, the present application provides a method for processing resource data, including:

[0006] Obtaining user data corresponding to a user who applies for a target resource within a resource period;

[0007] Obtaining business resource processing information corresponding to the target resource;

[0008] Processing the user data based on the business resource processing information to obtain application data corresponding to the user's application for the target resource within the resource period;

[0009] Inputting the application data into an abnormal behavior recognition model to analyze trend change information corresponding to the application data based on the abnormal behavior recognition model, and identifying application behavior data corresponding to the trend change information;

[0010] If the application behavior data corresponds to an abnormal behavior, obtaining adjustment information for the business resource processing information based on the trend change information;

[0011] Updating the business resource processing information based on the adjustment information to perform resource data processing based on the updated business resource processing information.

[0012] In a second aspect, the present application further provides an apparatus for processing resource data, including:

[0013] A data acquisition module, configured to acquire user data corresponding to users who apply for a target resource within a resource cycle;

[0014] A service information acquisition module, configured to acquire service resource processing information corresponding to the target resource;

[0015] An application processing module, configured to process the user data based on the service resource processing information to obtain application data corresponding to the user's application for the target resource within the resource cycle;

[0016] A model processing module, configured to input the application data into an abnormal behavior recognition model, analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model, and identify the application behavior data corresponding to the trend change information;

[0017] An adjustment information generation module, configured to, if the application behavior data corresponds to an abnormal behavior, obtain adjustment information for the service resource processing information based on the trend change information;

[0018] An update module, configured to update the service resource processing information based on the adjustment information, so as to perform resource data processing based on the updated service resource processing information.

[0019] In a third aspect, the present application further provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the steps described in the above method are implemented.

[0020] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps described in the above method are implemented.

[0021] In a fifth aspect, the present application further provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps described in the above method are implemented.

[0022] The above resource data processing method, device, computer equipment, computer-readable storage medium and computer program product obtain user data corresponding to users who apply for target resources within a resource cycle, and obtain business resource processing information corresponding to the target resources. The business resource processing information is used to determine whether the user data can successfully apply for the target resources. Therefore, the user data is processed based on the business resource processing information to obtain application data corresponding to the user's application for the target resources within the resource cycle. In this way, the result of whether the user can successfully apply for the target resources can be obtained. Since users may adopt some abnormal means to obtain the target resources, in this application, in order to successfully identify these abnormal means and adapt to the update of these abnormal means, the application data needs to be input into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model. In this way, it can be monitored whether there is abnormal data in the data for applying to obtain the target resources, and these abnormal data are analyzed through the abnormal behavior recognition model to identify the application behavior data corresponding to these abnormal data. If the application behavior data corresponds to abnormal behavior, that is, there is a user who has successfully obtained the target resources through these abnormal data, which is actually not allowed, then the business resource processing information needs to be updated. In this way, these abnormal data can be identified based on the updated business resource processing information, thereby reducing the probability of successfully obtaining the target resources through the use of abnormal means, and being able to timely identify the changed abnormal means, effectively preventing the abnormal acquisition of the target resources, and further improving the security verification of the application behavior for obtaining the target resources, and improving the reliability and robustness of processing the resource data. Brief Description of the Drawings

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for describing the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0024] Figure 1 It is an application environment diagram of the resource data processing method in an embodiment;

[0025] Figure 2 It is a flowchart of the resource data processing method in an embodiment;

[0026] Figure 3 It is a schematic diagram of business resource processing information in an embodiment;

[0027] Figure 4 It is a flowchart of processing user data based on business resource processing information in an embodiment;

[0028] Figure 5 It is a schematic flow chart of data pre - processing in an embodiment;

[0029] Figure 6 It is a schematic flow chart of model training in an embodiment;

[0030] Figure 7 It is a structural block diagram of a resource data processing device in an embodiment;

[0031] Figure 8 It is an internal structure diagram of a computer device in an embodiment. Specific embodiments

[0032] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0033] The resource data processing method provided by the embodiments of the present application can be applied to, for example Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or can be placed on the cloud or other network servers. The user sends a request to apply for a target resource to the server 104 through the terminal 102. In response to this request, the server 104 processes the user data corresponding to the user who applies for the target resource, and obtains the application data corresponding to the user's application for the target resource through the obtained service resource processing information corresponding to the target resource, so as to determine whether the user can successfully apply for the target resource. To successfully identify the behavior of successfully applying for a target resource through abnormal data, the server 104 can obtain the application data corresponding to the user's application for the target resource within the resource cycle. Input the application data into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model, and identify the application behavior data corresponding to the trend change information. If it is determined through the abnormal behavior recognition model that the application behavior data corresponds to an abnormal behavior, then obtain the adjustment information for the service resource processing information based on the trend change information. Update the service resource processing information based on the adjustment information, and perform resource data processing based on the updated service resource processing information. In this way, the behavior of abnormally applying for a target resource can be successfully intercepted based on the updated service resource processing information, thereby effectively preventing the abnormal acquisition of the target resource. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0034] In an exemplary embodiment, as Figure 2 shown, a resource data processing method is provided. Taking the method applied to Figure 1 the server 104 in

[0035] Step 202, obtain the user data corresponding to the user who applies for the target resource within the resource cycle.

[0036] Among them, the resource cycle refers to a time period. The resource cycle corresponds to the verification cycle of business resource processing information. For example, if the verification cycle is two days, then every two days, the abnormal behavior recognition model can be used to analyze the user data obtained within the verification cycle and determine whether the current business resource processing information can successfully intercept the abnormal behavior of applying for the target resource. Abnormal behavior refers to the behavior of applying for the target resource with user data that does not meet the preset conditions. If the abnormal behavior of applying for the target resource cannot be successfully intercepted, the business resource processing data needs to be updated so that the application resource result corresponding to applying for the target resource using the abnormal behavior is an application failure, that is, the target resource cannot be successfully applied for using the abnormal behavior.

[0037] Among them, the target resource is an entity or abstraction that can be allocated to the target user. For example, the target resource can be tickets, gift cards, credit products, insurance products, etc. The target user is the user who meets the preset conditions determined from multiple users by processing the user data through the business resource processing information. The target user can be an individual or other organizations such as enterprises.

[0038] Among them, the user data includes first user data. The first user data is the behavior data of the user on the terminal side. For example, the first user data can be the user's login time, login frequency, historical operation information of the user on the terminal side, etc. The user data can also include second user data. The second user data is the data related to the user obtained from a third-party data source. For example, the second user data can be credit data, invoice data, tax data, operation data in a third-party system, etc. Optionally, the user data can also include the data input by the user to the terminal when applying for the target resource.

[0039] Exemplarily, the user data can also be the data obtained after processing the original data of the collected user. After obtaining the original data corresponding to the user who applies for the target resource during the resource acquisition period, the original data is processed to obtain the user data. The user data can also include real-time behavior data and statistical behavior data. Real-time behavior data refers to the set of behavior-related information obtained at millisecond-to-second time intervals through real-time acquisition, transmission, and processing mechanisms during the process of the subject (such as the user) performing various activities. That is to say, the real-time behavior data is the data obtained by analyzing the original data in real time. For example, it is real-time click behavior data (such as click frequency, click time interval, etc.), real-time browsing behavior data (such as browsing time, etc.), and real-time form filling behavior data (such as data required for filling, filling range, etc.). Statistical behavior data is the data obtained by analyzing a batch of original data. For example, the target type data in the batch of original data corresponding to the resource acquisition period is obtained, and the distribution level of each target type data with respect to the overall target type data is statistically analyzed to obtain the corresponding statistical behavior data. Taking the target type data as the monthly consumption ability as an example, the statistical behavior data is, for example, the monthly consumption ability data, and the user data corresponding to each user includes the distribution level of the monthly consumption ability corresponding to the original data with respect to the overall original data corresponding to the resource period. For example, the monthly consumption ability of a certain user (such as 18,999 yuan) is at the second level. The overall original data corresponds to a first level (0 yuan - 10,000 yuan), a second level (10,000 yuan - 20,000 yuan), and a third level (20,000 yuan - 30,000 yuan).

[0040] Step 204, obtain the service resource processing information corresponding to the target resource.

[0041] Step 206, process the user data based on the service resource processing information to obtain the application data corresponding to the user's application for the target resource during the resource period.

[0042] Exemplarily, each target resource can be configured with corresponding service resource processing information. The application data includes the application resource result. The application resource result is the data corresponding to whether the user successfully applies for the target resource. The application resource result can be a quantitative value (such as 60 points, 90 points), or it can be a qualitative data (such as success, failure).

[0043] Exemplarily, the service resource processing information includes at least one processing parameter and each sub-processing information corresponding to each processing parameter.

[0044] Among them, the parameter value of the processing parameter is obtained based on user data. When the user data corresponds to multiple types of feature data, the parameter value of the processing parameter can be obtained by fusing multiple feature data. Optionally, the parameter value of the processing parameter can be obtained by calculating multiple feature data based on the configuration rule corresponding to the processing parameter. Specifically, the parameter value of the processing parameter can be obtained based on the product of multiple feature data. Or the parameter value of the processing parameter can also be obtained based on the difference between multiple feature data. For example, if the processing parameter is the revenue amount, the difference between the feature data "cost" and the feature data "income" needs to be obtained to obtain the parameter value of the user data corresponding to the "revenue amount". Another example is that when the processing parameter is the potential consumption ability, the product value between the feature data "age" and the feature data "income" can be obtained to obtain the parameter value of the user data corresponding to the "potential consumption ability".

[0045] In some other embodiments, the parameter value of the processing parameter can also be determined based on the target feature data. Exemplarily, the parameter value of the processing parameter can be determined based on the target feature data and the statistical feature data corresponding to the target feature data. Specifically, the target feature data is the transaction amount, and the statistical feature data can be the average value corresponding to the transaction amount of each user data within the resource cycle. The parameter value of the processing parameter can then be determined based on the ratio of the transaction amount corresponding to each user data to the average value. It can be understood that the parameter values of some processing parameters can also be directly determined by the target feature data.

[0046] Among them, the sub-processing information is used to determine the association relationship between the corresponding processing parameter and the application resource result. Exemplarily, each processing parameter is configured with a corresponding configuration weight. That is, each sub-processing information is configured with a corresponding configuration weight. For a user data, obtain the sub-processing results corresponding to each sub-processing information of the service resource processing information, and obtain the application resource result corresponding to the user data based on the weighted sum of each sub-processing result and the matching configuration weight. Optionally, if the application resource result is "application successful" or "application failed", the weighted sum of each sub-processing result and the matching configuration weight can be compared with the preset application threshold. If the sum is greater than or equal to the preset application threshold, it can be determined that the application resource result is "application successful". If the sum is less than the preset application threshold, it can be determined that the application resource result is "application failed". Specifically, the sub-processing result can be the parameter value corresponding to the processing parameter.

[0047] In some embodiments, the parameter values of each processing parameter can be configured with corresponding rank weights according to a data range. In this case, for a piece of user data, the parameter values of the user data for each processing parameter are obtained, and the rank weights corresponding to the parameter values are determined. The application resource result corresponding to the user data is obtained based on the weighted sum of the configured weights and the matching rank weights corresponding to each processing parameter. That is to say, the target abnormal feature corresponding to the trend change information can be obtained; the rank weight corresponding to the target abnormal feature in the business resource processing information can be obtained; and the application resource result data corresponding to the user data for the business resource processing information can be obtained.

[0048] It can be understood that if the parameter values of the processing parameter are not configured with corresponding rank weights according to the data range, the application resource result corresponding to the processing parameter can be directly obtained based on the product of the parameter value and the matching configured weight.

[0049] During the process of a user applying for a target resource, due to the user's exploration of the business resource processing information, it may lead to the user successfully applying for the target resource through abnormal means. However, the risk of allocating the target resource to such users is relatively high. Therefore, these abnormal means need to be identified to prevent such users from successfully applying for the target resource. Specifically, the following content can be referred to:

[0050] Step 208: Input the application data into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model and identify the application behavior data corresponding to the trend change information.

[0051] Among them, the abnormal behavior recognition model determines the data related to abnormal behavior in these user data by analyzing the user data.

[0052] Exemplarily, the application data includes service processing data. The service processing data may be the parameter values of the above-mentioned processing parameters. After the service processing data in the application data is input into the abnormal behavior recognition model, the abnormal behavior recognition model can analyze the service processing data in the application data and monitor the trend change information corresponding to the service processing data within the resource cycle. For example, it may be the trend change information between the service processing data in the currently obtained resource cycle and the historical service processing data input into the abnormal behavior recognition model before the resource cycle. When the user data changes, it is very likely that there is an event of abnormally obtaining the target resource. Therefore, it is necessary to analyze various aspects of data through the abnormal behavior recognition model to detect whether the trend of such data changes is a normal application behavior. That is to say, in this embodiment, in order to increase the probability that the service resource processing information can successfully block the application behavior of abnormally obtaining the target resource and cope with the changes in the application behavior of abnormally obtaining the target resource, the application data is input into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model and identify the application behavior data corresponding to the trend change information. If the trend change information corresponding to the application data corresponds to a normal behavior, that is, it is detected that the trend change information has nothing to do with the application behavior of abnormally obtaining the target resource, and the application resource result corresponding to the application data corresponds to an application success, then the service resource processing information may not be adjusted. If it is determined that the trend change information corresponding to the application data corresponds to a normal behavior, but the application resource result corresponding to the application data corresponds to an application failure, then the first adjustment information for the service resource processing information can be obtained based on the trend change information. The first adjustment information is different from the adjustment information in step 210.

[0053] If, after analysis by the abnormal behavior recognition model, it is recognized that the trend change information corresponding to the application data corresponds to an abnormal behavior, the following steps can be executed:

[0054] Step 210, if the application behavior data corresponds to an abnormal behavior, obtain the adjustment information for the service resource processing information based on the trend change information.

[0055] Among them, the abnormal behavior refers to the application behavior of abnormally obtaining the target resource. For example, applying for the target resource with improper operation behavior (such as obtaining the target resource multiple times by changing the IP address when there are restrictions on the target resource allocation), or applying for the target resource by forging user data.

[0056] Among them, the adjustment information can be used to adjust the configuration weight corresponding to the target sub-processing information in the service resource processing information. The target sub-processing information corresponds to the trend change information. The adjustment information can also be used to adjust the data range corresponding to the sub-processing parameters matched by the target sub-processing information.

[0057] Step 212: Update the service resource processing information based on the adjustment information, so as to process resource data based on the updated service resource processing information.

[0058] Exemplarily, update the configuration weight corresponding to the target sub - processing information in the service resource processing information based on the adjustment information. Or update the data range corresponding to the sub - processing parameters matched by the target sub - processing information based on the adjustment information. The application resource result of the application data obtained after processing the user data based on the updated service resource processing information is consistent with the target application result. Among them, the target application result corresponds to "application failed".

[0059] In the above - mentioned resource data processing method, by obtaining the user data of the user who applies for the target resource within the resource cycle, and obtaining the service resource processing information corresponding to the target resource, the service resource processing information is used to determine whether the user data can successfully apply for the target resource. Therefore, the user data is processed based on the service resource processing information to obtain the application data corresponding to the user's application for the target resource within the resource cycle. In this way, the result of whether the user can successfully apply for the target resource can be obtained. Since users may adopt some abnormal means to obtain the target resource, in this application, in order to successfully identify these abnormal means and adapt to the updates of these abnormal means, the application data needs to be input into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on this abnormal behavior recognition model. In this way, it can be monitored whether there is abnormal data in the data of applying for the target resource, and these abnormal data are analyzed through the abnormal behavior recognition model to identify the application behavior data corresponding to these abnormal data. If the application behavior data corresponds to abnormal behavior, that is, there is a user who has successfully obtained the target resource through these abnormal data, which is actually not allowed, then the service resource processing information needs to be updated. In this way, these abnormal data can be identified based on the updated service resource processing information, thereby reducing the probability of successfully obtaining the target resource through the use of abnormal means, and being able to promptly identify the changed abnormal means, effectively preventing the abnormal acquisition of the target resource, and further improving the security verification of the application behavior of obtaining the target resource, and improving the reliability and robustness of processing resource data.

[0060] In this case, through the combination of big data and AI model technology, a series of operation processes of resource data are carried out in the credit anti - fraud system, improving the reliability and robustness of data analysis and processing.

[0061] In an exemplary embodiment, the application resource result may be application resource result data. The application resource result data is quantitative data. For example, the user data is processed based on the business resource processing information to obtain the corresponding application resource result data. If the application resource result data is greater than or equal to a preset value, the application resource result data corresponds to "application successful"; if the application resource result data is less than the preset value, the application resource result data corresponds to "application failed". Therefore, if it is recognized that a certain trend change information corresponds to an abnormal behavior, and the user data corresponding to the trend change information determined based on the current business resource processing information corresponds to a normal behavior, then the business resource processing information needs to be adjusted so that the application resource result data obtained by processing the user data based on the adjusted business resource processing information corresponds to an abnormal behavior. In some embodiments, the application resource result data may be updated by adjusting the configuration weight. Optionally, the application resource result data may also be updated by adjusting the level weight. Wherein, step 210 includes steps 302 to 308. Among them:

[0062] Step 302, obtain the target abnormal feature corresponding to the trend change information.

[0063] Step 304, obtain the reference value corresponding to the target abnormal feature in the business resource processing information, and the reference value is used to determine the application resource result in the application data.

[0064] Among them, the business resource processing information is used to associate each data feature with the corresponding reference value.

[0065] Among them, there are multiple feature data corresponding to the business resource processing information in the user data. If a certain feature data changes within the resource cycle, the corresponding trend change information of the feature data can be obtained. If the trend change information is recognized as an abnormal behavior through the abnormal behavior recognition model, it is determined that the feature data corresponding to the trend change information is the data corresponding to the target abnormal feature, that is, the target abnormal feature can be determined. For example, the target abnormal features are "login time", "tax amount", "login address", etc.

[0066] As can be seen from the above, in the business resource processing information, each sub - processing information is used to analyze the corresponding processing parameters. The application data input into the abnormal behavior recognition model includes the parameter values corresponding to each processing parameter. The abnormal behavior recognition model determines whether the trend change information corresponding to these parameter values corresponds to an abnormal behavior by analyzing the trend change information. Therefore, after obtaining the target abnormal feature corresponding to the trend change information, the sub - processing information corresponding to the target abnormal feature in the business resource processing information can be determined. For example, determine the sub - processing parameter in the business resource processing information that is consistent with the target abnormal feature, and the sub - processing information corresponding to this sub - processing parameter is the sub - processing information corresponding to the target abnormal feature.

[0067] After determining the sub - processing information corresponding to the target abnormal feature in the business resource processing information, the reference value corresponding to the target abnormal feature in the business resource processing information can be obtained.

[0068] Exemplarily, the reference value can also be the grade weight corresponding to the target abnormal feature in the business resource processing information. As Figure 3 shown, the grade weight is Figure 3 the tax grade shown in

[0069] There are multiple distribution ranges for the parameter values of the sub - processing parameter corresponding to the target abnormal feature. Each distribution range is configured with a corresponding grade weight. The target abnormal feature corresponds to at least one of the distribution ranges. For example, the target abnormal feature corresponds to "tax amount less than 1 million", and the corresponding grade weight is E (E is less than D, D is less than C,...). In this way, the reference value of the target abnormal feature in the business resource processing information can be obtained through the grade weight corresponding to the target abnormal feature in the business resource processing information.

[0070] Optionally, the reference value corresponding to the target abnormal feature in the business resource processing information can also be the data range corresponding to the target abnormal feature. For example, the feature data belonging to this data range is successfully applied, and the feature data not belonging to this data range is rejected from applying for the target resource. Then, the data range corresponding to the target abnormal feature can also be obtained, and the corresponding data range can be narrowed based on the trend change information.

[0071] Step 306: Obtain the application resource result data corresponding to the user data for the business resource processing information.

[0072] Step 308: Update the reference value based on the difference between the application resource result data and the target application resource result data to obtain the corresponding target reference value.

[0073] Exemplarily, in order to prevent a user from successfully applying for the target resource using abnormal data, the application resource result data corresponding to the abnormal data can be reduced so that the application resource result data is less than a preset value.

[0074] Specifically, the application resource result data corresponding to the target abnormal feature data corresponding to the trend change information can be obtained. Based on the difference between the application resource result data and the target application resource result data, the reference value corresponding to the target abnormal feature data is updated to obtain the target reference value. The target reference value can replace the reference value in the business resource processing information to update the business resource processing information. Processing the user data corresponding to the target abnormal feature data using the updated business resource processing information, the obtained application resource result data is less than the preset value, so that it can be ensured that the user cannot successfully apply for the target resource.

[0075] In this embodiment, data is mined through a deep learning algorithm, and the target abnormal feature corresponding to the trend change information is obtained; using the feature engineering technology, the reference value corresponding to the target abnormal feature in the business resource processing information is obtained, and the reference value is used to determine the application resource result in the application data, and the business resource processing information is used to associate each data feature with the corresponding reference value; obtaining the application resource result data corresponding to the user data for the business resource processing information; calculating the difference between the application resource result data and the target application resource result data based on the loss function in machine learning, and updating the reference value to obtain the corresponding target reference value. In this way, the reference value corresponding to the abnormal data feature in the business resource processing information can be accurately adjusted according to specific values, reducing the possibility of over-adjustment, so as to ensure the security of resource data processing and the accuracy of resource data.

[0076] In some other embodiments, the level weight can also be updated to the target level weight based on the difference between the application resource result data and the target application resource result data, so that the difference between the target application resource result data and the application resource result data corresponding to the target level weight meets the preset conditions. Specifically, the business resource processing information may include login time processing information, and different login times correspond to different level weights. When the login time within the resource cycle changes, for example, the number of logged-in users increases during a certain period, resulting in an increased probability of abnormal acquisition of target resources, the reference value corresponding to the login time processing information can be adjusted through the following steps: obtaining the abnormal login time corresponding to the trend change information; obtaining the level weight corresponding to the abnormal login time in the login time processing information; obtaining the user data corresponding to the abnormal login time for the application resource result data corresponding to the login time processing information; updating the level weight to the target weight based on the difference between the application resource result data and the target application resource result data, so that the difference between the target application resource result data and the application resource result data corresponding to the target weight meets the preset conditions.

[0077] Among them, the target application resource result data can be the above-mentioned preset value. The difference meeting the preset conditions can mean that the difference is greater than or equal to zero.

[0078] In this embodiment, by identifying the abnormal trend of the login time, users who apply for target resources during abnormal time periods can be successfully intercepted. For example, the probability that a user who applies for a target resource late at night returns the target resource after successfully applying for it is relatively low. When the user changes the login time, the abnormality of the login time can also be detected in a timely manner, thereby improving the security and robustness of resource data processing, and then the abnormal application means of the user can be learned in a timely manner and corresponding measures can be taken.

[0079] In some embodiments, after inputting the application data obtained within the resource cycle into the abnormal behavior recognition model, the abnormal behavior recognition model can also analyze the trend change information corresponding to the application resource result. Exemplarily, the ratio between the first application resource result and the second application resource result is obtained. If the ratio is not within the preset range, the first business processing data corresponding to the first application resource result and the second business processing data corresponding to the second application resource result are obtained, and whether there is abnormal feature data in the first business processing data and the second business processing data is analyzed through the model. If so, the business resource processing information can be adjusted using the above steps 302 to 308. In this way, the dynamic changes of the application resource result can be monitored. When the distribution of the application resource result does not conform to the general law and is abnormal, the reasons can be analyzed and corresponding adjustments can be made.

[0080] In some embodiments, after obtaining the application data, an application report corresponding to the application data may also be generated. The application report is used to display the above-mentioned business processing data and application resource results. The template of the application report may be pre-configured. After obtaining the application data, the application data is filled into the template to obtain the final application report. Through the application report, it is convenient to intuitively display the application results and the corresponding business processing data, improving the readability of the application results.

[0081] In some embodiments, as Figure 4 shown, in the process of processing user data based on business resource processing information to obtain application data corresponding to the user's application for target resources within the resource cycle, performance information can be obtained to detect the system operation status during the process of processing user data. For example, abnormal CPU / memory consumption, throughput fluctuations, processing failures, etc. It is also possible to detect the business processing data to obtain the business processing data to be audited. For example, values outside the normal range or abnormal values. For example, if the user's tax data is 0, this is an abnormal value. Or if the user is a small and micro enterprise, but the tax amount exceeds the preset amount range for small and micro enterprises, the tax amount is also determined to be an abnormal value. For the business processing data to be audited, the corresponding user data can be sent to a third terminal to manually process the user data corresponding to the business processing data to be audited through the third terminal.

[0082] In an exemplary embodiment, the obtained user data is multi-dimensional data after processing the original data. Exemplarily, the original data stored by the user in a third-party server is obtained. Referring to Figure 5 , after obtaining the user's authorization, the user's invoice data, tax data, industrial and commercial data, credit data, etc. can be obtained. To improve the processing efficiency of the original data and perform matching processing methods for different types of data, the original data can be pushed to the corresponding message queue to process the original data. After processing the original data through the message queue, the obtained queue processing data can be stored in the matching storage space according to the data type. Specifically, the data is classified into different databases or file systems according to the data structure type. For example, structured invoice data falls into the Mysql database; documented industrial and commercial and judicial data falls into the Mongodb database. After that, the user data corresponding to the user can be obtained from the storage space.

[0083] Among them, in the process of processing the original data through the message queue, the data in the message queue can be pushed to the consumption application. The consumption application performs preliminary processing of the data on the message to process the data format. For example, checking whether the data format is normal and whether the data is correct. In the case of missing data, the missing data can be filled based on the average value corresponding to the data.

[0084] Optionally, it is also possible to obtain the queue processing data in the storage space and process the queue processing data according to a preset processing method to obtain multi-dimensional user data. Among them, the multi-dimensional user data can be the above-mentioned real-time behavior data and statistical behavior data. Specifically, the synchronization of data sources (such as user data stored in each storage space) uses oplog for real-time synchronization or scheduled batch synchronization. It can be understood that during the synchronization process, unified standardization processing also needs to be completed, such as unifying data units, filtering illegal characters, data encoding, etc. Then the data is aggregated to the metadata layer of the data warehouse. After that, the user data of the preset dimension (such as real-time behavior data, statistical behavior data) can be obtained by processing the data in the metadata layer. Specifically, the data stream of the data warehouse source data layer is accessed through Kafka, database CDC, etc., and at the same time, batch data sources such as HDFS are regularly imported. The stream processing task uses Flink for real-time calculation to generate second-level / minute-level behavior tags (i.e., real-time behavior data, such as real-time click heatmaps), and the intermediate results are temporarily stored in Redis. The batch processing task processes historical data through Spark to produce statistical behavior data (such as monthly consumption ability grading). After the two types of tags pass the quality verification, they are uniformly stored in the data warehouse.

[0085] In some embodiments, after obtaining the user data, the user data can be structurally processed to classify the user data according to data attributes. In this way, the user data with consistent attributes can be grouped into one category, which is convenient for subsequent correspondence between the user data of each attribute and the corresponding business resource processing information, so as to specifically process the data of each attribute. For example: there are 4 types of user data (age, gender, number of houses, number of vehicles), and the user data related to personal characteristics can be grouped into one category, and the user data related to personal property can be grouped into another category, resulting in the following structure: {"Personal_Features":{"age":45,"gender":"man"},"Personal_Assets":{"car_cnt":3,"house_cnt":4}}.

[0086] In this way, it can not only improve the efficiency of processing user data based on business resource processing information, but also improve the degree of fit between the subsequent data and the business resource processing information by classifying the data.

[0087] In this embodiment, by processing the data, multi-dimensional user data can be obtained. In this way, potential features, statistical features, etc. of the user data can be obtained. Furthermore, when analyzing the user data subsequently, multi-dimensional and multi-level user data can be considered, so as to evaluate from multiple aspects whether the user can apply for the target resource, thereby improving the accuracy and rationality of the application result.

[0088] In an exemplary embodiment, when there is a change in business processing data, the abnormal behavior recognition model can trigger behavior analysis of the change, such as whether it is a normal behavior or an abnormal behavior. Exemplarily, the application data includes business processing data, and the application data is input into the abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model, and identify the application behavior data corresponding to the trend change information, including: inputting the application data into the abnormal behavior recognition model to obtain the data distribution information corresponding to the business processing data based on the abnormal behavior recognition model, comparing the data distribution information with the historical data distribution information corresponding to the business processing data to obtain the first change information corresponding to the business processing data, and if the first change information corresponds to exceeding the target threshold, identifying the application behavior data corresponding to the first change information.

[0089] Among them, the data distribution information corresponding to the business processing data can refer to the data range corresponding to the business processing data, or the distribution information corresponding to the business processing data (such as the number of users corresponding to each login time period). Among them, the historical data distribution information can be the training data of the model, or the historical data distribution information corresponding to the application data input into the model before the current resource cycle. Then, the data distribution information can be compared with the historical data distribution information to obtain the first change information corresponding to the business processing data. For example, the number of users in the login time period from 2 o'clock to 3 o'clock has increased by 10%. Considering that the data distribution fluctuates above and below the baseline, when the first change information exceeds the target threshold, the application behavior data corresponding to the first change information can be analyzed through the abnormal behavior recognition model.

[0090] In this embodiment, by analyzing whether the trend change information generated in the user data by the abnormal behavior recognition model is an abnormal behavior, the abnormal behavior of the user can be monitored in a timely manner, and then effective prevention measures can be executed in a timely manner.

[0091] In some embodiments, before determining that the application behavior data corresponding to the business processing data is an abnormal behavior if the first change information corresponds to exceeding the target threshold, the method further includes: analyzing the historical data distribution information corresponding to the business processing data based on the abnormal behavior recognition model to predict the predicted data distribution information corresponding to the business processing data within the resource cycle; obtaining the difference information between the predicted data distribution information and the historical data distribution information; determining the target threshold based on the difference information.

[0092] Exemplarily, the abnormal behavior recognition model analyzes the historical data distribution information corresponding to the business processing data, and can predict the predicted data distribution information corresponding to the business processing data within the resource cycle. For example, the number of logged-in users from Monday to Friday is about 100, and the number of logged-in users from Saturday to Sunday will rise to about 199. If the current resource cycle corresponds to the promotion period, the number of logged-in users from Saturday to Sunday will rise to about 399. That is, it rises from the historical data of 199 to about 399. If the application data from Saturday to Sunday is input into the abnormal behavior recognition model, the predicted data distribution information corresponding to the number of logged-in users generated is about 599. Since the difference between this data and the historical data exceeds the predicted difference, there may be a certain probability of abnormal acquisition of the target resource.

[0093] In this embodiment, by predicting the data distribution information through the abnormal behavior recognition model, the predicted data distribution information corresponding to various scenarios can be generated. That is, the target threshold can adapt to changes, thus avoiding misjudging the trend change of a certain business processing data as an abnormal behavior, and improving the reliability of resource data processing.

[0094] In some exemplary embodiments, the model needs to be updated regularly to adapt to changing abnormal application means or to adapt to changes in some external information. Therefore, the method of this application further includes: obtaining an external data source and application resource training data corresponding to the external data source, where the application resource training data corresponds to actual application result data; extracting features from the external data source to obtain external data features; extracting features from the application resource training data to obtain application data features; obtaining the association relationship between the external data features and the application data features based on the actual application result data; predicting the predicted application result data corresponding to the application resource training data based on the association relationship, and training the abnormal behavior recognition model based on the difference information between the predicted application result data and the actual application result data to update the abnormal behavior recognition model.

[0095] Among them, the external data source may be a standard application rule related to the target resource, such as an industry standard rule promulgated by a certain organization. The application resource training data corresponding to the external data source is application resource training data that matches the external data source. The application resource training data includes business processing data and updated application resource results (i.e., actual application result data). The updated application resource results match the external data source. After inputting the external data source and the corresponding application resource training data into the abnormal behavior recognition model, the model can be retrained based on the external data source and the corresponding application resource training data.

[0096] Specifically, feature extraction can be performed on an external data source to obtain external data features. Feature extraction is performed on the application resource training data to obtain application data features. Based on the actual application result data, the association relationship between the external data features and the application data features is obtained. Based on the association relationship, the predicted application result data corresponding to the application resource training data is predicted. Based on the difference information between the predicted application result data and the actual application result data, the abnormal behavior recognition model is trained to update the abnormal behavior recognition model.

[0097] In this embodiment, the model can adjust the parameters of the model by mining the association relationship between the external data features and the application data features, and thus can generate a model that matches the external data source, and can adapt to the changes of the external environment in a timely manner.

[0098] In some embodiments, to improve the recognition ability of the abnormal behavior recognition model, the abnormal behavior recognition model can be evaluated. Exemplarily, the complexity evaluation result of the updated abnormal behavior recognition model is obtained; if the complexity evaluation result is higher than or equal to a preset threshold, the first correlation degree between each external data feature and other external data features and the second correlation degree between each application data feature and other application data features are obtained. If the first correlation degree meets the first preset condition, at least one application data feature corresponding to the first correlation degree is deleted. If the second correlation degree meets the second preset condition, at least one application data feature corresponding to the second correlation degree is deleted; if the complexity evaluation result is lower than the preset threshold, the fused external data features are generated based on each external data feature, so as to perform feature extraction on the external data source based on each external data feature and the fused external data features, and / or the fused application data features are generated based on each application data feature, so as to perform feature extraction on the application resource training data based on each application data feature and the fused application data features.

[0099] That is to say, when the model complexity is too low, new features can be generated based on the existing feature fusion. Specifically, features highly relevant to the application resource result can be added. For example, feature interaction terms can be created (such as the product of age and income as a new feature). When the model complexity is too high, redundant features or some highly correlated features are deleted, such as age and age groups (youth, middle age, old age). Among them, redundant features or some highly correlated features can be determined by obtaining the correlation degree between each feature and other features. If the correlation degree between certain features is too high, these features are determined as redundant features and can be appropriately deleted.

[0100] In this embodiment, by deleting or adding features in combination with the complexity of the model, it is possible to increase the complexity of the model, enabling the model to consider the feature output results at multiple levels, and at the same time reducing the redundant modules of the model and improving the computational efficiency of the model. In addition, according to the evaluation results of the model, optimizing and adjusting the model, such as adjusting the parameters of the model, adding or deleting features, optimizing feature engineering, etc., can improve the accuracy and recall rate of the model. Over time and with the development of the business, the abnormal application means and user behavior patterns may change. Therefore, it is necessary to update and retrain the model regularly to ensure the timeliness and effectiveness of the model.

[0101] In an exemplary embodiment, in the field of AI credit anti-fraud, the training process of the abnormal behavior recognition model can refer to Figure 6. First, data collection and preprocessing. Obtain multi-dimensional data related to users - collect users' basic information (such as age, gender, contact information, etc.), transaction records (such as transaction time, amount, location, merchant type, etc.), behavioral data (such as login time, login location, operation habits, etc.), and historical abnormal application records, etc. Clean the collected data to remove duplicate, incorrect, or missing data. Standardize different types of user data, such as unifying the value range, format, etc., for subsequent analysis and modeling. Secondly, feature transformation. Select features with strong correlation with abnormal behaviors from user data, such as the size and fluctuation of transaction amounts, the abnormality of transaction times, multiple logins from the same IP address or device within a short period of time, etc. Through the model, learn the correlation between user data and abnormal behaviors, and generate target features to reflect the abnormal behavior risks of users. For example, calculate the transaction frequency of a user within a certain period of time, the ratio of the average transaction amount to the current transaction amount, etc. For categorical features, such as gender, merchant type, etc., perform encoding transformation. Common ones include One-Hot Encoding, Label Encoding, etc., to convert them into numerical features recognizable by the model. It can be understood that corresponding weights can also be configured for the target features. After that, during the process of model training and evaluation, determine the target algorithm for training based on data characteristics and actual requirements. The target algorithms include at least one of logistic regression, decision tree, random forest, XGBoost, support vector machine, and artificial neural network. Specifically, the dataset can be divided into a training set and a test set, usually with a ratio of 7:3 or 8:2, etc. Use the training set to train the model, and use the test set to evaluate the trained model to test the performance and generalization ability of the model. Model ability evaluation indicators use metrics such as Accuracy, Recall, F1 value, AUC-ROC curve, etc. to evaluate the effect of the model. The larger the AUC value, the stronger the discrimination ability of the model, and it can better identify abnormal application new behaviors and normal application behaviors. Then, after the model is trained, use the trained model to predict the samples to obtain the probability related to the application resource result. Convert the probability into a score. For example, based on the mapping relationship of the score card, a benchmark score and a multiple can be set, and corresponding scores are assigned according to the size of the probability. For example, set the benchmark score to 500 points and the multiple to 50, and calculate the corresponding score according to the probability. According to business requirements and risk tolerance, determine a score threshold. Samples above the threshold are determined to be abnormal behaviors, and samples below or equal to the threshold are determined to be normal behaviors. The setting of the threshold can refer to the evaluation results of the model, such as selecting the point with a higher accuracy while ensuring a certain recall rate as the threshold.

[0102] It should be understood that although the steps in the flowcharts involved in the above embodiments are sequentially shown according to the indications of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0103] Based on the same inventive concept, an embodiment of the present application also provides a resource data processing device for implementing the resource data processing method involved above. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the resource data processing device provided below can refer to the limitations on the resource data processing method in the above text, and will not be repeated here.

[0104] In an exemplary embodiment, as Figure 7 shown, a resource data processing device 700 is provided, including: a data acquisition module 701, a service information acquisition module 702, an application processing module 703, a model processing module 704, an adjustment information generation module 705, and an update module 706, where:

[0105] The data acquisition module 701 is configured to acquire user data corresponding to a user who applies for a target resource within a resource cycle.

[0106] The service information acquisition module 702 is configured to acquire service resource processing information corresponding to the target resource.

[0107] The application processing module 703 is configured to process the user data based on the service resource processing information to obtain application data corresponding to the user's application for the target resource within the resource cycle.

[0108] The model processing module 704 is configured to input the application data into an abnormal behavior recognition model, analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model, and identify the application behavior data corresponding to the trend change information.

[0109] The adjustment information generation module 705 is configured to, if the application behavior data corresponds to an abnormal behavior, obtain adjustment information for the service resource processing information based on the trend change information.

[0110] An update module 706 is configured to update service resource processing information based on adjustment information, so as to perform resource data processing based on the updated service resource processing information.

[0111] In some embodiments, in terms of obtaining adjustment information for service resource processing information based on trend change information, the adjustment information generation module 705 is further configured to: obtain a target abnormal feature corresponding to the trend change information; obtain a reference value corresponding to the target abnormal feature in the service resource processing information, where the reference value is used to determine the application resource result in the application data, and the service resource processing information is used to associate each data feature with the corresponding reference value; obtain the application resource result data corresponding to the user data for the service resource processing information; and update the reference value based on the difference between the application resource result data and the target application resource result data to obtain the corresponding target reference value.

[0112] In some embodiments, the service resource processing information includes login time processing information, different login times correspond to different weights, and the weights are used to calculate the application resource result data. In terms of obtaining a target abnormal feature corresponding to the trend change information, the adjustment information generation module 705 is further configured to: obtain an abnormal login time corresponding to the trend change information; obtain the level weight corresponding to the abnormal login time in the login time processing information; obtain the application resource result data corresponding to the user data corresponding to the abnormal login time for the login time processing information; and update the level weight to the target level weight based on the difference between the application resource result data and the target application resource result data, so that the difference between the target application resource result data and the application resource result data corresponding to the target level weight meets a preset condition.

[0113] In some embodiments, the application data includes service processing data. In terms of inputting the application data into an abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model and identifying the application behavior data corresponding to the trend change information, the model processing module 704 is further configured to: input the application data into the abnormal behavior recognition model to obtain the data distribution information corresponding to the service processing data based on the abnormal behavior recognition model, compare the data distribution information with the historical data distribution information corresponding to the service processing data to obtain the first change information corresponding to the service processing data, and if the first change information corresponds to exceeding a target threshold, identify the application behavior data corresponding to the first change information.

[0114] In some embodiments, before determining that the application behavior data corresponding to the service processing data is an abnormal behavior if the first change information corresponds to exceeding a target threshold, the model processing module 704 is further configured to: analyze the historical data distribution information corresponding to the service processing data based on the abnormal behavior recognition model to predict the predicted data distribution information corresponding to the service processing data within a resource cycle; obtain the difference information between the predicted data distribution information and the historical data distribution information; and determine the target threshold based on the difference information.

[0115] In some embodiments, the resource data processing device 700 is further configured to: obtain an external data source and application resource training data corresponding to the external data source, where the application resource training data corresponds to actual application result data; extract features from the external data source to obtain external data features; extract features from the application resource training data to obtain application data features; obtain the association relationship between the external data features and the application data features based on the actual application result data; predict the predicted application result data corresponding to the application resource training data based on the association relationship, and train the abnormal behavior recognition model based on the difference information between the predicted application result data and the actual application result data to update the abnormal behavior recognition model.

[0116] In some embodiments, the resource data processing device 700 is further configured to: obtain the complexity evaluation result of the updated abnormal behavior recognition model; if the complexity evaluation result is higher than or equal to a preset threshold, obtain the first correlation degree between each external data feature and other external data features and the second correlation degree between each application data feature and other application data features, and if the first correlation degree meets the first preset condition, delete at least one application data feature corresponding to the first correlation degree, and if the second correlation degree meets the second preset condition, delete at least one application data feature corresponding to the second correlation degree; if the complexity evaluation result is lower than the preset threshold, generate fused external data features based on each external data feature to extract features from the external data source based on each external data feature and the fused external data features, and / or generate fused application data features based on each application data feature to extract features from the application resource training data based on each application data feature and the fused application data features.

[0117] Each module in the above resource data processing device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or stored in a memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0118] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 8As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data such as user data and business resource processing information. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a resource data processing method.

[0119] Those skilled in the art can understand that Figure 8 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0120] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0121] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0122] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0123] 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 for analysis, stored data, displayed data, 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 relevant data need to comply with relevant regulations.

[0124] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory 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), magnetoresistive 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 be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0125] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered as the scope recorded in the present application.

[0126] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for processing resource data, characterized in that The method includes: Obtaining user data corresponding to users who apply for target resources during a resource cycle; Obtaining business resource processing information corresponding to the target resources; Processing the user data based on the business resource processing information to obtain application data corresponding to the users' application for the target resources during the resource cycle; Inputting the application data into an abnormal behavior recognition model to analyze trend change information corresponding to the application data based on the abnormal behavior recognition model and identify application behavior data corresponding to the trend change information; If the application behavior data corresponds to an abnormal behavior, obtaining adjustment information for the business resource processing information based on the trend change information; Updating the business resource processing information based on the adjustment information to perform resource data processing based on the updated business resource processing information.

2. The method according to claim 1, wherein The obtaining adjustment information for the business resource processing information based on the trend change information includes: Obtaining a target abnormal feature corresponding to the trend change information; Obtaining a reference value corresponding to the target abnormal feature in the business resource processing information, where the reference value is used to determine the application resource result in the application data, and the business resource processing information is used to associate each data feature with the corresponding reference value; Obtaining application resource result data corresponding to the user data for the business resource processing information; Updating the reference value based on the difference between the application resource result data and target application resource result data to obtain a corresponding target reference value.

3. The method according to claim 1, wherein The obtaining adjustment information for the business resource processing information based on the trend change information includes: Obtaining a target abnormal feature corresponding to the trend change information; Obtaining a level weight corresponding to the target abnormal feature in the business resource processing information; Obtaining application resource result data corresponding to the user data for the business resource processing information; Updating the level weight to a target level weight based on the difference between the application resource result data and target application resource result data so that the difference between the target application resource result data and the application resource result data corresponding to the target level weight meets a preset condition.

4. The method according to claim 1, wherein The application data includes business processing data. The inputting the application data into an abnormal behavior recognition model to analyze trend change information corresponding to the application data based on the abnormal behavior recognition model and identify application behavior data corresponding to the trend change information includes: Inputting the application data into the abnormal behavior recognition model to obtain data distribution information corresponding to the business processing data based on the abnormal behavior recognition model; Comparing the data distribution information with historical data distribution information corresponding to the business processing data to obtain first change information corresponding to the business processing data; If the first change information corresponds to exceeding a target threshold, identifying application behavior data corresponding to the first change information.

5. The method according to claim 4, wherein Before the if the first change information corresponds to exceeding a target threshold, identifying application behavior data corresponding to the first change information, the method further includes: Analyze the historical data distribution information corresponding to the service processing data based on the abnormal behavior recognition model to predict the predicted data distribution information corresponding to the service processing data within the resource cycle; Obtain the difference information between the predicted data distribution information and the historical data distribution information; Determine the target threshold based on the difference information.

6. The method according to claim 1, characterized in that The method further includes: Obtain an external data source and application resource training data corresponding to the external data source, and the application resource training data corresponds to actual application result data; Extract features from the external data source to obtain external data features; extract features from the application resource training data to obtain application data features; Based on the actual application result data, obtain the correlation relationship between the external data features and the application data features; Predict the predicted application result data corresponding to the application resource training data based on the correlation relationship, and train the abnormal behavior recognition model based on the difference information between the predicted application result data and the actual application result data to update the abnormal behavior recognition model.

7. The method according to claim 6, characterized in that The method further includes: Obtain the complexity evaluation result of the updated abnormal behavior recognition model; If the complexity evaluation result is higher than or equal to a preset threshold, obtain the first correlation degree between each of the external data features and other external data features and the second correlation degree between each of the application data features and other application data features. If the first correlation degree meets the first preset condition, delete at least one application data feature corresponding to the first correlation degree. If the second correlation degree meets the second preset condition, delete at least one application data feature corresponding to the second correlation degree; If the complexity evaluation result is lower than the preset threshold, generate fused external data features based on each of the external data features to extract features from the external data source based on each of the external data features and the fused external data features, and / or generate fused application data features based on each of the application data features to extract features from the application resource training data based on each of the application data features and the fused application data features.

8. A resource data processing device, characterized in that, The device includes: A data acquisition module for acquiring user data of a user who applies for a target resource within a resource cycle; A service information acquisition module for acquiring service resource processing information corresponding to the target resource; An application processing module for processing the user data based on the service resource processing information to obtain application data corresponding to the user's application for the target resource within the resource cycle; A model processing module for inputting the application data into an abnormal behavior recognition model to analyze the trend change information corresponding to the application data based on the abnormal behavior recognition model and identify the application behavior data corresponding to the trend change information; An adjustment information generation module for, if the application behavior data corresponds to an abnormal behavior, obtaining adjustment information for the service resource processing information based on the trend change information; An update module, configured to update the service resource processing information based on the adjustment information, so as to perform resource data processing based on the updated service resource processing information.

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

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

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images