Object data determination method and apparatus, storage medium, and electronic device

Abstract

The present disclosure provides a subject data determination method and device, a storage medium and an electronic device. The method comprises: in the case of receiving a target data processing request, obtaining an account operation feature of a target object account; determining at least one target feature label matched with the target object account and a label matching degree corresponding to each of the at least one target feature label according to the account operation feature; determining at least one target data operation condition matched with the account operation feature from a plurality of candidate data operation conditions according to the at least one target feature label and the label matching degree corresponding to each of the at least one target feature label; and processing current credit data of the target object account in a data processing mode matched with the at least one target data operation condition to obtain target credit data. The present disclosure solves the technical problem of poor accuracy of related resource quota adjustment.

Description

Technical Field

[0001] This invention relates to the field of computers, and more specifically, to a method and apparatus for determining object data, a storage medium, and an electronic device. Background Technology

[0002] In scenarios involving interaction based on object resources, to improve interaction efficiency, relevant entities typically pre-allocate a portion of object resources to object accounts for permissible operations. The allocated resource quota for each object account is then adjusted based on its resource usage over a certain period.

[0003] Current methods for adjusting resource limits typically rely on traditional statistical analysis and human experience. By statistically analyzing basic information and operation records of target accounts, fixed resource characteristic rules are determined. These rules are then manually adjusted based on the user's experience. However, this manual adjustment method depends solely on the matching results of specific rules and fails to adequately consider the differences among multiple target accounts across various dimensions. This limitation prevents the development of personalized data processing methods to address these differences, resulting in inefficient limit adjustment strategies. In other words, existing technologies suffer from poor accuracy in resource limit adjustments.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method and apparatus for determining object data, a storage medium, and an electronic device, to at least solve the technical problem of poor accuracy in resource quota adjustment in related fields.

[0006] According to one aspect of the present invention, a method for determining object data is provided, comprising: upon receiving a target data processing request, acquiring account operation characteristics of a target object account, wherein the account operation characteristics are determined based on at least one account operation of the target object account within a target period; determining at least one target feature tag matching the target object account and a tag matching degree corresponding to each of the at least one target feature tag based on the account operation characteristics; determining at least one target data operation condition matching the account operation characteristics from a plurality of candidate data operation conditions based on the at least one target feature tag and the tag matching degree corresponding to each of the at least one target feature tag; and processing the current credit data of the target object account according to a data processing method matching the at least one target data operation condition to obtain target credit data, wherein the target credit data is used to indicate the total resource limit allowed for the target object account to operate in a target interaction activity.

[0007] According to another aspect of the present invention, an apparatus for determining object data is also provided, comprising: an acquisition unit, configured to acquire account operation characteristics of a target object account upon receiving a target data processing request, wherein the account operation characteristics are determined based on at least one account operation of the target object account within a target period; a first determination unit, configured to determine at least one target feature tag matching the target object account and a tag matching degree corresponding to each of the at least one target feature tag based on the account operation characteristics; a second determination unit, configured to determine at least one target data operation condition matching the account operation characteristics from a plurality of candidate data operation conditions based on the at least one target feature tag and the tag matching degree corresponding to each of the at least one target feature tag; and a processing unit, configured to process the current credit data of the target object account according to a data processing method matching the at least one target data operation condition to obtain target credit data, wherein the target credit data is used to indicate the total resource limit allowed for the target object account to operate in a target interaction activity.

[0008] Optionally, the first determining unit is configured to: obtain cluster center features corresponding to each of the plurality of fuzzy clusters, wherein the plurality of fuzzy clusters correspond to a plurality of operation feature labels; determine the feature distance between the account operation feature and the plurality of cluster center features; determine the cluster matching degree between the account operation feature and the plurality of fuzzy clusters based on the plurality of feature distances; and determine at least one target feature label that matches the target object account, and the label matching degree corresponding to each of the at least one target feature label, based on the plurality of cluster matching degrees.

[0009] Optionally, the first determining unit is configured to: obtain fuzzy control coefficients, wherein the fuzzy control coefficients determine the distribution of the cluster matching degree; obtain a first feature distance between the account operation feature and the current cluster center feature among the multiple cluster center features, wherein the current cluster center feature is a cluster center feature that matches the current fuzzy cluster; obtain multiple second feature distances between the account operation feature and the multiple cluster center features; and determine the current cluster matching degree between the account operation feature and the current fuzzy cluster based on the first feature distance, the multiple second feature distances, and the fuzzy control coefficients.

[0010] Optionally, the first determining unit is configured to: determine a matching reference value based on the cumulative result of multiple cluster matching degrees; determine the cluster label corresponding to the current fuzzy cluster among multiple fuzzy clusters as the target feature label; and determine the ratio between the current cluster matching degree corresponding to the current fuzzy cluster and the matching reference value as the label matching degree corresponding to the target feature label.

[0011] Optionally, the above-mentioned object data determining device further includes: a clustering unit, used to obtain an operation feature set determined based on the object account set; obtain clustering control coefficients, wherein the clustering control coefficients are used to determine the number of clusters in the fuzzy clusters; determine a fuzzy partitioning matrix based on the feature distance between every two account operation features in the operation feature set and the clustering control coefficients, wherein the fuzzy partitioning matrix is ​​used to indicate the clustering matching degree between any one of the account operation features in the operation feature set and multiple fuzzy clusters; and determine the cluster center features corresponding to each of the multiple fuzzy clusters when the fuzzy partitioning matrix satisfies the target clustering conditions.

[0012] Optionally, the second determining unit is configured to: determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the target account has performed a conditional account operation within the conditional period; determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the first account reference indicator of the target account is within the conditional indicator range; or determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the changing trend of the second account reference indicator of the target account satisfies the conditional trend.

[0013] Optionally, the processing unit is configured to: acquire a condition priority coefficient corresponding to at least one of the target data operation conditions; determine an execution condition from the at least one target data operation condition based on the condition priority coefficient; process the current credit data of the target account according to the data processing method matched by the execution condition; acquire a condition weight coefficient corresponding to at least one of the target data operation conditions; weight the data operation values ​​corresponding to at least one target data operation condition according to the condition weight coefficient to obtain a target operation value; process the current credit data of the target account according to the data processing method matched by the target operation value; acquire an execution order coefficient corresponding to at least one of the target data operation conditions; determine a target execution order from the data processing methods corresponding to at least one target data operation condition based on the execution order coefficient; and process the current credit data of the target account according to the target execution order corresponding to at least one data processing method.

[0014] Optionally, the processing unit is configured to: obtain multiple first data operation conditions from a data cache, wherein the first data operation conditions are candidate data operation conditions with a matching hit rate greater than or equal to a target value; if the multiple first data operation conditions include at least one target data operation condition, obtain at least one data processing method matching the target data operation condition; if the multiple first data operation conditions do not include the target data operation condition, obtain multiple second data operation conditions from a database, wherein the second data operation conditions are candidate data operation conditions with a matching hit rate less than the target value; and determine at least one target data operation condition matching the account operation feature from the multiple second data operation conditions based on at least one target feature tag and the tag matching degree corresponding to each of the at least one target feature tag.

[0015] Optionally, the above-mentioned object data determination device further includes at least one of the following: a first update unit, configured to update at least one data processing parameter indicating at least one of the above-mentioned data processing methods when the data processing method corresponding to at least one of the above-mentioned candidate data operation conditions satisfies the first update condition, wherein the above-mentioned data processing parameter includes a processing trend parameter indicating a data processing trend and a processing value parameter indicating a data processing value; and a second update unit, configured to update at least one clustering parameter indicating a clustering method of the above-mentioned fuzzy clustering when the data processing method corresponding to at least one of the above-mentioned candidate data operation conditions satisfies the second update condition, wherein the above-mentioned clustering parameter includes a cluster quantity parameter, a fuzzy control coefficient, an initial cluster center, and a data preprocessing parameter.

[0016] Optionally, the aforementioned object data determination device further includes: a condition judgment unit, configured to acquire a first set of accounts matching the first data processing method corresponding to the aforementioned reference data operation conditions, and a first set of operation features corresponding to the aforementioned first set of accounts, wherein the aforementioned first set of accounts includes first object accounts that process credit data according to the aforementioned first data processing method, and the aforementioned first set of operation features includes first operation features corresponding to each of the aforementioned first object accounts; based on the aforementioned first set of operation features, determine at least one evaluation indicator matching the aforementioned first data processing method; based on the weighted summation result between at least one of the aforementioned evaluation indicators, determine a target evaluation indicator matching the aforementioned first data processing method; if the aforementioned target evaluation indicator is within a first indicator range, determine that the aforementioned first data processing method satisfies the aforementioned first update condition; if the aforementioned target evaluation indicator is within a second indicator range, determine that the aforementioned first data processing method satisfies the aforementioned second update condition.

[0017] Optionally, the aforementioned condition judgment unit is used for at least one of the following: determining a risk indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein the risk indicator is used to indicate the degree of operational risk of the target institution configuring the aforementioned total resource quota for the aforementioned first target account; determining a return indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein the risk indicator is used to indicate the degree of return corresponding to the aforementioned target institution configuring the aforementioned total resource quota for the aforementioned first target account; determining a risk indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein after the target institution configures the aforementioned total resource quota for the aforementioned first target account, the risk indicator is determined based on the feedback operation of the aforementioned first target account.

[0018] Optionally, the aforementioned object data determining device further includes: a third updating unit, configured to acquire a second set of accounts matching the second data processing method corresponding to the aforementioned reference data operation conditions, and a second set of operation features corresponding to the aforementioned second set of accounts, wherein the aforementioned second set of accounts includes second object accounts that process credit data according to the aforementioned second data processing method, and the aforementioned second set of operation features includes second operation features corresponding to each of the aforementioned second object accounts; predict a third set of operation features corresponding to the second operation period and a second set of transaction platform description features corresponding to the aforementioned second operation period based on the aforementioned second set of operation features corresponding to the first operation period and the aforementioned first set of transaction platform description features corresponding to the aforementioned first operation period, wherein the aforementioned second operation period is later than the aforementioned first operation period; and update at least one data processing parameter of the aforementioned second data processing method when the aforementioned third set of operation features and the aforementioned second set of transaction platform description features satisfy the third updating condition.

[0019] Optionally, the third update unit is further configured to: obtain a set of reference operation features corresponding to the second operation period and a set of reference trading platform description features corresponding to the second operation period, wherein the set of reference operation features includes reference operation features determined based on the actual account operations of the second object account in the second operation period, and the reference trading platform description features are the actual trading platform description features corresponding to the second operation period; and determine that the third operation feature set and the second trading platform description features satisfy the third update condition based on a first degree of difference between the set of reference operation features and the third set of operation features, and a second degree of difference between the reference trading platform description features and the second trading platform description features.

[0020] Optionally, the third update unit is configured to: input the second set of operational features and the first trading platform description features into at least one prediction model to obtain at least one prediction result, wherein the prediction result includes a third set of reference operational features corresponding to the second operational period and a second set of reference trading platform description features corresponding to the second operational period; and determine a target prediction result based on at least one of the prediction results, wherein the target prediction result includes the third set of operational features corresponding to the second operational period and the second set of reference trading platform description features corresponding to the second operational period.

[0021] According to another aspect of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to execute the above-described method for determining object data at runtime.

[0022] According to another aspect of the embodiments of this application, a computer program product or computer program is provided, which includes 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 method for determining object data as described above.

[0023] According to another aspect of the present invention, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to execute the object data determination method through the computer program.

[0024] In the embodiments of this application, when a target data processing request is received, the account operation characteristics of the target account are first obtained. Based on these account operation characteristics, at least one target feature tag matching the target account is determined, along with a tag matching degree corresponding to each target feature tag. In other words, in the embodiments of this application, the operation characteristics of an account can be matched with multiple feature tags, and each tag will have a corresponding tag matching degree. This method of combining feature tags and their matching degrees achieves fuzzy classification of account operation characteristics.

[0025] Furthermore, conditional matching based on the results of this fuzzy classification can fully consider the personalized characteristics reflected in the operational features of each account, achieving more accurate conditional matching. By utilizing a data feature matching mechanism, personalized resource adjustment strategies are formulated for each account, thereby solving the technical problem of insufficient accuracy in existing resource adjustment methods. Through the above implementation method, not only is the accuracy of resource adjustment improved, but the adaptability and flexibility of the strategy are also enhanced. Attached Figure Description

[0026] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0027] Figure 1 This is a schematic diagram of the hardware environment for an optional method for determining object data according to an embodiment of the present invention;

[0028] Figure 2 This is a flowchart of an optional method for determining object data according to an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of an optional clustering method according to an embodiment of the present invention;

[0030] Figure 4This is a schematic diagram of another optional clustering method according to an embodiment of the present invention;

[0031] Figure 5 This is a flowchart of another method for determining object data according to an embodiment of the present invention;

[0032] Figure 6 This is a schematic diagram of an optional object data determination architecture according to an embodiment of the present invention;

[0033] Figure 7 A schematic diagram of an optional object data determination device according to an embodiment of the present invention;

[0034] Figure 8 This is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present invention. Detailed Implementation

[0035] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0036] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0037] It should be noted that in the relevant embodiments of this application, the acquisition and processing of relevant data information of the target account are all obtained in advance through methods that comply with the relevant normative legal documents, and the authorization permission of the corresponding account subject must be obtained before acquiring the above-mentioned relevant information and data.

[0038] According to one aspect of the present invention, a method for determining object data is provided. As an optional implementation, the above-described method for determining object data may be applied, but is not limited to, to applications such as... Figure 1The system for determining object data in the hardware environment shown may include, but is not limited to, terminal device 102, network 110, and server 104, wherein server 104 further includes a database and a processing engine. The aforementioned network may include, but is not limited to, wired networks and wireless networks, wherein the wired network includes local area networks (LANs), metropolitan area networks (MANs), and wide area networks (WANs), and the wireless network includes Bluetooth, Wi-Fi, and other networks that enable wireless communication.

[0039] Terminal device 102 runs a target object account client (taking a resource operation object account client that can perform account operations based on object resources as an example). Terminal device 102 includes a human-computer interaction screen, a processor, and a memory. The human-computer interaction screen in terminal device 102 is used to display interactive screens during resource operations. Terminal device 102 also provides a human-computer interaction interface to receive object interaction operations for performing resource operations. These object interaction operations may include, but are not limited to, resource exchange operations, resource consumption operations, and resource storage operations. The processor is used to generate interaction instructions in response to the aforementioned human-computer interaction operations and execute corresponding local operations according to the interaction instructions.

[0040] The aforementioned server 104 can be a server cluster consisting of multiple servers, or a cloud server. The aforementioned server includes a database and a processing engine.

[0041] Assumption Figure 1 The terminal device 102 runs a resource operation object account client for performing resource operations, and the server 104 is used to respond to data processing requests initiated by at least one resource operation object account client. The specific process of this embodiment is as follows: the terminal device 102 executes step S102 to run the resource operation application; then, as in step S104, the terminal device 102 sends a scene image to the server 104 through the network 110.

[0042] Next, on server 104, steps S106, S108, S110, and S112 are executed. Upon receiving a target data processing request, the account operation characteristics of the target account are obtained. These account operation characteristics are determined based on at least one account operation performed by the target account within the target period. Based on the account operation characteristics, at least one target feature tag matching the target account and the tag matching degree corresponding to each of the at least one target feature tag are determined. Based on the at least one target feature tag and the tag matching degree corresponding to each of the at least one target feature tag, at least one target data operation condition matching the account operation characteristics is determined from multiple candidate data operation conditions. The current credit data of the target account is processed according to the data processing method matching the at least one target data operation condition to obtain target credit data. The target credit data is used to indicate the total resource limit that the target account is allowed to operate in the target interaction activity.

[0043] Next, server 104 executes S114, pushing the processing result to terminal device 102 based on target credit data via network 110;

[0044] Finally, terminal device 102 executes S116 and displays the data processing results.

[0045] In another alternative implementation, if the terminal device 102 is a trusted terminal, the steps S106 to S112 can also be implemented in the terminal device 102.

[0046] Optionally, in this embodiment, the terminal device 102 can be a terminal device configured with a target account client, which may include, but is not limited to, at least one of the following: mobile phone (such as Android phone, iOS phone, etc.), laptop computer, tablet computer, PDA, MID (Mobile Internet Devices), PAD, desktop computer, smart TV, etc. The target account client can be a video target account client, instant messaging target account client, browser target account client, educational target account client, or other target account client that supports providing game tasks. The network may include, but is not limited to, wired networks and wireless networks. The wired network includes: local area network (LAN), metropolitan area network (MAN), and wide area network (WAN). The wireless network includes: Bluetooth, Wi-Fi, and other networks that enable wireless communication. The server may be a single server, a server cluster composed of multiple servers, or a cloud server. The above is merely an example, and no limitation is made in this embodiment.

[0047] In the embodiments of this application, when a target data processing request is received, the account operation characteristics of the target account are first obtained. Based on these account operation characteristics, at least one target feature tag matching the target account is determined, along with a tag matching degree corresponding to each target feature tag. In other words, in the embodiments of this application, the operation characteristics of an account can be matched with multiple feature tags, and each tag will have a corresponding tag matching degree. This method of combining feature tags and their matching degrees achieves fuzzy classification of account operation characteristics.

[0048] Furthermore, conditional matching based on the results of this fuzzy classification can fully consider the personalized characteristics reflected in the operational features of each account, achieving more accurate conditional matching. By utilizing a data feature matching mechanism, personalized resource adjustment strategies are formulated for each account, thereby solving the technical problem of insufficient accuracy in existing resource adjustment methods. Through the above implementation method, not only is the accuracy of resource adjustment improved, but the adaptability and flexibility of the strategy are also enhanced.

[0049] The above is merely an example, and no limitation is made in this embodiment.

[0050] As an optional implementation method, such as Figure 2 As shown, the method for determining the object data described above can be applied to a server, and may specifically include the following steps:

[0051] S202, Upon receiving a target data processing request, obtain the account operation characteristics of the target object account, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period;

[0052] S204, Based on the account operation characteristics, determine at least one target feature tag that matches the target account, and the tag matching degree corresponding to each of the at least one target feature tag;

[0053] S206, Based on at least one target feature label and the label matching degree corresponding to each of the at least one target feature label, determine at least one target data operation condition matching the account operation feature from multiple candidate data operation conditions;

[0054] S208, the current credit data of the target object account is processed according to a data processing method that matches at least one target data operation condition to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

[0055] It should be noted that the above-described embodiments of this application can be applied to scenarios or platforms that support object accounts performing object operations based on object resources.

[0056] Optionally, when the aforementioned object resource is specifically a computing resource, the above implementation method can be applied to cloud computing scenarios. Correspondingly, in the aforementioned cloud computing scenario, an object account can request a certain amount of computing resources (such as CPU, memory, and storage space) to be pre-allocated from the cloud service provider according to its business needs. Based on the object account's resource usage over a certain period and the object account's account level, the cloud service provider can adjust its computing resource allocation quota according to the above steps S202 to S208.

[0057] Furthermore, in cloud computing scenarios, account operation characteristics can reflect the resource utilization of an object account within a certain period, and the target feature tags and corresponding tag matching degrees of the object account can be determined based on these characteristics. These target feature tags can be used for fuzzy classification of object accounts in cloud computing scenarios. For example, classification methods in cloud computing scenarios may include, but are not limited to: First tag: object accounts with high computational load but low frequency; Second tag: object accounts with low computational load but high frequency; Third tag: object accounts with high computational load and high frequency; Fourth tag: object accounts with high computational load and service request times typically distributed at night.

[0058] Understandably, the boundaries between the various tags mentioned above are often quite blurred. For example, the same account can be configured with both the first and fourth tags simultaneously. Furthermore, by simultaneously acquiring the two target feature tags and obtaining the tag matching degree corresponding to the two target feature tags, a more precise description and positioning of the account can be achieved.

[0059] Optionally, when the aforementioned object resource is specifically a virtual resource, the above implementation method can be applied to a game scenario. Correspondingly, in the aforementioned game scenario, the object account can obtain virtual items or virtual transaction resources pre-allocated by the game platform. Based on the object account's game operation behavior (such as online time, virtual consumption records) and account information such as the object account level within a certain period, the game platform can adjust the resource quota of virtual items and other virtual resources according to the above steps S202 to S208.

[0060] Furthermore, in the context of gaming platforms, the game operation characteristics of player accounts can reflect the player's game participation and virtual resource utilization over a certain period. Based on these operation characteristics, target feature tags and corresponding tag matching degrees can be determined for the player account. These target feature tags can be used to fuzzily classify the player account's behavior in the game. For example, classification methods may include, but are not limited to: First tag: high-spending player, but less game time; Second tag: low-spending player, but more game time; Third tag: high-spending and high-frequency player; Fourth tag: player who prefers to play during specific time periods (such as nighttime).

[0061] Similarly, the boundaries between the various tags mentioned above are often quite blurred. For example, the same player account can be configured with both the first and fourth tags simultaneously. Therefore, by simultaneously acquiring the two target feature tags and obtaining the tag matching degree corresponding to the two target feature tags, a more accurate description and positioning of the player account can be achieved.

[0062] Furthermore, gaming platforms can determine additional social tags based on players' social interaction characteristics (such as team-up frequency and community activity) to further refine players' gaming behavior. In this way, gaming platforms can not only provide players with a more personalized gaming experience but also dynamically adjust game resource allocation strategies based on player behavior, thereby improving resource utilization efficiency and player satisfaction.

[0063] Optionally, when the aforementioned object resource is specifically a financial resource, the above implementation method can be applied to a financial resource pre-authorization scenario. Correspondingly, in the aforementioned financial resource pre-authorization scenario, the object account can apply to the financial entity for pre-authorized financial resources, and the financial entity can adjust the amount of financial resources authorized to the object account based on the object account's relevant financial operations within a certain period, i.e., according to the aforementioned steps S202 to S208.

[0064] Let's continue with the example of financial resource pre-authorization. In this scenario, the resource operation characteristics of a target account can reflect the player's resource consumption and repayment status over a certain period. Based on these characteristics, target feature tags and corresponding tag matching degrees can be determined for the target account. These target feature tags can be used to fuzzily classify the player's behavior in the game. For example, classification methods may include, but are not limited to: First tag: High-risk, high-consumption group; Second tag: Low-risk, low-consumption group; Third tag: Medium-risk, medium-consumption group.

[0065] Similarly, the boundaries between the various tags mentioned above can be somewhat blurred. For example, if the risk indicators and consumption indicators for the same account both fall within the boundaries defined by the tags, the account can be configured with multiple tags simultaneously. Therefore, by simultaneously acquiring multiple target feature tags and obtaining the tag matching degree corresponding to these multiple target feature tags, a more accurate description and positioning of the target account can be achieved.

[0066] It should be noted that in the above-mentioned resource pre-authorization scenario, the account operation characteristics in step S202 can be obtained, but are not limited to, from the following operation characteristic information after data cleaning, noise reduction, and normalization: The above operation characteristic information includes: basic account information, such as account holding duration and account type; resource operation records (such as resource repayment records, expected number of resource repayments, etc.); resource usage operations (such as consuming resources to exchange for other objects, resource usage frequency, etc.) and other multi-dimensional data. It should be noted that obtaining the above-mentioned relevant information requires authorization from the relevant entity and must be carried out in accordance with the provisions of relevant normative documents.

[0067] It is understood that the above application scenarios are merely illustrative examples and do not limit the specific scenarios in which the embodiments of this application can be applied.

[0068] In the aforementioned application scenarios, the resource allocation for a target account can be dynamically adjusted based on the account's activity over a certain period, thereby improving resource utilization and platform operational efficiency. Existing resource allocation adjustments based on traditional statistical analysis methods and human experience may not adequately consider the differences between accounts across multiple dimensions, resulting in suboptimal efficiency and accuracy in resource adjustment strategies.

[0069] According to the above-described embodiments of this application, when a target data processing request is received, the account operation characteristics of the target account are first obtained. Based on these account operation characteristics, at least one target feature tag matching the target account is determined, along with a tag matching degree corresponding to each target feature tag. In other words, in the embodiments of this application, the operation characteristics of an account can be matched with multiple feature tags, and each tag will have a corresponding tag matching degree. This combination of feature tags and their matching degrees achieves fuzzy classification of account operation characteristics.

[0070] Furthermore, conditional matching based on the results of this fuzzy classification can fully consider the personalized characteristics reflected in the operational features of each account, achieving more accurate conditional matching. By utilizing a data feature matching mechanism, personalized resource adjustment strategies are formulated for each account, thereby solving the technical problem of insufficient accuracy in existing resource adjustment methods. Through the above implementation method, not only is the accuracy of resource adjustment improved, but the adaptability and flexibility of the strategy are also enhanced.

[0071] The following provides a further explanation of the aforementioned fuzzy clustering method. In the embodiments of this application, the above-mentioned method of determining at least one target feature tag that matches the target object account, and the tag matching degree corresponding to each of the at least one target feature tag, includes:

[0072] S1, obtain the cluster center features corresponding to each of the multiple fuzzy clusters, wherein the multiple fuzzy clusters correspond to multiple operational feature labels respectively;

[0073] S2, determine the feature distance between account operation features and multiple cluster center features;

[0074] S3, determine the clustering matching degree between account operation features and multiple fuzzy clusters based on multiple feature distances;

[0075] S4. Based on multiple clustering matching degrees, determine at least one target feature label that matches the target object account, and the label matching degree corresponding to each of the at least one target feature label.

[0076] It should be noted that in the above embodiments of this application, multiple fuzzy clusters determined by the fuzzy clustering method are used in the process of determining the target feature labels and label matching degrees. In the embodiments of this application, data points can belong to multiple clusters to varying degrees, rather than, as in traditional hard clustering (such as K-Means), where data points can only belong to one cluster.

[0077] like Figure 3 As shown, in an optional fuzzy clustering method, node 302 is used to represent account operation features, and nodes 304, 306, 308, 310, and 312 are used to represent the cluster center features corresponding to the five different fuzzy clusters, respectively. Figure 3In the fuzzy clustering method shown, node 302 can simultaneously belong to the clusters corresponding to nodes 304, 306, 308, 310, and 312. The cluster matching degree is determined based on the distances between node 302 and each of these clusters. In this embodiment, each cluster can be configured with an operational feature label. Therefore, based on the five operational feature labels and the five cluster matching degrees, five corresponding target feature labels and their respective label matching degrees can be determined.

[0078] like Figure 4 As shown, in another optional fuzzy clustering method, node 402 is used to represent account operation features, and nodes 404, 406, etc., are used to represent the cluster center features corresponding to the five different fuzzy clusters, respectively. Figure 4 In the fuzzy clustering method shown, each fuzzy cluster is configured with a cluster radius. Since node 402 is simultaneously within the cluster radii of nodes 404 and 406, it can belong to the respective clusters of nodes 404 and 406. Based on the distances between node 402 and each of nodes 404 and 406, the corresponding cluster matching degree is determined. In this embodiment, each cluster can be configured with an operational feature label. Therefore, based on the two operational feature labels and the two cluster matching degrees, two corresponding target feature labels and their respective label matching degrees can be determined.

[0079] The following further explains the method for determining target feature labels based on fuzzy clustering. The above method determines the clustering matching degree between account operation features and multiple fuzzy clusters based on multiple feature distances, including:

[0080] S1, obtain the fuzzy control coefficients, where the fuzzy control coefficients determine the distribution of the cluster matching degree;

[0081] S2, obtain the first feature distance between the account operation feature and the current cluster center feature among multiple cluster center features, where the current cluster center feature is the cluster center feature that matches the current fuzzy cluster;

[0082] S3, obtain the distances between multiple second features between account operation features and multiple cluster center features;

[0083] S4. Based on the first feature distance, multiple second feature distances, and fuzzy control coefficients, determine the current clustering matching degree between the account operation features and the current fuzzy clustering.

[0084] In the above embodiments of this application, the current clustering matching degree between account operation features and the current fuzzy clustering can be characterized as u i,j, representing data point x i The degree to which a point belongs to cluster j. This can be a fuzzy value, between 0 and 1. It characterizes the probability or confidence that a point belongs to the current fuzzy cluster. Specifically, it can be determined as follows:

[0085]

[0086] Where m is the fuzzification parameter, i.e., the fuzzy control coefficient mentioned above, and v j Let be the j-th cluster center, and c be the number of clusters. |x i -v j |x| represents the first feature distance. i -v k This is the second feature distance.

[0087] It should be noted that the aforementioned fuzzy control coefficients can be used to control the degree of fuzziness in the membership of data points within clusters, determining the distribution range of membership degrees. When the value of m is large, the distribution of membership degrees is smoother, and data points are more likely to belong to multiple clusters with smaller membership degrees, thus increasing the fuzziness of the clustering. Conversely, when the value of m is close to 2, the distribution of membership degrees is sharper, and data points tend to belong entirely to one cluster, resulting in a clustering result closer to hard clustering.

[0088] For example Figure 3 In the fuzzy clustering method shown, node 302 can belong to every cluster, while Figure 4 The fuzzy clustering method shown indicates that node 402 belongs only to the clusters corresponding to nodes 404 and 406. It can be determined that... Figure 3 The first m value in the fuzzy clustering method shown is greater than Figure 4 The second m value in the fuzzy clustering method shown.

[0089] The following explains how to determine tag matching degree. It involves determining at least one target feature tag that matches the target account, and the tag matching degree corresponding to each of the at least one target feature tag, including:

[0090] S1, determine the matching reference value based on the cumulative result of multiple cluster matching degrees;

[0091] S2, determine the cluster label corresponding to the current fuzzy cluster among multiple fuzzy clusters as the target feature label;

[0092] S3, the ratio between the current cluster matching degree corresponding to the current fuzzy cluster and the matching reference value is determined as the label matching degree corresponding to the target feature label.

[0093] Specifically, the tag matching degree can be expressed as w i,j The above implementation method can be achieved using the following formula:

[0094]

[0095] In other words, in this embodiment, the label matching degree can be obtained by normalizing the cluster matching degree. That is, the label matching degree obtained by normalization can reflect the relative importance of data points in each fuzzy cluster.

[0096] The following describes the construction method for fuzzy clustering. In an optional implementation, before obtaining the account operation characteristics of the target object account upon receiving a target data processing request, the method further includes:

[0097] S1, Obtain the set of operational features determined based on the set of object accounts;

[0098] S2, obtain the clustering control coefficients, where the clustering control coefficients are used to determine the number of clusters in fuzzy clustering;

[0099] S3. Based on the feature distance between every two account operation features in the operation feature set and the clustering control coefficient, determine the fuzzy partitioning matrix, where the fuzzy partitioning matrix is ​​used to indicate the clustering matching degree between any account operation feature in the operation feature set and multiple fuzzy clusters.

[0100] S4. If the fuzzy partitioning matrix satisfies the target clustering conditions, determine the cluster center features corresponding to each of the multiple fuzzy clusters.

[0101] The following combination Figure 5 This paper explains a specific method for establishing fuzzy clustering in the field of financial credit granting.

[0102] S502, Obtain sample data;

[0103] This involves acquiring relevant data about the target account, including basic account information such as account holding duration and account type; resource operation records (such as resource repayment records and expected repayment times); and resource usage operations (such as consuming resources to exchange for other objects and resource usage frequency), among other multi-dimensional data. It should be noted that acquiring the above-mentioned information requires authorization from the relevant entity and must be conducted in accordance with the provisions of relevant regulatory documents.

[0104] S504, Data Preprocessing;

[0105] Next, the collected data undergoes preprocessing, including data cleaning, denoising, and normalization. Data cleaning primarily removes duplicate, erroneous, and outlier data to ensure accuracy and integrity. Denoising can be achieved using methods such as filtering to remove noise interference from the data. Normalization transforms the data to a uniform scale for subsequent analysis and processing.

[0106] Let the original data matrix be X = [x1, x2, ..., xn]. i ], where x i =[x i1 ,x i2 ,...,x im [] represents the m-dimensional feature vector of the i-th object account. After normalization, the normalized data matrix Y = [y1, y2, ..., y3] is obtained. i ], where y i =[y i1 ,y i2 ,...,y im ], and satisfy Normalization can be determined using the following formula:

[0107]

[0108] The above implementation methods can improve the quality and availability of data, providing a reliable data foundation for subsequent fuzzy clustering and rule engine processing.

[0109] S506, Initialize fuzzy clustering;

[0110] Fuzzy clustering is a method for dividing data objects into several fuzzy groups or categories, allowing each data object to belong to different categories to varying degrees. In this invention, the Fuzzy C-means clustering algorithm (FCM) is used to classify the accounts of credit resource objects.

[0111] The basic idea of ​​the FCM algorithm is: given a set of data points X = [x1, x2, ..., x...], ... i The cluster is divided into c fuzzy clusters, each consisting of a cluster center v1, v2, ..., v3. j The algorithm aims to minimize the following objective function:

[0112]

[0113] Among them, U=[u i,j ] is an n×c fuzzy partitioning matrix, u i,j U represents the degree to which the i-th data point xi belongs to the j-th cluster, satisfying 0 ≤ u i,j ≤1 and

[0114] m is the fuzzy weighting exponent, usually taken as m>1; ||x i -v j ‖ represents data point x i With cluster center v i The Euclidean distance between them.

[0115] To minimize the objective function J(U,V), an iterative optimization method is adopted. The specific steps are as follows: First, an initialization operation is performed, randomly initializing the fuzzy partition matrix U and the cluster centers V.

[0116] S508, Calculate cluster centers; based on the current fuzzy partition matrix U, calculate the center v of each cluster. j Specifically, it can be obtained in the following ways:

[0117]

[0118] S510, Update the fuzzy partitioning matrix, which can be determined in the following way:

[0119]

[0120] Where m is the fuzzification parameter, i.e., the fuzzy control coefficient mentioned above, and v j Let be the j-th cluster center, and c be the number of clusters. |x i -v j |x| represents the first feature distance. i -v k This is the second feature distance.

[0121] Next, execute judgment S512 to check if the condition is met;

[0122] If the termination condition is met, proceed to step S512 to determine multiple fuzzy clusters; if the iteration condition is not met, return to step S508.

[0123] It is understandable that the above termination condition can be the convergence of the objective function J(U,V) or the reaching of the maximum number of iterations.

[0124] Through the above-described embodiments of this application, target accounts can be divided into different fuzzy clusters, each cluster representing a group of target accounts with similar characteristics. For example, target accounts can be divided into high-risk, high-spending groups, low-risk, low-spending groups, and medium-risk, medium-spending groups.

[0125] Suppose that after fuzzy clustering, c cluster centers are obtained, V = [v1, v2, ..., v3]. j ], and the corresponding fuzzy partitioning matrix U = [u i,j For any object account x i The degree to which it belongs to the j-th cluster is u. i,j Its weight in each cluster can be calculated using the following formula:

[0126]

[0127] Among them, u i,j Represents the object account x i The weights of the account operation features in the j-th cluster satisfy the following conditions:

[0128] The results of fuzzy clustering analysis provide input for the subsequent rule engine, enabling the rule engine to select the appropriate resource data processing method based on the classification results of the object account.

[0129] In one optional implementation, the process of determining at least one target data operation condition that matches the account operation feature from a plurality of candidate data operation conditions based on at least one target feature label and the label matching degree corresponding to each target feature label includes at least one of the following:

[0130] Method 1: If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one target feature label and label matching degree, and the target object account has performed the condition account operation within the condition period, then the current candidate data operation condition shall be determined as the target data operation condition.

[0131] Method 2: If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one target feature label and label matching degree, and the first account reference index of the target object account is within the condition index range, then the current candidate data operation condition is determined as the target data operation condition.

[0132] Method 3: If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one target feature label and label matching degree, and the changing trend of the second account reference indicator of the target object account meets the condition trend, then the current candidate data operation condition is determined as the target data operation condition.

[0133] Understandably, label matching degree can serve as an important input parameter in the rule matching process, assisting in the matching of conditions. For example, under the current conditions: if the target account belongs to a high-risk cluster and its weight in that cluster exceeds a certain threshold (e.g., 0.6), and has overdue payment records in the last three months, the credit limit should be significantly reduced; if the weight is between 0.3 and 0.6, the credit limit should be moderately reduced. By introducing label matching degree, the matching of conditions becomes more refined and flexible, more accurately reflecting the relationship between the target account and the cluster, thereby enabling the formulation of a more appropriate credit limit data processing method.

[0134] The above-described embodiments of this application provide a method for determining target data operation conditions based on target feature tags and tag matching degrees. These embodiments also allow for dynamic adjustment of credit resource limits based on the behavioral characteristics and risk level of the target account, adapting to changes in the target account's situation. The following will further explain these three methods in conjunction with specific credit resource scenarios:

[0135] In Method 1, for accounts exhibiting specific behavioral patterns, such as those belonging to a high-risk, high-spending group and having overdue payment records in the past three months, the credit limit is reduced. In this case, the account's spending behavior (high spending) and credit history (overdue payments) match specific characteristic tags (high risk, high spending) and tag matching degree, triggering the conditions for reducing the credit limit.

[0136] In Method 1, for situations where credit limits need to be adjusted based on certain key indicators of the target account, for example, if the target account belongs to a low-risk, low-consumption group and has a high credit score, the credit limit remains unchanged. Here, the target account's credit score serves as a reference indicator; if it falls within a certain range (e.g., the credit score is above a certain threshold), the credit limit remains unchanged. This method dynamically adjusts the credit limit based on the target account's credit status to reduce risk.

[0137] In Method 1, the credit limit needs to be adjusted based on the development trend of the target account. For example, if the target account belongs to a medium-risk, medium-consumption group and its income is steadily increasing, the credit limit will be increased. In this case, the target account's income growth trend serves as a reference indicator; if specific trend conditions are met (e.g., continuous income growth), the credit limit will be increased. This method dynamically adjusts the credit limit based on the target account's economic situation and development potential to support its growth.

[0138] Through the above-described embodiments of this application, credit limits can be dynamically adjusted based on the different characteristics and behavioral patterns of target accounts. By combining target feature tags and tag matching degrees, the risk level and credit status of target accounts can be identified more accurately, thereby formulating more suitable credit resource strategies. This method not only improves the intelligence level of credit resource management but also better adapts to market changes and dynamic changes in target account behavior, improving the accuracy and efficiency of credit resource decisions.

[0139] In one optional implementation, the above-described data processing method, which matches at least one target data operation condition, processes the current credit data of the target account, including one of the following:

[0140] Method 1: Obtain the condition priority coefficient corresponding to at least one target data operation condition; determine the conditions to be executed from at least one target data operation condition based on the condition priority coefficient; process the current credit data of the target account according to the data processing method matched by the conditions to be executed.

[0141] Method 2: Obtain the condition weight coefficients corresponding to at least one target data operation condition; weight the data operation values ​​corresponding to each of the at least one target data operation conditions according to the condition weight coefficients to obtain the target operation value; process the current credit data of the target account according to the data processing method matching the target operation value.

[0142] Method 3: Obtain the execution order coefficient corresponding to each of the at least one target data operation conditions; determine the target execution order of the data processing methods corresponding to each of the at least one target data operation conditions based on the execution order coefficients; process the current credit data of the target account according to the target execution order corresponding to each of the at least one data processing methods.

[0143] In the above embodiments of this application, the current credit data of the target account is processed according to the matching target data operation conditions, which can be achieved in one of the following three ways:

[0144] In Method 1, processing can be based on a condition priority coefficient. In this implementation, each target data operation condition is assigned a condition priority coefficient, which indicates which condition should be given priority when multiple conditions are met simultaneously. It is understood that operations under conditions with higher priority coefficients will be executed first.

[0145] In Method Two, processing can be based on conditional weighting coefficients. Specifically, a conditional weighting coefficient can be assigned to each target data operation condition, representing the importance of the condition in the final decision. Through weighted processing, a comprehensive target operation value can be obtained.

[0146] For example, the condition weight coefficient for each target data operation condition can be determined first. Then, based on the condition weight coefficient, the data operation value under each condition is weighted and calculated to obtain a target operation value. Finally, based on the calculated target operation value, the corresponding data processing method is matched to process the current credit data of the target account.

[0147] In Method 3, each target data operation condition is assigned an execution order coefficient to determine the execution order of conditions when multiple conditions need to be executed. Specifically, the execution order coefficient for each target data operation condition can be determined first. Then, based on the execution order coefficient, the execution order of each data processing method is determined. Finally, the current credit data of the target account is processed sequentially according to the determined execution order.

[0148] The above-described embodiments of this application provide a flexible mechanism for processing the current credit data of target accounts, allowing for the selection of the most suitable processing method based on different business needs and strategies. Method one emphasizes the priority of conditions, method two emphasizes the weight and overall impact of conditions, while method three emphasizes the execution order of conditions. In specific implementations, the appropriate method can be selected as needed, thereby enabling more precise control of the data processing flow to adapt to complex business rules and strategy requirements, thus improving the intelligence and automation level of data processing.

[0149] In an optional implementation, the process of determining at least one target data operation condition matching the account operation feature from multiple candidate data operation conditions based on at least one target feature label and the label matching degree corresponding to each target feature label includes:

[0150] S1, retrieve multiple first data operation conditions from the data cache, wherein the first data operation conditions are candidate data operation conditions with a matching hit rate greater than or equal to the target value;

[0151] S2, when at least one target data operation condition is included among multiple first data operation conditions, obtain at least one data processing method that matches the target data operation condition;

[0152] S3, if the target data operation condition is not included in the multiple first data operation conditions, obtain multiple second data operation conditions from the database, wherein the second data operation conditions are candidate data operation conditions with a matching hit rate less than the target value.

[0153] S4, based on at least one target feature label and the label matching degree corresponding to each of the at least one target feature label, determine at least one target data operation condition that matches the account operation feature from multiple second data operation conditions.

[0154] It is understandable that, in the above implementation, the execution speed of conditions can be improved by caching the regular conditions and corresponding execution methods in the server.

[0155] In the process of adjusting credit limits, there are some commonly used conditions, such as maintaining the existing credit limit for accounts with credit scores within a certain range and good repayment records. When new account data requires adjustment of credit limits, if the results of these commonly used conditions have been cached, the server does not need to perform the complete calculation process from data reading and condition judgment to conclusion again.

[0156] Taking a large bank that processes a large number of credit limit adjustment requests every day as an example, assuming that 30% of the requests can match common conditions, after caching the results of these conditions, the processing time of these requests can be reduced from the original average of 100 milliseconds to a negligible cache read time (may only take a few milliseconds), thereby significantly improving the overall processing speed.

[0157] Furthermore, many conditional judgments require accessing the database to retrieve various data about the target account, such as credit records and consumption behavior data. Caching the results of frequently used conditions reduces the server's need to constantly retrieve relevant data from the database for conditional judgments. For example, caching the results of frequently accessed conditions related to high-quality target accounts (such as high credit scores and stable income) can reduce the number of database queries, alleviate database access pressure, and thus improve the overall system response speed. Under high concurrency, this reduction in database access pressure can effectively prevent system performance degradation due to excessive database load, ensuring that credit limit adjustment requests are processed promptly.

[0158] Furthermore, in a server environment, multiple credit limit adjustment requests are typically processed simultaneously. When frequently used condition results are cached, multiple requests can quickly retrieve the cached results without queuing for condition calculation. This helps improve the system's concurrency capabilities, enabling the server to handle more credit limit adjustment requests per unit of time. For example, during major e-commerce promotions, a large number of merchants may apply for credit limit adjustments. Caching frequently used conditions allows the server to handle these concurrent requests more efficiently, avoiding long user wait times or request backlogs due to slow processing speeds.

[0159] In one specific implementation, the above-mentioned condition matching and data adjustment methods can be implemented through a rule engine.

[0160] In this embodiment, the rule engine can be a software system that allows users to define business logic in the form of rules and automatically execute these rules based on input data. In this embodiment, a rule engine is designed to select an appropriate credit resource data processing method based on the fuzzy clustering results of object accounts and other relevant factors.

[0161] The design of the rules engine includes the following aspects:

[0162] Rule Definition: Based on business needs and expert experience, define a series of rules to describe the credit resource data processing methods for different target account groups. For example, rules including but not limited to the following can be defined:

[0163] If the target account belongs to a high-risk, high-consumption group and has overdue payment records in the past three months, the credit limit will be reduced.

[0164] If the target account belongs to a low-risk, low-consumption group and has a high credit score, the credit limit will remain unchanged.

[0165] If the target account belongs to a medium-risk, medium-spending group and has a stable income growth, the credit limit will be increased.

[0166] Rule Priority: Define a priority for each rule to determine the order of execution in case of rule conflicts. Priorities can be determined based on factors such as business importance and risk level.

[0167] Rule execution: When inputting object account data, the rule engine will automatically execute the corresponding rules according to the rule definition and priority, and select the appropriate credit resource data processing method.

[0168] To improve the efficiency and accuracy of the rule engine, the following improvements can be made to enhance rule matching and execution: First, rules can be optimized by removing redundant and conflicting rules, improving readability and maintainability. Second, frequently used rule results can be cached to increase execution speed. Finally, machine learning and other methods can be used to automatically learn and optimize rules, improving the rule engine's adaptability and intelligence.

[0169] The following section provides a detailed explanation of the condition matching method implemented based on the aforementioned rule engine.

[0170] Let the rule set be R = [r1, r2, ..., r i ], where ri represents a rule. For any object account x i Its eigenvector is f(x) i )=[f1(x i ),f2(x i ),...,f m (x i )], where f k (x i ) represents the object account x i The k-th feature. Rule r i The conditional part can be represented as a Boolean function C(r)i ,f(x i If the condition is met, return true; otherwise, return false. Rule r i The conclusion can be represented as a function A(r) i ,f(x i )) is used to determine the data processing method for credit resources.

[0171] The execution process of the rule engine can be represented by the following formula:

[0172] For each r i in R:

[0173] if C(r i ,f(x i )):

[0174] return A(r i ,f(x i ))

[0175] The above formula specifically describes the following process: First, traverse the rule set: that is, traverse each rule r in the rule set R. i ;

[0176] Next, check the conditions: for each rule r i Check its conditional part C(r) i ,f(x i Check if the condition is met. If it is met (returns true), proceed to the next step; if it is not met (returns false), skip the rule and continue to check the next rule.

[0177] Execution conclusion: If the conditions are met, execute the conclusion part A(r) of the execution rule. i ,f(x i The processing of credit resource data is determined according to the strategy defined in the rules.

[0178] Through the above implementation methods, the design of the rule engine makes the selection of credit resource data processing methods more automated and intelligent, and can make decisions quickly and accurately based on the specific circumstances of the target account.

[0179] It should be noted that, in the embodiments of this application, the aforementioned data operation conditions and their corresponding data processing methods can be dynamically adjusted during execution. The following provides a further explanation of the dynamic adjustment method involved in this case.

[0180] Upon receiving a request to process target data, before obtaining the account operation characteristics of the target account, at least one of the following is also included:

[0181] Method 1: When the data processing method corresponding to at least one of the multiple candidate data operation conditions satisfies the first update condition, at least one data processing parameter used to indicate at least one data processing method is updated, wherein the data processing parameter includes a processing trend parameter used to indicate the data processing trend and a processing value parameter used to indicate the data processing value.

[0182] Understandably, in Method 1, the data processing parameters for data processing methods that meet the first update condition can be adjusted. For example, the adjustment could be the value of the credit resource or the processing trend of the credit resource (increase, decrease, or remain unchanged).

[0183] Method 2: If the data processing method corresponding to at least one of the multiple candidate data operation conditions satisfies the second update condition, at least one clustering parameter used to indicate the clustering method of multiple fuzzy clusterings is updated, wherein the clustering parameter includes cluster quantity parameter, fuzzy control coefficient, initial cluster center, and data preprocessing parameter.

[0184] In this embodiment, the parameters related to fuzzy clustering can also be adjusted if the data processing method meets the second update condition.

[0185] The cluster number parameter determines how many different groups the object accounts are divided into. If the value is too small, the division of object account groups will be too general and may not accurately reflect the differences between object accounts. Conversely, if the value is too large, the number of object accounts in each cluster will be too small, which may lead to over-segmentation, increasing the complexity of strategy formulation, and may also cause unstable clustering results due to data sparsity.

[0186] The fuzzy control coefficient controls the degree of fuzziness in clustering. When the value is small (close to 1), the fuzzy clustering result is closer to hard clustering, meaning each account tends to belong explicitly to a particular cluster, with a strong dependence on the cluster center. When the value is large, the fuzziness of the clustering increases, and an account can belong to multiple clusters simultaneously to a greater extent. While this better reflects the fuzziness and diversity of account characteristics, if it is too large, it may make the clustering results too fuzzy, losing the meaning of clustering and also affecting the formulation of the strategy.

[0187] Clustering algorithms are highly sensitive to the choice of initial cluster centers. Different initial cluster centers can lead to different clustering results. If the initial cluster centers are chosen inappropriately, the clustering result may get stuck in a local optimum instead of a global optimum.

[0188] Before performing fuzzy clustering, data preprocessing, such as normalization, is typically performed. The parameters used in the normalization process (such as the normalization range and method) affect the distribution and characteristics of the data. If the normalization parameters are not chosen appropriately, it may alter the original characteristic relationships of the data, leading to inaccurate clustering results.

[0189] It is understood that, in the embodiments of this application, the above parameters can be adjusted in a way that optimizes the evaluation indicators. The following is a detailed explanation of how the evaluation indicators are determined.

[0190] In an optional implementation, before obtaining the account operation characteristics of the target object account upon receiving a target data processing request, the method further includes:

[0191] S1, obtain a first set of accounts that match the first data processing method corresponding to the reference data operation conditions, and a first set of operation features corresponding to the first set of accounts, wherein the first set of accounts includes a first object account that processes the credit data according to the first data processing method, and the first set of operation features includes the first operation features corresponding to each of the first object accounts.

[0192] S2, based on the first set of operational features, determine at least one evaluation index that matches the first data processing method;

[0193] S3, determine the target evaluation index that matches the first data processing method based on the weighted summation result between at least one evaluation index;

[0194] S4-1, when the target evaluation index is within the first index range, determine that the first data processing method meets the first update condition;

[0195] S4-2, when the target evaluation index is within the second index range, determine that the first data processing method meets the second update condition.

[0196] In one optional implementation, the determination of at least one evaluation index matching the first data processing method based on the first set of operational features includes at least one of the following:

[0197] Method 1: Determine risk indicators based on the first set of operational features corresponding to the first set of accounts, wherein the risk indicators are used to indicate the degree of operational risk of the target institution allocating the total resource quota for the first object account;

[0198] Method 2: Determine the profit indicators based on the first set of operational features corresponding to the first set of accounts, wherein the risk indicators are used to indicate the degree of profit corresponding to the total resource allocation of the target institution for the first target account;

[0199] Method 3: Determine risk indicators based on the first set of operational features corresponding to the first set of accounts. In this case, after the target institution allocates a total resource quota for the first target account, the risk indicators are determined based on the feedback operations of the first target account.

[0200] Strategy evaluation involves assessing and providing feedback on the selected credit resource data processing methods to determine the effectiveness and rationality of the strategy. In this invention, multiple evaluation indicators are used to assess the credit resource data processing methods, including risk indicators, return indicators, and target account satisfaction indicators.

[0201] Among them, risk indicators are mainly used to assess the impact of credit resource data processing methods on bank risk. Indicators such as delinquency rate and non-performing loan rate can be used to measure the degree of risk. Let the delinquency rate be E and the non-performing loan rate be B, then the risk indicators can be expressed as:

[0202] Risk = αE + βB

[0203] Here, α and β are weighting coefficients, which are determined based on business needs and risk appetite.

[0204] Profitability metrics: These are primarily used to assess the impact of credit resource data processing methods on the profitability of financial entities. Indicators such as interest income and fee income can be used to measure profitability. Let interest income be I and fee income be F, then the profitability metrics can be expressed as:

[0205] Profit=γI+δF

[0206] Here, γ and δ are weighting coefficients, which are determined based on business needs and profit targets.

[0207] Account Satisfaction Index: This index is primarily used to assess the impact of credit resource data processing methods on account satisfaction. Indicators such as account complaint rate and account churn rate can be used to measure account satisfaction. Let the account complaint rate be C and the account churn rate be L, then the account satisfaction index can be expressed as:

[0208] Satisfaction=∈C+ζL

[0209] Here, ε and ζ are weighting coefficients, which are determined based on business needs and object account relationship management goals.

[0210] Comprehensive Evaluation Index: This index combines risk indicators, return indicators, and target account satisfaction indicators to arrive at a comprehensive evaluation index. A weighted average method can be used to calculate the comprehensive evaluation index; the formula is as follows:

[0211] Overall=ω1Risk+ω2Profit+ω3Satisfaction

[0212] Among them, ω1, ω2 and ω3 are weighting coefficients, which are determined according to business needs and decision-making objectives.

[0213] Strategy adjustment involves modifying and optimizing the data processing methods for credit resources based on the results of strategy evaluation. If the evaluation results indicate that the strategy is ineffective, it can be optimized by adjusting the rules of the rule engine, the parameters of fuzzy clustering, and other methods.

[0214] Let the adjusted strategy be S. ′ The strategy adjustment process can then be represented as:

[0215] S ′ =Adjust(S,Overall)

[0216] Here, Adjust is an adjustment function that adjusts strategy S based on the overall evaluation metric Overall.

[0217] The strategy evaluation and adjustment process is a closed-loop feedback system that can continuously optimize the data processing methods for credit resources and improve the accuracy and effectiveness of decision-making.

[0218] In an optional implementation, this application further employs intelligent dynamic adjustment. Specifically, this adjustment method can adjust the credit resource data processing strategy in real time according to market changes and dynamic changes in the behavior of the target account, thereby achieving intelligent credit resource management. Before obtaining the account operation characteristics of the target account upon receiving a target data processing request, the method further includes:

[0219] S1, obtain a second set of accounts that match the second data processing method corresponding to the reference data operation conditions, and a second set of operation features corresponding to the second set of accounts, wherein the second set of accounts includes second object accounts that process credit data according to the second data processing method, and the second set of operation features includes the second operation features corresponding to each of the second object accounts;

[0220] S2, based on the second set of operation features corresponding to the first operation cycle and the first trading platform description features corresponding to the first operation cycle, predict the third set of operation features corresponding to the second operation cycle and the second trading platform description features corresponding to the second operation cycle, wherein the second operation cycle is later than the first operation cycle;

[0221] S3, if the third set of operational features and the second trading platform description features satisfy the third update condition, update at least one data processing parameter of the second data processing method.

[0222] It should be noted that the adjustments to the data processing method described above are based on the results of strategy evaluation. Adjustments to the rules or fuzzy clustering parameters of the rule engine are only triggered when evaluation metrics (such as risk metrics, return metrics, and target account satisfaction metrics) show poor strategy performance. This adjustment is relatively passive, being a post-implementation adjustment based on existing evaluation results after the strategy has been implemented for a period of time.

[0223] In this embodiment, an intelligent dynamic adjustment strategy is adopted, with more diversified and real-time triggering conditions. It not only considers the results of strategy evaluation but also monitors the dynamic changes in market data (such as interest rate changes and economic conditions) and target account behavior data (such as consumption behavior and repayment behavior) in real time. Once a significant change in market or target account behavior is detected, an adjustment will be triggered even if the current strategy evaluation results do not yet indicate a negative outcome, in order to adapt to the changes in advance and avoid potential risks or losses. The triggering conditions of this embodiment are more forward-looking and timely.

[0224] The specific adjustment methods are explained below.

[0225] In an optional implementation, after predicting the third set of operation features corresponding to the second set of accounts corresponding to the second operation period and the second set of trading platform description features based on the second set of operation features corresponding to the second set of accounts corresponding to the first operation period and the first trading platform description features, the method further includes:

[0226] S1, obtain a set of reference operation features corresponding to the second operation cycle, and a set of reference trading platform description features corresponding to the second operation cycle. The set of reference operation features includes reference operation features determined based on the actual account operation of the second object account in the second operation cycle, and the reference trading platform description features are the actual trading platform description features corresponding to the second operation cycle.

[0227] S2, based on the first degree of difference between the reference operation feature set and the third operation feature set, and the second degree of difference between the reference trading platform description feature and the second trading platform description feature, determine that the third operation feature set and the second trading platform description feature satisfy the third update condition.

[0228] In an optional implementation, the above-mentioned prediction of a third set of operational features corresponding to a second operational period and a second set of trading platform description features corresponding to a second operational period, based on a second set of operational features corresponding to a first operational period and a first set of trading platform description features corresponding to the first operational period, wherein the second operational period is later than the first operational period, includes:

[0229] S1, input the second set of operation features and the first set of trading platform description features into at least one prediction model to obtain at least one prediction result, wherein the prediction result includes a third set of reference operation features corresponding to the second operation cycle and a second set of reference trading platform description features corresponding to the second operation cycle.

[0230] S2, based on at least one prediction result, determine a target prediction result, wherein the target prediction result includes a third set of operational features corresponding to the second operational cycle, and a second set of trading platform description features corresponding to the second operational cycle.

[0231] In the embodiments of this application, the above-mentioned intelligent dynamic adjustment process requires the execution of data monitoring operations, namely, real-time monitoring of market data and target account behavior data, including interest rate changes, financial trends, target account consumption behavior, repayment behavior, etc.

[0232] Next, based on the monitored data, trend forecasting techniques are used to predict future trends in the market and the behavior of target accounts. Methods such as time series analysis and machine learning can be employed for trend forecasting. For example, the ARIMA model can be used to predict interest rate changes, and neural network models can be used to predict the consumption behavior of target accounts.

[0233] Finally, based on the trend forecast results, the data processing method for credit resources is automatically adjusted. If a rise in market interest rates is predicted, the credit resource limit may need to be reduced to mitigate risk; if an increase in consumer spending by a target account is predicted, the credit resource limit may need to be increased to meet the account's needs.

[0234] Specifically, let the market data be M = [m1, m2, ..., m i The target account behavior data is C = [c1, c2, ..., c i ], where m i and c i Let M and C represent the i-th market data and the target account behavior data, respectively. The trend prediction function is P(M,C), used to predict the future trend of market and target account behavior. The strategy adjustment function is A(S,P(M,C)), used to adjust the credit resource data processing method based on the trend prediction results.

[0235] The process of intelligent dynamic adjustment can be represented by the following formula:

[0236] P = P(M, C)

[0237] S ′ =A(S,P)

[0238] To improve the accuracy and reliability of intelligent dynamic adjustment, the following optimization schemes can be further adopted:

[0239] Firstly, in the embodiments of this application, multi-source data fusion can be performed: data from multiple data sources, including internal data of financial entities, external data, market data, etc., can be fused to improve the comprehensiveness and accuracy of the data.

[0240] Furthermore, the results of multiple prediction models can be combined to improve the accuracy and reliability of trend prediction. For example, a weighted average method can be used to fuse the prediction results of different models.

[0241] Simultaneously, a feedback mechanism should be established to adjust and optimize the trend prediction model and strategy adjustment function based on actual results. For example, if the actual results do not match the predictions, the model parameters or the rules of the strategy adjustment function can be adjusted.

[0242] Let the fused prediction result be P. ′ The fusion function is F(P1,P2,…,P). n ), where P1, P2, ..., P n This represents the prediction results of different models. The feedback function is B(S,P) ′ Actual (,) is used to adjust the model and strategy based on the actual results.

[0243] The process of intelligent dynamic adjustment can be further represented as:

[0244] P1, P2, ..., P n =Predict(M,C)

[0245] P ′ =F(P1,P2,…,P) n )

[0246] S ′ =A(S,P) ′ )

[0247] Adjust = B(S,P) ′ Actual)

[0248] In the above implementation, different models (M) and parameters (C) are first used to predict certain characteristics or behaviors of the target account. Each model produces a prediction result, denoted as P1, P2, ..., P n .

[0249] Furthermore, model fusion operations can be performed. The fusion function F(P1,P2,…,P) n The prediction results from different models are integrated through a fusion function F to obtain a comprehensive prediction result P. ′This fusion process can employ weighted averaging, voting mechanisms, or other methods to improve the accuracy and reliability of predictions.

[0250] Furthermore, the adjustment function A(S,P) is adopted. ′ Based on the fused prediction result P ′ Given the current state S, calculate the new state S using the adjustment function A. ′ This new status reflects adjustments made to the credit limit or other relevant parameters of the target account based on the forecast results.

[0251] Finally, in the feedback phase, the feedback function B(S,P) ′ The system collects actual performance data (Actual) and uses a feedback function B to adjust the model and policy based on the actual performance. If the actual performance does not match the prediction results, the system will adjust the model parameters or the rules of the policy adjustment function to optimize future predictions and decisions.

[0252] Intelligent dynamic adjustment enables the credit resource data processing method to adapt to market changes and dynamic changes in the behavior of target accounts in a timely manner, thereby improving the intelligence level of credit resource management and the accuracy of decision-making.

[0253] The following combination Figure 6 The illustrated scheme architecture provides a complete description of one embodiment of this application. For example... Figure 6 As shown, this application can adjust the credit resources through the fuzzy clustering module 602, the rule engine 604, the first adjustment module 606, and the second adjustment module 608.

[0254] The fuzzy clustering module 602 can be used to perform fuzzy clustering on a set of object accounts. Based on the received data processing requests for the object accounts, it outputs fuzzy clustering results, i.e., the label matching degree corresponding to at least one of the target feature labels. By classifying object accounts through the fuzzy clustering algorithm executed in the fuzzy clustering module 602, the multi-dimensional characteristics of the object accounts are fully considered, making the classification results more accurate and personalized, thus providing a foundation for subsequent credit limit adjustment strategy selection.

[0255] Next, the fuzzy clustering results will be input into the rule engine 604 to perform condition matching and credit data processing. The rule engine automatically selects appropriate credit limit processing strategies based on the object account classification results and other relevant factors, improving the efficiency and accuracy of decision-making.

[0256] Finally, the results feedback from the rule engine 604 can be obtained based on the first adjustment module 606 and the second adjustment module 608. On the one hand, the first adjustment module 606 can adjust the credit limit processing strategy in real time according to market changes and dynamic changes in the behavior of the target account, so as to realize intelligent credit resource management; on the other hand, the second adjustment module 608 can execute corresponding strategy adjustments according to the fluctuations of specific indicators.

[0257] Understandably, the feedback results of the first adjustment module 606 and the second adjustment module 608 can also influence the clustering process in the fuzzy clustering module 602, thereby improving the accuracy and fitness of clustering.

[0258] Finally, the adjustment methods for the selected credit resources can be appropriately communicated to relevant personnel or systems. Reports, push notifications, and other methods can be used to notify credit resource managers and target accounts of the strategy results. Furthermore, relevant feedback can be received. For example, feedback from relevant personnel or systems regarding the strategy results can be collected to further optimize and improve the strategy. Feedback can be collected through questionnaires, target account complaint channels, etc. Finally, based on the feedback, the strategy can be adjusted and optimized to continuously improve its accuracy and effectiveness.

[0259] Through the above-described embodiments of this application, a personalized credit limit adjustment strategy is adopted. By using fuzzy clustering technology to accurately classify merchants, personalized credit limit adjustment strategies can be formulated for different types of merchants, meeting their individual needs and improving their satisfaction.

[0260] Furthermore, the use of rules engines automates and intelligently selects credit limit adjustment strategies, enabling rapid and accurate decision-making based on merchant characteristics and market changes, thus improving efficiency and accuracy. Moreover, through risk indicator assessment and strategy adjustments, credit resource risks can be identified and mitigated in a timely manner, safeguarding the funds of financial institutions.

[0261] Finally, the intelligent dynamic adjustment mechanism can monitor market changes and changes in merchant behavior in real time, adjust credit limit adjustment strategies in a timely manner, adapt to market changes, and improve the competitiveness of financial institutions.

[0262] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0263] According to another aspect of the present invention, an apparatus for determining object data for implementing the above-described object data determination method is also provided. For example... Figure 7 As shown, the device includes:

[0264] The acquisition unit 702 is used to acquire the account operation characteristics of the target object account when a target data processing request is received, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period.

[0265] The first determining unit 704 is used to determine, based on the account operation characteristics, at least one target feature tag that matches the target object account, and the tag matching degree corresponding to each of the at least one target feature tag.

[0266] The second determining unit 706 is used to determine at least one target data operation condition matching the account operation feature from multiple candidate data operation conditions based on at least one of the target feature labels and the label matching degree corresponding to each of the target feature labels.

[0267] The processing unit 708 is configured to process the current credit data of the target object account according to a data processing method that matches at least one of the target data operation conditions to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

[0268] Optionally, the first determining unit 704 is configured to: acquire cluster center features corresponding to each of the plurality of fuzzy clusters, wherein the plurality of fuzzy clusters correspond to a plurality of operation feature labels; determine the feature distance between the account operation feature and the plurality of cluster center features; determine the cluster matching degree between the account operation feature and the plurality of fuzzy clusters based on the plurality of feature distances; and determine at least one target feature label that matches the target object account, and the label matching degree corresponding to each of the at least one target feature label, based on the plurality of cluster matching degrees.

[0269] Optionally, the first determining unit 704 is configured to: obtain fuzzy control coefficients, wherein the fuzzy control coefficients determine the distribution of the cluster matching degree; obtain a first feature distance between the account operation feature and the current cluster center feature among the multiple cluster center features, wherein the current cluster center feature is a cluster center feature that matches the current fuzzy cluster; obtain multiple second feature distances between the account operation feature and the multiple cluster center features; and determine the current cluster matching degree between the account operation feature and the current fuzzy cluster based on the first feature distance, the multiple second feature distances, and the fuzzy control coefficients.

[0270] Optionally, the first determining unit 704 is configured to: determine a matching reference value based on the cumulative result of multiple cluster matching degrees; determine the cluster label corresponding to the current fuzzy cluster among multiple fuzzy clusters as the target feature label; and determine the ratio between the current cluster matching degree corresponding to the current fuzzy cluster and the matching reference value as the label matching degree corresponding to the target feature label.

[0271] Optionally, the above-mentioned object data determining device further includes: a clustering unit, used to obtain an operation feature set determined based on the object account set; obtain clustering control coefficients, wherein the clustering control coefficients are used to determine the number of clusters in the fuzzy clusters; determine a fuzzy partitioning matrix based on the feature distance between every two account operation features in the operation feature set and the clustering control coefficients, wherein the fuzzy partitioning matrix is ​​used to indicate the clustering matching degree between any one of the account operation features in the operation feature set and multiple fuzzy clusters; and determine the cluster center features corresponding to each of the multiple fuzzy clusters when the fuzzy partitioning matrix satisfies the target clustering conditions.

[0272] Optionally, the second determining unit 706 is configured to: determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the target account has performed a conditional account operation within the conditional period; determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the first account reference indicator of the target account is within the conditional indicator range; or determine the current candidate data operation condition as the target data operation condition if the feature label condition and label matching degree condition indicated by the current candidate data operation condition among the plurality of candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the changing trend of the second account reference indicator of the target account satisfies the conditional trend.

[0273] Optionally, the processing unit 708 is configured to: acquire a condition priority coefficient corresponding to at least one of the target data operation conditions; determine a condition to be executed from the at least one target data operation condition based on the condition priority coefficient; process the current credit data of the target account according to the data processing method matched by the condition to be executed; acquire a condition weight coefficient corresponding to at least one of the target data operation conditions; perform weighted processing on the data operation value corresponding to at least one of the target data operation conditions based on the condition weight coefficient to obtain a target operation value; process the current credit data of the target account according to the data processing method matched by the target operation value; acquire an execution order coefficient corresponding to at least one of the target data operation conditions; determine a target execution order from the data processing methods corresponding to at least one of the target data operation conditions based on the execution order coefficient; and process the current credit data of the target account according to the target execution order corresponding to at least one of the data processing methods.

[0274] Optionally, the processing unit 708 is configured to: obtain multiple first data operation conditions from a data cache, wherein the first data operation conditions are candidate data operation conditions with a matching hit rate greater than or equal to a target value; if the multiple first data operation conditions include at least one target data operation condition, obtain at least one data processing method matching the target data operation condition; if the multiple first data operation conditions do not include the target data operation condition, obtain multiple second data operation conditions from a database, wherein the second data operation conditions are candidate data operation conditions with a matching hit rate less than the target value; and determine at least one target data operation condition matching the account operation feature from the multiple second data operation conditions based on at least one target feature tag and the tag matching degree corresponding to each of the at least one target feature tag.

[0275] Optionally, the above-mentioned object data determination device further includes at least one of the following: a first update unit, configured to update at least one data processing parameter indicating at least one of the above-mentioned data processing methods when the data processing method corresponding to at least one of the above-mentioned candidate data operation conditions satisfies the first update condition, wherein the above-mentioned data processing parameter includes a processing trend parameter indicating a data processing trend and a processing value parameter indicating a data processing value; and a second update unit, configured to update at least one clustering parameter indicating a clustering method of the above-mentioned fuzzy clustering when the data processing method corresponding to at least one of the above-mentioned candidate data operation conditions satisfies the second update condition, wherein the above-mentioned clustering parameter includes a cluster quantity parameter, a fuzzy control coefficient, an initial cluster center, and a data preprocessing parameter.

[0276] Optionally, the aforementioned object data determination device further includes: a condition judgment unit, configured to acquire a first set of accounts matching the first data processing method corresponding to the aforementioned reference data operation conditions, and a first set of operation features corresponding to the aforementioned first set of accounts, wherein the aforementioned first set of accounts includes first object accounts that process credit data according to the aforementioned first data processing method, and the aforementioned first set of operation features includes first operation features corresponding to each of the aforementioned first object accounts; based on the aforementioned first set of operation features, determine at least one evaluation indicator matching the aforementioned first data processing method; based on the weighted summation result between at least one of the aforementioned evaluation indicators, determine a target evaluation indicator matching the aforementioned first data processing method; if the aforementioned target evaluation indicator is within a first indicator range, determine that the aforementioned first data processing method satisfies the aforementioned first update condition; if the aforementioned target evaluation indicator is within a second indicator range, determine that the aforementioned first data processing method satisfies the aforementioned second update condition.

[0277] Optionally, the aforementioned condition judgment unit is used for at least one of the following: determining a risk indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein the risk indicator is used to indicate the degree of operational risk of the target institution configuring the aforementioned total resource quota for the aforementioned first target account; determining a return indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein the risk indicator is used to indicate the degree of return corresponding to the aforementioned target institution configuring the aforementioned total resource quota for the aforementioned first target account; determining a risk indicator based on a first set of operational features corresponding to the aforementioned first set of accounts, wherein after the target institution configures the aforementioned total resource quota for the aforementioned first target account, the risk indicator is determined based on the feedback operation of the aforementioned first target account.

[0278] Optionally, the aforementioned object data determining device further includes: a third updating unit, configured to acquire a second set of accounts matching the second data processing method corresponding to the aforementioned reference data operation conditions, and a second set of operation features corresponding to the aforementioned second set of accounts, wherein the aforementioned second set of accounts includes second object accounts that process credit data according to the aforementioned second data processing method, and the aforementioned second set of operation features includes second operation features corresponding to each of the aforementioned second object accounts; predict a third set of operation features corresponding to the second operation period and a second set of transaction platform description features corresponding to the aforementioned second operation period based on the aforementioned second set of operation features corresponding to the first operation period and the aforementioned first set of transaction platform description features corresponding to the aforementioned first operation period, wherein the aforementioned second operation period is later than the aforementioned first operation period; and update at least one data processing parameter of the aforementioned second data processing method when the aforementioned third set of operation features and the aforementioned second set of transaction platform description features satisfy the third updating condition.

[0279] Optionally, the third update unit is further configured to: obtain a set of reference operation features corresponding to the second operation period and a set of reference trading platform description features corresponding to the second operation period, wherein the set of reference operation features includes reference operation features determined based on the actual account operations of the second object account in the second operation period, and the reference trading platform description features are the actual trading platform description features corresponding to the second operation period; and determine that the third operation feature set and the second trading platform description features satisfy the third update condition based on a first degree of difference between the set of reference operation features and the third set of operation features, and a second degree of difference between the reference trading platform description features and the second trading platform description features.

[0280] Optionally, the third update unit is configured to: input the second set of operational features and the first trading platform description features into at least one prediction model to obtain at least one prediction result, wherein the prediction result includes a third set of reference operational features corresponding to the second operational period and a second set of reference trading platform description features corresponding to the second operational period; and determine a target prediction result based on at least one of the prediction results, wherein the target prediction result includes the third set of operational features corresponding to the second operational period and the second set of reference trading platform description features corresponding to the second operational period.

[0281] Optionally, in this embodiment, the implementation of each of the above-mentioned unit modules can be referred to the above-mentioned method embodiments, which will not be repeated here.

[0282] According to another aspect of the present invention, an electronic device for implementing the above-described method for determining object data is also provided. This electronic device may be... Figure 8 The terminal device or server shown. This embodiment uses this electronic device as an example for illustration. Figure 8 As shown, the electronic device includes a memory 802 and a processor 804. The memory 802 stores a computer program, and the processor 804 is configured to execute the steps in any of the above method embodiments via the computer program.

[0283] Optionally, in this embodiment, the aforementioned electronic device may be located in at least one of a plurality of network devices in a computer network.

[0284] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0285] S1, Upon receiving a target data processing request, obtain the account operation characteristics of the target object account, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period.

[0286] S2, based on the above account operation characteristics, determine at least one target feature tag that matches the above target object account, and the tag matching degree corresponding to each of the above target feature tags;

[0287] S3, based on at least one of the above-mentioned target feature labels and the label matching degree corresponding to each of the above-mentioned target feature labels, determine at least one target data operation condition matching the above-mentioned account operation feature from multiple candidate data operation conditions;

[0288] S4, process the current credit data of the target object account according to the data processing method that matches at least one of the above target data operation conditions to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

[0289] Alternatively, as those skilled in the art will understand, Figure 8 The structure shown is for illustrative purposes only. Electronic devices can also be in-vehicle terminals, smartphones (such as Android phones, iOS phones, etc.), tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 8 This does not limit the structure of the aforementioned electronic devices. For example, the electronic device may also include components that are more... Figure 8 The more or fewer components shown (such as network interfaces, etc.), or having the same Figure 8The different configurations shown.

[0290] The memory 802 can be used to store software programs and modules, such as the program instructions / modules corresponding to the object data determination method and apparatus in this embodiment of the invention. The processor 804 executes various functional applications and data processing by running the software programs and modules stored in the memory 802, thereby realizing the aforementioned object data determination method. The memory 802 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 802 may further include memory remotely located relative to the processor 804, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. Specifically, the memory 802 may be used, but is not limited to, to store file information such as target logical files. As an example, such as Figure 8 As shown, the memory 802 may include, but is not limited to, the acquisition unit 702, the first determination unit 704, the second determination unit 706, and the processing unit 708 in the object data determination device. Furthermore, it may include, but is not limited to, other module units in the object data determination device, which will not be described further in this example.

[0291] Optionally, the transmission device 806 described above is used to receive or send data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 806 includes a Network Interface Controller (NIC), which can be connected to other network devices and a router via a network cable to communicate with the Internet or a local area network. In another example, the transmission device 806 is a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0292] In addition, the above-mentioned electronic device also includes a display 808 and a connection bus 810 for connecting the various module components in the above-mentioned electronic device.

[0293] In other embodiments, the aforementioned terminal device or server can be a node in a distributed system, wherein the distributed system can be a blockchain system, which is a distributed system formed by connecting multiple nodes through network communication. The nodes can form a peer-to-peer network, and any form of computing device, such as a server, terminal, or other electronic device, can become a node in the blockchain system by joining this peer-to-peer network.

[0294] According to one aspect of this application, a computer program product is provided, comprising a computer program / instructions containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions provided in embodiments of this application.

[0295] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0296] According to one aspect of this application, a computer-readable storage medium is provided, wherein a processor of a computer device reads computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform the aforementioned method for determining object data.

[0297] Optionally, in this embodiment, the computer-readable storage medium may be configured to store a computer program for performing the following steps:

[0298] S1, Upon receiving a target data processing request, obtain the account operation characteristics of the target object account, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period.

[0299] S2, based on the above account operation characteristics, determine at least one target feature tag that matches the above target object account, and the tag matching degree corresponding to each of the above target feature tags;

[0300] S3, based on at least one of the above-mentioned target feature labels and the label matching degree corresponding to each of the above-mentioned target feature labels, determine at least one target data operation condition matching the above-mentioned account operation feature from multiple candidate data operation conditions;

[0301] S4, process the current credit data of the target object account according to the data processing method that matches at least one of the above target data operation conditions to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

[0302] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0303] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0304] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0305] In the several embodiments provided in this application, it should be understood that the disclosed object account can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0306] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0307] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0308] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for determining object data, characterized in that, include: Upon receiving a target data processing request, the account operation characteristics of the target object account are obtained, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period. Based on the account operation characteristics, determine at least one target feature tag that matches the target account, and the tag matching degree corresponding to each of the at least one target feature tag; Based on at least one target feature label and the label matching degree corresponding to each of the target feature labels, at least one target data operation condition matching the account operation feature is determined from multiple candidate data operation conditions; The current credit data of the target object account is processed according to a data processing method that matches at least one of the target data operation conditions to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

2. The method according to claim 1, characterized in that, The step of determining at least one target feature tag matching the target account based on the account operation characteristics, and the tag matching degree corresponding to each of the at least one target feature tag, includes: Obtain the cluster center features corresponding to each of the multiple fuzzy clusters, where the multiple fuzzy clusters correspond to multiple operational feature labels; Determine the feature distance between the account operation features and the multiple cluster center features; The clustering matching degree between the account operation features and the multiple fuzzy clusters is determined based on the multiple feature distances; Based on multiple clustering matching degrees, at least one target feature tag that matches the target object account is determined, as well as the tag matching degree corresponding to each of the at least one target feature tag.

3. The method according to claim 2, characterized in that, The step of determining the clustering matching degree between the account operation features and the multiple fuzzy clusters based on the multiple feature distances includes: Obtain fuzzy control coefficients, wherein the fuzzy control coefficients determine the distribution of the cluster matching degree; Obtain the first feature distance between the account operation feature and the current cluster center feature among multiple cluster center features, wherein the current cluster center feature is the cluster center feature that matches the current fuzzy cluster; Obtain multiple second feature distances between the account operation features and multiple cluster center features; Based on the first feature distance, multiple second feature distances, and the fuzzy control coefficient, the current clustering matching degree between the account operation feature and the current fuzzy cluster is determined.

4. The method according to claim 2, characterized in that, The step of determining at least one target feature tag that matches the target object account based on multiple clustering matching degrees, and the tag matching degree corresponding to each of the at least one target feature tag, includes: A matching reference value is determined based on the sum of multiple cluster matching scores; The cluster label corresponding to the current fuzzy cluster among multiple fuzzy clusters is determined as the target feature label; The ratio between the current cluster matching degree corresponding to the current fuzzy cluster and the matching reference value is determined as the label matching degree corresponding to the target feature label.

5. The method according to claim 2, characterized in that, Before obtaining the account operation characteristics of the target object account upon receiving the target data processing request, the process further includes: Obtain the set of operational characteristics determined based on the object account set; Obtain clustering control coefficients, wherein the clustering control coefficients are used to determine the number of clusters in the fuzzy clustering; Based on the feature distance between every two account operation features in the operation feature set and the clustering control coefficient, a fuzzy partitioning matrix is ​​determined, wherein the fuzzy partitioning matrix is ​​used to indicate the clustering matching degree between any one of the account operation features in the operation feature set and multiple fuzzy clusters; If the fuzzy partitioning matrix satisfies the target clustering conditions, the cluster center features corresponding to each of the multiple fuzzy clusters are determined.

6. The method according to claim 1, characterized in that, The step of determining at least one target data operation condition matching the account operation feature from multiple candidate data operation conditions based on at least one target feature label and the label matching degree corresponding to each of the at least one target feature label includes at least one of the following: If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the target object account has performed a condition account operation within the condition period, then the current candidate data operation condition shall be determined as the target data operation condition. If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the first account reference index of the target object account is within the condition index range, then the current candidate data operation condition is determined as the target data operation condition. If the feature label condition and label matching degree condition indicated by the current candidate data operation condition among multiple candidate data operation conditions match at least one of the target feature labels and the label matching degree, and the changing trend of the second account reference index of the target object account satisfies the condition trend, then the current candidate data operation condition is determined as the target data operation condition.

7. The method according to claim 6, characterized in that, The process of processing the current credit data of the target account according to a data processing method that matches at least one of the target data operation conditions includes one of the following: Obtain the condition priority coefficient corresponding to at least one of the target data operation conditions; determine the conditions to be executed from at least one of the target data operation conditions based on the condition priority coefficients; The current credit data of the target account is processed according to the data processing method matched by the conditions to be executed. Obtain the condition weight coefficient corresponding to at least one of the target data operation conditions; perform weighted processing on the data operation values ​​corresponding to at least one of the target data operation conditions according to the condition weight coefficients to obtain the target operation value; process the current credit data of the target object account according to the data processing method matching the target operation value; Obtain the execution order coefficient corresponding to at least one of the target data operation conditions; Based on the execution order coefficient, determine the target execution order of the data processing methods corresponding to at least one of the target data operation conditions; process the current credit data of the target object account according to the target execution order corresponding to at least one of the data processing methods.

8. The method according to claim 6, characterized in that, The step of determining at least one target data operation condition matching the account operation feature from multiple candidate data operation conditions based on at least one target feature label and the label matching degree corresponding to each of the at least one target feature label includes: Multiple first data operation conditions are obtained from the data cache, wherein the first data operation conditions are the candidate data operation conditions whose matching hit rate is greater than or equal to the target value; In the case where at least one of the target data operation conditions is included among multiple first data operation conditions, at least one data processing method matching the target data operation condition is obtained; If the target data operation condition is not included in the plurality of first data operation conditions, a plurality of second data operation conditions are obtained from the database, wherein the second data operation conditions are the candidate data operation conditions whose matching hit rate is less than the target value; Based on at least one of the target feature tags and the tag matching degree corresponding to each of the target feature tags, at least one of the target data operation conditions that match the account operation features is determined from a plurality of second data operation conditions.

9. The method according to claim 2, characterized in that, Upon receiving a request to process target data, before obtaining the account operation characteristics of the target account, at least one of the following is also included: If the data processing method corresponding to at least one of the candidate data operation conditions satisfies the first update condition, at least one data processing parameter used to indicate at least one of the data processing methods is updated, wherein the data processing parameter includes a processing trend parameter used to indicate a data processing trend and a processing value parameter used to indicate a data processing value. If the data processing method corresponding to at least one of the candidate data operation conditions satisfies the second update condition, at least one clustering parameter used to indicate the clustering method of the multiple fuzzy clusterings is updated, wherein the clustering parameter includes a cluster number parameter, a fuzzy control coefficient, an initial cluster center, and a data preprocessing parameter.

10. The method according to claim 9, characterized in that, Upon receiving a target data processing request, before obtaining the account operation characteristics of the target object account, the process also includes: Obtain a first set of accounts that match the first data processing method corresponding to the reference data operation conditions, and a first set of operation features corresponding to the first set of accounts. The first set of accounts includes a first object account that processes the credit data according to the first data processing method, and the first set of operation features includes the first operation features corresponding to each of the first object accounts. Based on the first set of operational features, determine at least one evaluation index that matches the first data processing method; Based on the weighted summation result among at least one of the evaluation indicators, a target evaluation indicator matching the first data processing method is determined. If the target evaluation index is within the first index range, it is determined that the first data processing method satisfies the first update condition. If the target evaluation index is within the second index range, it is determined that the first data processing method satisfies the second update condition.

11. The method according to claim 10, characterized in that, The step of determining at least one evaluation index matching the first data processing method based on the first set of operational features includes at least one of the following: Based on the first set of operational features corresponding to the first set of accounts, a risk indicator is determined, wherein the risk indicator is used to indicate the degree of operational risk of the target institution configuring the total resource quota for the first object account; Based on the first set of operational features corresponding to the first set of accounts, a revenue indicator is determined, wherein the risk indicator is used to indicate the degree of revenue corresponding to the total resource quota configured by the target institution for the first object account; Risk indicators are determined based on a first set of operational features corresponding to the first set of accounts, wherein the risk indicators are determined based on the feedback operations of the first set of accounts after the target institution configures the total resource quota for the first target account.

12. The method according to claim 9, characterized in that, Upon receiving a target data processing request, before obtaining the account operation characteristics of the target object account, the process also includes: Obtain a second set of accounts that matches the second data processing method corresponding to the reference data operation conditions, and a second set of operation features corresponding to the second set of accounts. The second set of accounts includes second object accounts that process credit data according to the second data processing method, and the second set of operation features includes the second operation features corresponding to each of the second object accounts. Based on the second set of operation features corresponding to the first operation cycle and the first trading platform description features corresponding to the first operation cycle, predict the third set of operation features corresponding to the second operation cycle and the second trading platform description features corresponding to the second operation cycle, wherein the second operation cycle is later than the first operation cycle; If the third set of operational features and the second transaction platform description features meet the third update condition, at least one data processing parameter of the second data processing method is updated.

13. The method according to claim 12, characterized in that, After predicting the third set of operation features corresponding to the second set of accounts corresponding to the second operation period, and the second set of trading platform description features, based on the second set of operation features corresponding to the second set of accounts corresponding to the first operation period and the first trading platform description features, the method further includes: Obtain a set of reference operation features corresponding to the second operation period, and a set of reference trading platform description features corresponding to the second operation period. The set of reference operation features includes reference operation features determined based on the actual account operation of the second object account in the second operation period, and the reference trading platform description features are the actual trading platform description features corresponding to the second operation period. Based on the first degree of difference between the reference operation feature set and the third operation feature set, and the second degree of difference between the reference trading platform description feature and the second trading platform description feature, it is determined that the third operation feature set and the second trading platform description feature satisfy the third update condition.

14. The method according to claim 12, characterized in that, The step of predicting a third set of operation features corresponding to a second operation period and a second set of trading platform description features corresponding to a second operation period based on the second set of operation features corresponding to a first operation period and the first set of trading platform description features corresponding to the first operation period, wherein the second operation period is later than the first operation period, includes: The second set of operational features and the first set of trading platform description features are input into at least one prediction model to obtain at least one prediction result, wherein the prediction result includes a third set of reference operational features corresponding to the second operational cycle and a second set of reference trading platform description features corresponding to the second operational cycle. A target prediction result is determined based on at least one of the prediction results, wherein the target prediction result includes the third set of operational features corresponding to the second operational cycle, and the second trading platform description features corresponding to the second operational cycle.

15. A device for determining object data, characterized in that, include: The acquisition unit is used to acquire the account operation characteristics of the target object account when a target data processing request is received, wherein the account operation characteristics are determined based on at least one account operation of the target object account within the target period. The first determining unit is configured to determine, based on the account operation characteristics, at least one target feature tag that matches the target object account, and the tag matching degree corresponding to each of the at least one target feature tag; The second determining unit is used to determine at least one target data operation condition that matches the account operation feature from multiple candidate data operation conditions based on at least one target feature label and the label matching degree corresponding to each of the at least one target feature label. The processing unit is configured to process the current credit data of the target object account according to a data processing method that matches at least one of the target data operation conditions to obtain target credit data, wherein the target credit data is used to indicate the total resource quota that the target object account is allowed to operate in the target interaction activity.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method described in any one of claims 1 to 14.

17. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1 to 14.

18. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 14 through the computer program.

Images