Loan consumption fund use monitoring method and device and electronic equipment

Abstract

The invention provides a consumption loan fund use monitoring method and device and electronic equipment. In the scheme, a first clustering algorithm is used to extract a plurality of typical daily expenditure curves of each user in different time periods from a plurality of pieces of user limit expenditure data, and a typical daily expenditure curve of a user group is determined according to the typical daily expenditure curves; performing similarity comparison on the real-time daily support curve of the day of application support of each user and a plurality of typical daily support curves of the user, and selecting a minimum similarity comparison result; performing trend comparison on the real-time daily expenditure curve of each user and the typical daily expenditure curve of the user group, and selecting a maximum trend comparison result; and comparing the minimum similarity comparison result and the maximum trend comparison result with corresponding preset thresholds, and monitoring whether the supporting behavior of the user is abnormal or not. According to the scheme, the problem that in the prior art, the identification capability on the abnormal fund use risk is limited in the consumption loan expenditure application stage is solved.

Description

Technical Field

[0001] The present application relates to the technical field of data analysis and processing. Specifically, it relates to a method, device, computer-readable storage medium, and electronic device for monitoring the use of funds for consumer loans. Background Art

[0003] In the prior art, the flow of loan funds is monitored through a graph theory model or a decision tree model. This method only realizes the monitoring of the post-loan fund flow and cannot expose risks in advance. That is, for abnormal transactions that have already occurred, it is impossible to fully recover, resulting in inevitable economic losses. In addition, the prior art conducts a risk assessment of loan disbursement based on the historical data of accounts and account groups. The value cycle of the historical data used is short, and the data is accidental and cannot accurately represent the behavior habits of the target account and the account group where it is located. At the same time, the determination of the account group where the account is located in this method is only based on the quota interval where the disbursement quota of the current account application is located, and the judgment factor is single. Therefore, it is impossible to accurately identify the risk of abnormal fund use. Summary of the Invention

[0004] The main purpose of the present application is to provide a method, device, computer-readable storage medium, and electronic device for monitoring the use of funds for consumer loans, so as to at least solve the problem of limited ability to identify the risk of abnormal fund use in the consumer loan disbursement application stage in the prior art.

[0005] To achieve the above object, according to one aspect of the present application, there is provided a method for monitoring the use of funds for consumer loans, including: obtaining user quota disbursement data of multiple users, and performing clustering analysis on the user quota disbursement data of multiple users using a first clustering algorithm to extract multiple typical daily disbursement curves of each user in different time periods, where the time periods include a time period composed of multiple working days within a preset time period and a time period composed of multiple holidays within the preset time period; determining a typical daily disbursement curve of the user group according to the typical daily disbursement curves of all users; comparing the real-time daily disbursement curve of each user on the day of applying for disbursement with the multiple typical daily disbursement curves of the user to obtain multiple similarity comparison results, and selecting the minimum similarity comparison result, where the minimum similarity comparison result is the minimum value among the multiple similarity comparison results; comparing the trend of the real-time daily disbursement curve of each user with the typical daily disbursement curve of the user group to obtain multiple trend comparison results, and selecting the maximum trend comparison result, where the maximum trend comparison result is the maximum value among the multiple trend comparison results; comparing the minimum similarity comparison result and the maximum trend comparison result with corresponding preset thresholds respectively to monitor whether there is an abnormality in the disbursement behavior of the user.

[0006] According to another aspect of the present application, there is provided a monitoring device for the use of funds for consumer loans, including: an acquisition unit configured to acquire the user quota usage data of multiple users, and perform clustering analysis on the user quota usage data of the multiple users using a first clustering algorithm to extract multiple typical daily usage curves of each user in different time periods, where the time periods include a time period composed of multiple working days within a preset time period and a time period composed of multiple holidays within the preset time period; a first determination unit configured to determine the typical daily usage curve of the user group according to the typical daily usage curves of all the users; a similarity comparison unit configured to compare the real-time daily usage curve of each user on the day of applying for the quota with the multiple typical daily usage curves of the user to obtain multiple similarity comparison results, and select the minimum similarity comparison result, where the minimum similarity comparison result is the minimum value among the multiple similarity comparison results; a trend comparison unit configured to compare the trend of the real-time daily usage curve of each user with the typical daily usage curve of the user group to obtain multiple trend comparison results, and select the maximum trend comparison result, where the maximum trend comparison result is the maximum value among the multiple trend comparison results; a comparison unit configured to compare the minimum similarity comparison result and the maximum trend comparison result with corresponding preset thresholds respectively to monitor whether there is an abnormality in the usage behavior of the user.

[0007] According to yet another aspect of the present application, there is provided a computer-readable storage medium, where the computer-readable storage medium includes a stored program, and when the program runs, it controls the device where the computer-readable storage medium is located to execute any one of the methods for monitoring the use of funds for consumer loans.

[0008] According to still another aspect of the present application, there is provided an electronic device, including: one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for monitoring the use of funds for consumer loans.

[0009] Applying the technical solution of the present application, obtaining the user quota utilization data of multiple users, and using the first clustering algorithm to perform clustering analysis on the user quota utilization data of multiple users, extracting multiple typical daily utilization curves of each user in different time periods, where the time periods include the time period composed of multiple working days within a preset time period and the time period composed of multiple holidays within a preset time period; determining the typical daily utilization curve of the user group according to the typical daily utilization curves of all users; comparing the real-time daily utilization curve on the day when each user applies for utilization with the multiple typical daily utilization curves of the user to obtain multiple similarity comparison results, and selecting the minimum similarity comparison result, where the minimum similarity comparison result is the minimum value among the multiple similarity comparison results; comparing the trend of the real-time daily utilization curve of each user with the typical daily utilization curve of the user group to obtain multiple trend comparison results, and selecting the maximum trend comparison result, where the maximum trend comparison result is the maximum value among the multiple trend comparison results; comparing the minimum similarity comparison result and the maximum trend comparison result with the corresponding preset thresholds respectively to monitor whether there are abnormalities in the utilization behavior of the user. This solution immediately performs risk assessment on the real-time utilization application data of the user on the day of applying for utilization, overcomes the lag of traditional post-loan fund monitoring in the prior art, and at the same time extracts the utilization behavior habit curves of users and user groups based on a large amount of historical user quota utilization data, making the extracted reference curves more referenceable and enabling more accurate prediction of whether there are abnormalities in the current utilization of the user, thus solving the problem of limited ability to identify the risk of abnormal fund use in the consumption loan utilization application stage in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0011] Figure 1 The hardware structure block diagram of a mobile terminal showing a method for monitoring the use of funds for consumer loans provided in an embodiment of the present application is shown;

[0012] Figure 2 The flowchart showing a method for monitoring the use of funds for consumer loans provided in an embodiment of the present application is shown;

[0013] Figure 3 The flowchart showing a specific method for monitoring the use of funds for consumer loans provided in an embodiment of the present application is shown;

[0014] Figure 4 The structure block diagram of a device for monitoring the use of funds for consumer loans provided in an embodiment of the present application is shown.

[0015] Among them, the above-mentioned drawings include the following reference numerals:

[0016] 102, processor; 104, memory; 106, transmission device; 108, input / output device. Detailed implementation manners

[0017] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0018] In order to enable those skilled in the art of the present technology to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present application.

[0019] It should be noted that the terms "first", "second", etc. in the specification, claims and drawings of the present application are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances for the embodiments of the present application described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0020] As introduced in the background art, in the prior art, the monitoring of the fund flow is usually only carried out after the loan, and the historical data used has a short value-taking period and a single judgment factor, and it is impossible to accurately identify the risk of abnormal fund use. To solve the problem that the ability to identify the risk of abnormal fund use in the consumer loan disbursement application stage in the prior art is limited, the embodiments of the present application provide a method, device, computer-readable storage medium and electronic device for monitoring the fund use of consumer loans.

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention.

[0022] The method embodiments provided in the embodiments of the present application can be executed on a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 is a hardware structure block diagram of a mobile terminal of a method for monitoring the fund use of consumer loans according to an embodiment of the present invention. AsFigure 1 As shown, the mobile terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Among them, the above mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown in the figure is only schematic and does not limit the structure of the above mobile terminal. For example, the mobile terminal may further include more or fewer components than Figure 1 shown in the figure, or have a different configuration from Figure 1 shown in the figure.

[0023] The memory 104 can be used to store computer programs. For example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implements the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories may be connected to the mobile terminal through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and their combinations. The transmission device 106 is used to receive or send data via a network. Specific examples of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency (Radio Frequency, abbreviated as RF) module, which is used to communicate with the Internet wirelessly.

[0024] In this embodiment, a method for monitoring the use of funds for consumer loans running on a mobile terminal, a computer terminal, or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And, although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0025] Figure 2It is a schematic flowchart of a method for monitoring the use of funds for consumer loans according to an embodiment of the present application. As Figure 2 shown, the method includes the following steps:

[0026] Step S201: Obtain the user quota utilization data of multiple users, and use the first clustering algorithm to perform clustering analysis on the user quota utilization data of the multiple users, and extract multiple typical daily utilization curves of each user in different time periods. The time periods include a time period composed of multiple working days within a preset time period and a time period composed of multiple holidays within the preset time period;

[0027] Specifically, obtain the user quota utilization data of multiple users from a relevant database. These user quota utilization data contain the consumption loan utilization records of each user at each time point (such as every day), including information such as the utilized amount and timestamp. The time range covered by the data can be the past one year, two years or a longer time period to ensure that there is sufficient historical data for analysis. Use the first clustering algorithm to perform clustering analysis on the user quota utilization data of multiple users. In this embodiment, the first clustering algorithm refers to the K-means++ clustering algorithm, and its purpose is to group the utilization behaviors of users and analyze the utilization habits of each user in different time periods. The clustering analysis is based on the characteristics of the user utilization data, such as the utilization time and the utilized amount, and assigns the data points to different clusters (or groups). Each cluster represents a pattern of utilization behavior. Through clustering, the consumption habits and patterns of each user in different time periods can be revealed, providing a basis for subsequent risk assessment.

[0028] Through cluster analysis, curves representing the average spending behavior of users within each cluster are extracted, namely the typical daily spending curves. The typical daily spending curves reflect the average spending patterns of users in different time periods (weekdays, holidays). Specifically, since the consumption patterns on holidays are quite different from those on weekdays, by performing cluster analysis on the historical quota spending data of users, clusters with similar spending habits can be identified from numerous historical spending data, and then the cluster centers of each cluster are calculated. The cluster center of each cluster is a typical daily spending curve. Users' spending habits vary on weekdays and holidays. Among them, weekdays can be further divided into the beginning of the month, the middle of the month, the end of the month, etc., and holidays can be divided into weekends, long holidays, short holidays, shopping festivals, etc. The quota spending data of multiple users in the past year can be obtained, and various scenarios of weekdays and holidays are included during this period. By performing cluster analysis on these users' quota spending data, the typical daily spending curves that can represent the spending behavior characteristics of users in different time periods are automatically identified and extracted, that is, these typical daily spending curves are automatically generated according to the clustering patterns naturally formed in the users' historical spending data. For example, the results of cluster analysis will reveal that one curve particularly conforms to the spending behavior of users on most weekdays; another curve represents the spending behavior of users on weekends or public holidays; and there is also a curve reflecting the spending behavior of users during shopping festivals. By clustering to obtain the typical daily spending curves of users, the spending behavior characteristics of users in different time periods can be captured more accurately, and thus potential abnormal spending behaviors can be better identified.

[0029] Step S202: Determine the typical daily spending curve of the user group according to the above-mentioned typical daily spending curves of all the above users;

[0030] Specifically, the purpose of step S202 is to further extract the typical daily expenditure curve of the user group from the typical daily expenditure curves of all users. In step S201, clustering analysis has been performed on the user quota expenditure data of each user, so as to extract multiple typical daily expenditure curves of each user in different time periods. These typical daily expenditure curves reflect the consumption patterns and habits of a single user based on their historical data. Further, in order to more comprehensively evaluate whether there are abnormalities in the user's expenditure behavior, it is not enough to only consider the typical daily expenditure curve of a single user. The behavior of a wider user group also needs to be considered. Therefore, it is necessary to extract the typical daily expenditure curve of the user group based on the typical daily expenditure curves of all users. By dividing users into different groups and further analyzing the typical daily expenditure curves of all users within each group, a curve that can represent the consumption pattern of the group, that is, the typical daily expenditure curve of the user group, can be obtained. The typical daily expenditure curve of each group reflects the common consumption habits of the users in that group. In short, the purpose of extracting the typical daily expenditure curve of the user group is to establish a reference benchmark, give the normal consumption pattern of the group to which a single user belongs, and provide a comparison for the subsequent judgment of the user's abnormal expenditure behavior.

[0031] Step S203: Compare the real-time daily expenditure curve of each of the above users on the day of applying for expenditure with the multiple above-mentioned typical daily expenditure curves of the above users to obtain multiple similarity comparison results, and select the minimum similarity comparison result. The above minimum similarity comparison result is the minimum value among the multiple above-mentioned similarity comparison results.

[0032] Specifically, compare the real-time daily expenditure curve of the user with each typical daily expenditure curve of the user to obtain multiple similarity comparison results, so as to quantify the similarity between the real-time daily expenditure curve of the user and each typical daily expenditure curve of the user. Select the minimum similarity comparison result from the multiple obtained similarity comparison results. The minimum similarity comparison result means that the difference between the real-time daily expenditure curve of the user and a certain typical daily expenditure curve of the user is the smallest, that is, the real-time expenditure behavior of the user is closest to the typical expenditure behavior pattern of the user. By selecting the minimum similarity comparison result, a quantitative indicator can be provided for the subsequent judgment of the user's abnormal behavior.

[0033] Step S204: Compare the trend of the real-time daily expenditure curve of each of the above users with the typical daily expenditure curve of the above user group to obtain multiple trend comparison results, and select the maximum trend comparison result. The above maximum trend comparison result is the maximum value among the multiple above-mentioned trend comparison results.

[0034] Specifically, the real-time daily disbursement curve on the day when the user applies for disbursement refers to the change in the actual disbursement amount of the user over time on the day when the user applies for a consumer loan. By comparing the trends of the real-time daily disbursement curve of each user with multiple typical daily disbursement curves of the user group determined previously for that user, multiple trend comparison results can be obtained. These trend comparison results reflect the differences in the consumption behaviors of the user on the day of applying for disbursement and the consumption behaviors of the group to which the user belongs in terms of trends. And the maximum trend comparison result is selected from the obtained multiple trend comparison results. This maximum trend comparison result is an indication of the highest correlation, that is, the trend of the user's real-time daily disbursement curve is closest to that of a certain typical daily disbursement curve of the user group, which means the user's real-time behavior is most consistent with the historical behavior of the group. By selecting the maximum trend comparison result, a quantitative indicator can be provided for the subsequent judgment of abnormal user behavior.

[0035] Step S205: Compare the above minimum similarity comparison result and the above maximum trend comparison result with their corresponding preset thresholds respectively to monitor whether there is any abnormality in the disbursement behavior of the above user.

[0036] Specifically, compare the maximum trend comparison result with the corresponding preset threshold. This preset threshold is obtained through a large amount of actual data for simulation training and is used to judge whether the trend of the user's behavior is normal. If the maximum trend comparison result is lower than this preset threshold, it indicates that there is a significant difference between the current disbursement trend of the user and the group behavior, suggesting that there is an abnormality in the user's disbursement behavior. Similar to the above process, compare the minimum similarity comparison result with the corresponding preset threshold. This preset threshold is also obtained through a large amount of actual data for simulation training. If the minimum similarity comparison result exceeds this preset threshold, it means that the current disbursement behavior of the user deviates significantly from the previous typical disbursement pattern, which also indicates that there is an abnormality in the disbursement behavior.

[0037] By comparing the minimum similarity comparison result and the maximum trend comparison result with their corresponding preset thresholds respectively, it is possible to judge whether there is any abnormality in the user's disbursement behavior. Among them, the selection of the preset threshold is very crucial. It is set based on a large amount of historical data and in-depth understanding of the user group, aiming to ensure that it can capture abnormal disbursement behaviors without misjudging normal disbursement activities. By using the minimum similarity comparison result and the maximum trend comparison result, combined with the preset threshold, it is possible to monitor the abnormality of the user's disbursement behavior in real time, identify potential risks in a timely manner, and thus improve the risk identification ability of fund use.

[0038] Through this embodiment, the real-time fund application data of the user is immediately subjected to risk assessment on the day of applying for fund disbursement, overcoming the lag of traditional post-loan fund monitoring in the prior art. At the same time, based on a large amount of historical user quota disbursement data, the disbursement behavior habit curves of the user and the user group are extracted, making the extracted control curves more referenceable and enabling more accurate prediction of whether there is an abnormality in the user's current disbursement. Thus, the problem of limited ability to identify the risk of abnormal fund use in the consumer loan disbursement application stage in the prior art is solved.

[0039] In the specific implementation process, according to the above-mentioned typical daily disbursement curves of all the above-mentioned users, the typical daily disbursement curve of the user group is determined, including: using the second clustering algorithm to perform clustering analysis on the above-mentioned typical daily disbursement curves of all the above-mentioned users to obtain multiple typical daily disbursement curve clusters; determining the cluster center curve of the multiple above-mentioned typical daily disbursement curve clusters as the above-mentioned typical daily disbursement curve of the user group. Among them, using the second clustering algorithm to perform clustering analysis on the above-mentioned typical daily disbursement curves of all the above-mentioned users to obtain multiple typical daily disbursement curve clusters includes: using the above-mentioned second clustering algorithm, for each of the above-mentioned users, taking the above-mentioned typical daily disbursement curve of each user as the initial centroid, calculating the Euclidean distance between the above-mentioned typical daily disbursement curves of all other users except this user and the above-mentioned initial centroid to obtain multiple Euclidean distances of each of the above-mentioned users; allocating the above-mentioned typical daily disbursement curve of each user to the cluster where the initial centroid corresponding to the minimum Euclidean distance among the multiple above-mentioned Euclidean distances of this user is located, and recalculating the centroid of each of the above-mentioned clusters until a preset stop condition is reached to obtain multiple above-mentioned typical daily disbursement curve clusters of each of the above-mentioned users.

[0040] The above content further details how to determine the typical daily expenditure curve of the user group. Specifically, after obtaining the above-mentioned typical daily expenditure curves of all users, the second clustering algorithm is used to cluster all the typical daily expenditure curves. The second clustering algorithm in this embodiment is the K-means++ clustering algorithm. The purpose of clustering is to gather curves with similar consumption behavior patterns together to form multiple clusters of typical daily expenditure curves. The clustering algorithm automatically groups the curves based on the similarity or difference between the curves. The curves within each group have a high degree of similarity in expenditure behavior, while the curves between different groups have significant differences in behavior patterns. The typical daily expenditure curve of each user is used as the initial centroid, that is, the typical daily expenditure curve of each user is regarded as a clustering center, which helps to ensure that the algorithm can fully consider the historical behavior patterns of each user. For example, when determining the user group to which a certain user belongs, the typical daily expenditure curve of this user is used as the initial centroid. Then, the Euclidean distance between each typical daily expenditure curve of other users and the initial centroid of this user is calculated to obtain multiple Euclidean distances for each user. The Euclidean distance is a method for measuring the spatial distance between two points and is used in this embodiment to quantify the difference between curves. The smaller the distance, the more similar the two curves are. Then, the typical daily expenditure curve of each user is assigned to the cluster where the initial centroid corresponding to the smallest Euclidean distance is located. In this way, similar curves will be gathered together to form a preliminary cluster. This process is essentially an automatic classification of the historical expenditure behavior of the user group, and each classification represents the consumption habits of the user group in a specific period.

[0041] Whenever a typical daily expenditure curve of a user is analyzed and added, it is assigned to the closest cluster according to its similarity to the curves in the existing clusters, and the centroid of the assigned cluster is recalculated. The new centroid can be the average of all the curves in the cluster. After that, the entire process (calculating the Euclidean distance, assigning curves, and updating the centroid) is repeated until a preset stop condition is reached. The preset stop condition in this embodiment can be that the centroid no longer changes, or the change in the centroid is less than a certain threshold, or a preset number of iterations is reached, indicating that the division of the cluster has stabilized and no further adjustment is required.

[0042] Through the above process, multiple clusters of typical daily usage curves are finally obtained. Each cluster contains a set of curves with similar usage behaviors. The cluster center curve reflects the typical usage behavior of the user group to which the user belongs during a certain time period. The determination of the cluster center curve enables the clear identification of the usage behavior patterns of the user group to which the user belongs in different time periods, providing a benchmark for subsequent comparison of group behavior trends. Finally, the cluster center curve of each user's cluster of typical daily usage curves is determined as the typical daily usage curve of the user group to which the user belongs. Each typical daily usage curve of the user group represents the typical consumption behavior of a type of user within a specific time period. These typical daily usage curves of the user groups will be used to compare trends with the real-time daily usage curve of a single user on the day when the user applies for usage, so as to evaluate whether the user's behavior deviates from the normal pattern of the group to which the user belongs and identify potential abnormal usage behaviors.

[0043] In some embodiments, before performing clustering analysis on the above-mentioned typical daily usage curves of all the above-mentioned users using the second clustering algorithm, the above method includes: using the elbow method to determine the number of initial cluster centers, and comparing the number of initial cluster centers with the number of the above-mentioned typical daily usage curves of all the above-mentioned users; in the case where the number of initial cluster centers is greater than or equal to the number of the above-mentioned typical daily usage curves of all the above-mentioned users, determining the number of initial cluster centers as the number of cluster centers, and in the case where the number of initial cluster centers is less than the number of the above-mentioned typical daily usage curves of all the above-mentioned users, determining the number of the above-mentioned typical daily usage curves of all the above-mentioned users as the number of cluster centers.

[0044] The above is the process of determining the number of cluster centers before performing the clustering algorithm analysis. First, the elbow method is used to determine a reasonable initial number of cluster centers. The elbow method is a method for finding the optimal k value for data clustering. By calculating the clustering cost function (such as the sum of squared errors within clusters) for different k values, an inflection point is selected as the k value. This inflection point usually appears as the point where the rate of decrease of the cost function slows down significantly, meaning that increasing more cluster centers has a diminishing marginal benefit in further reducing the error. After obtaining the initial number of cluster centers through the elbow method, it is compared with the number of typical daily usage curves of all users. The number of typical daily usage curves reflects the number of historical behavior data of all users clustered into typical curves. When the initial number of cluster centers determined by the elbow method is greater than or equal to the number of typical daily usage curves of all users, then this initial number of cluster centers is used as the number of cluster centers because a sufficient number of cluster centers can ensure that each typical daily usage curve belongs to at least one cluster, avoiding the absence of clusters. If the initial number of cluster centers is less than the number of typical daily usage curves of all users, the number of typical daily usage curves of all users is used as the number of cluster centers. This step ensures that the typical daily usage curves of each user can be taken into account, avoiding the omission of some user behavior patterns due to insufficient number of cluster centers.

[0045] Through the above process, before starting the clustering analysis, based on the elbow method and the number of typical daily usage curves of all users, a number of cluster centers that is neither too cumbersome nor too simplified can be selected to achieve the best clustering effect. This not only helps to improve the efficiency and accuracy of clustering, ensuring that the historical usage patterns of each user can be properly represented and analyzed, but also can avoid the waste of computing resources caused by too many cluster centers, or the omission of usage patterns caused by too few cluster centers.

[0046] After obtaining multiple typical daily usage curves of each user in different time periods and the typical daily usage curve of the user group through the clustering algorithm, the real-time daily usage curve of each user on the day of applying for usage is compared with multiple typical daily usage curves and multiple typical daily usage curves of the user group respectively. First, the similarity between the real-time daily usage curve of each user on the day of applying for usage and multiple typical daily usage curves of the above users is compared to obtain multiple similarity comparison results, and the minimum similarity comparison result is selected. The above minimum similarity comparison result is the minimum value among multiple above similarity comparison results, including: obtaining the above real-time daily usage curve of each user; using the Euclidean distance method to compare the similarity between the above real-time daily usage curve of each user and multiple typical daily usage curves of the above users to obtain multiple similarity comparison results; determining the minimum value among multiple above similarity comparison results as the above minimum similarity comparison result.

[0047] Specifically, first, collect the real-time daily withdrawal data of each user on the day of applying for withdrawal, and construct a real-time daily withdrawal curve. This curve reflects the actual withdrawal situation of the user on the application day, including information such as the withdrawal amount and timestamp. Then, use the Euclidean distance method to measure the similarity between the user's real-time daily withdrawal curve and multiple typical daily withdrawal curves of the user. The Euclidean distance method is a method for measuring the distance between two points (in this embodiment, it is for two sets of curve data points) in a multi-dimensional space, which can intuitively reflect the closeness of the two curves numerically. The formula is used to calculate the Euclidean distance between the user's real-time daily withdrawal curve and each typical daily withdrawal curve of the user, obtaining multiple similarity comparison results. Among them, x 1i represents the i-th data point of the user's real-time daily withdrawal curve, x 2i represents the i-th data point of the user's typical daily withdrawal curve, N represents the number of data points, and d represents the Euclidean distance between the user's real-time daily withdrawal curve and the user's typical daily withdrawal curve.

[0048] After obtaining multiple similarity comparison results, select the minimum value among the similarity comparison results, that is, the minimum similarity comparison result. This minimum similarity comparison result indicates that the Euclidean distance between the user's real-time daily withdrawal curve and a certain typical daily withdrawal curve is the smallest, that is, the user's real-time daily withdrawal behavior is closest to the user's own typical withdrawal behavior in a certain time period. The determination of the minimum similarity comparison result can help understand whether the user's withdrawal behavior on the current day is consistent with the past behavior pattern. If the minimum similarity comparison result is relatively large, it indicates that the user's current withdrawal behavior is abnormal and does not conform to the previous behavior pattern, which needs attention.

[0049] Next, compare the real-time daily withdrawal curves of each user on the day of applying for withdrawal with multiple typical daily withdrawal curves of the user group. In this embodiment, compare the trends of the real-time daily withdrawal curves of each of the above users on the day of applying for withdrawal with the typical daily withdrawal curves of the above user group to obtain multiple trend comparison results, and select the maximum trend comparison result. The above maximum trend comparison result is the maximum value among multiple above trend comparison results, including: using the Pearson correlation coefficient to compare the trends of the real-time daily withdrawal curves of each of the above users with the typical daily withdrawal curves of the above user group to obtain multiple above trend comparison results; determining the maximum value among multiple above trend comparison results as the above maximum trend comparison result.

[0050] Specifically, the Pearson correlation coefficient is a statistical index for measuring the degree of linear correlation between two variables, and its value range is between -1 and 1, where 1 represents a perfect positive correlation, -1 represents a perfect negative correlation, and 0 represents no correlation. In this embodiment, the formula Calculate the Pearson correlation coefficient between the real-time daily usage curve of each user on the application disbursement date and the typical daily usage curve of the previously determined user group. Here, X represents each data point on the real-time daily usage curve, Y represents each corresponding data point on the typical daily usage curve of the user group, N represents the number of data points, and P X,Y represents the trend comparison result between the real-time daily usage curve and the typical daily usage curve of the user group. This means that for each real-time daily usage curve, the trend comparison result with each typical daily usage curve of the user group will be calculated, and these trend comparison results reflect the matching degree between the user's current usage trend and the typical usage trend in the group's history.

[0051] Through the above calculation process, multiple trend comparison results are obtained. Find and determine the maximum value among these trend comparison results, that is, the maximum trend comparison result. This maximum trend comparison result represents the highest degree of correlation between the real-time daily usage curve of the user on the application disbursement date and a certain typical daily usage curve of the user group. The core of the above calculation process lies in using the Pearson correlation coefficient as a trend comparison tool, which can effectively measure the similarity of trends between two curves, even if there are numerical differences between the two curves. By comparing the trends of the user's real-time daily usage curve and the typical daily usage curve of the user group, it can be found whether the user's current usage behavior conforms to the behavior pattern of their group. Especially when the maximum trend comparison result is low, it indicates that the user's current usage behavior deviates significantly from the group's historical behavior, that is, there is a risk of abnormal disbursement.

[0052] To further enhance the depth and breadth of the monitoring of the use of consumer loan funds, in some embodiments, the social network data of the user can be analyzed. Specifically, after obtaining the user's authorization, collect the interaction data on their social network, including contact frequency, communication content, etc. Analyze the interactions related to the use of consumer loan funds in the social network and evaluate the degree of influence of the social network on the user's disbursement behavior. Combine the social network influence evaluation result with the user's historical and real-time disbursement data for comprehensive risk assessment. For users affected by negative social influence, issue a risk warning in advance. Social network analysis can help understand the social factors behind the user's behavior, improve the comprehensiveness of risk assessment, and combined with the influence of the user's social network, more personalized risk prevention and management strategies can be formulated to improve the security of fund use.

[0053] After obtaining the minimum similarity comparison result and the maximum trend comparison result, compare the above minimum similarity comparison result and the above maximum trend comparison result with the corresponding preset thresholds respectively to monitor whether there is any abnormality in the user's disbursement behavior, including: when the above minimum similarity comparison result is greater than or equal to the first preset threshold and the above maximum trend comparison result is less than or equal to the second preset threshold, determine that there is an abnormality in the user's disbursement behavior; when the above minimum similarity comparison result is less than the first preset threshold and the above maximum trend comparison result is greater than the second preset threshold, determine that there is no abnormality in the user's disbursement behavior; when the above minimum similarity comparison result is greater than or equal to the first preset threshold and the above maximum trend comparison result is greater than the second preset threshold, determine that there is no abnormality in the user's disbursement behavior; when the above minimum similarity comparison result is less than the first preset threshold and the above maximum trend comparison result is less than or equal to the second preset threshold, determine that there is no abnormality in the user's disbursement behavior.

[0054] The above is the specific logic for judging whether there is an abnormality in the user's disbursement behavior based on the minimum similarity comparison result and the maximum trend comparison result, combined with a preset threshold. When the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold, it is determined that the user has an abnormal disbursement behavior. This indicates that the user's current disbursement behavior has changed significantly compared with the user's historical behavior habits, there is an abnormal risk, and the user's real-time disbursement curve has a small correlation with the historical typical disbursement curve of the user group, that is, the user's current disbursement behavior does not conform to the disbursement behavior habits of the user group. In this case, the user's disbursement behavior can be determined as an abnormal application and should be monitored and warned. When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that the user's disbursement behavior is not abnormal. This indicates that the user's real-time disbursement curve is similar to the user's historical typical disbursement curve, that is, the user's current disbursement behavior is consistent with the user's historical behavior habits, and the user's real-time disbursement curve has a high correlation with the historical typical disbursement curve of the user group, that is, the user's current disbursement behavior conforms to the disbursement behavior habits of the user group. In this case, the user's disbursement behavior can be determined as a normal application and conforms to the historical disbursement habits of itself and the user group. When the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that the user's disbursement behavior is not abnormal. It shows that the user's real-time disbursement curve is quite different from the user's historical typical disbursement curve, that is, the user's current disbursement behavior has changed compared with the user's historical behavior habits, but the user's real-time disbursement curve has a high correlation with the historical typical disbursement curve of the user group, that is, the user's current disbursement behavior conforms to the disbursement behavior habits of the user group. In this case, although the user's own disbursement behavior has changed but conforms to the habit trend of the user group, it means that the change is within the expected range, so the user's disbursement behavior can also be determined as a normal application. When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold, it is determined that the user's disbursement behavior is not abnormal. It shows that the user's real-time disbursement curve is similar to the user's historical typical disbursement curve, that is, the user's current disbursement behavior is consistent with the user's historical behavior habits, but the user's real-time disbursement curve has a small correlation with the historical typical disbursement curve of the user group, that is, the user's current disbursement behavior is different from the disbursement behavior habits of the user group. In this case, although the user's disbursement behavior does not conform to the overall trend of the user group, it is consistent with the user's own historical disbursement habits, so the user's disbursement behavior can also be determined as a normal application.

[0055] Through the above logic, based on the user's real-time and historical utilization data, combined with group behavior patterns, multi-dimensional risk assessment can be carried out to timely detect and warn of potential abnormal behaviors, thereby effectively managing the risk of fund utilization of consumer loans. The advantage of this method is that it not only considers the behavioral changes of individual users, but also takes into account the behavioral consistency between users and their groups, achieving comprehensiveness and forward-looking in risk assessment at the stage of consumer loan application.

[0056] In some embodiments, deep learning models (such as long short-term memory networks or convolutional neural networks) can also be used to optimize the prediction of utilization behavior and risk assessment. Specifically, the user's utilization data is converted into a format suitable for input into the deep learning model, such as time series data. The deep learning model is trained using historical utilization data, with the goal of predicting the user's utilization behavior pattern and its potential risks. The user's real-time utilization data is input into the trained model to predict the risk level of the user's utilization behavior in real time. Based on the model prediction results, the above-mentioned preset threshold is dynamically adjusted to adapt to the changing situation and user behavior patterns. Using deep learning models can learn and capture complex patterns in utilization behavior, improve the accuracy of prediction, and the model can update the prediction strategy in real time to more flexibly adapt to the dynamic changes in the market and user behavior.

[0057] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the method for monitoring the utilization of funds for consumer loans of the present application will be described in detail below in conjunction with specific embodiments.

[0058] This embodiment relates to a specific method for monitoring the utilization of funds for consumer loans, as Figure 3 shown, including the following steps:

[0059] Step S1: Access to stock data. Since there may be missing data in the accessed data, in this embodiment, the formula is used to take the average value of the data of the user at the current time point to replace the missing data. Where x it represents the missing data of the i-th user at the t-th time point. n is the number of days of the selected data sample. xtk is the user's utilization amount on the k-th day at the t-th time point. z represents the number of time points when the user's quota is utilized.

[0060] Step S2: Data preprocessing. The user's utilization data of the quota varies greatly with the user's consumption habits. In the process of data clustering, if the data is not standardized, it is easy to amplify the influence of those attributes with a larger quantity set and ignore the data attributes with a smaller quantity set during the clustering process, resulting in inaccurate clustering results or even errors. In this embodiment, the deviation normalization algorithm is used for data normalization processing to linearly transform the data.

[0061] Let the data set be X i ={x i1 ,x i2 ....,x in}}, with a total of n attributes. Then the standardized value is '

[0062] where x ij is the standardized data, and are the maximum and minimum values in X i respectively. After standardization, all the data is within the range of [0, 1], avoiding the influence of magnifying some values with larger magnitudes, thus making the clustering results more accurate.

[0063] Step S3: Extraction and reception of the user's typical daily expenditure curve. Since the consumption behaviors and habits of the same user are different at different times, to avoid accidental situations, in this embodiment, the elbow method is first used to determine the optimal number of clusters, and the K-means++ clustering algorithm is used to analyze the user's historical consumption data to accurately extract multiple daily expenditure curves of the user in different time periods. So far, multiple exclusive typical daily expenditure curves of each user have been extracted.

[0064] Step S4: Measure the real-time daily expenditure curve of the user against the user's typical daily expenditure curves.

[0065] (1) Use the Euclidean distance to measure the similarity between the real-time daily expenditure curve of each user every day and the user's various typical daily expenditure curves, reflecting the differences in the user's expenditure data values.

[0066] (2) Record the minimum measurement result a.

[0067] Step S5: Extraction of the typical daily expenditure curve of the user group. Use the clustering algorithm to cluster all the users' typical daily expenditure curves obtained in Step S3, and determine the cluster center curve of the cluster where the user's typical daily expenditure curve is located as the typical daily expenditure curve of the user group.

[0068] The traditional K-means clustering algorithm is simple to operate and has good scalability, and can be applied to large-scale data sets, but there are problems such as the need to set the number of initial clustering centers in advance and the selection of initial centroids. To effectively extract a group of users with the same consumption behavior as the user, this embodiment improves the algorithm as follows:

[0069] 1) Use the elbow method to determine the number of clustering centers k. If the k value is greater than the number of the user's typical curves, then use k as the number of clustering centers; if the k value is less than the number of the user's typical curves, then use the number of the user's typical curves as the number of clustering centers.

[0070] 2) Use the typical curve of this user calculated in step S3 as the initial centroid.

[0071] 3) Calculate the Euclidean distances from other points in the data (points other than the cluster centers) to each cluster center.

[0072] 4) Recalculate the centroid and repeat step S3 until the centroid no longer changes.

[0073] To a certain extent, this method avoids the interference of big data on user characteristic behaviors and the trouble of falling into local optimal solutions due to randomly selecting cluster centers.

[0074] Step S6: Measure the real-time daily expenditure curve of this user against the typical daily expenditure curve of the user group to which the user belongs.

[0075] (1) Use the Pearson correlation coefficient to measure the trend between the real-time expenditure curve of the user every day and the typical daily expenditure curve of the user group to which the user belongs, reflecting the difference between the actual expenditure data curve of the user and the daily habitual expenditure trend of the user group to which the user belongs.

[0076] (2) Record the measurement result b with the highest correlation.

[0077] Step S7: Determine whether the user has abnormal behaviors based on the measurement results, so as to monitor and give early warnings.

[0078] Compare the obtained measurement results a and b with the predetermined thresholds A and B. The thresholds A and B need to be obtained through a large amount of actual data for simulation training. In this embodiment, the following several situations can be referred to for judgment:

[0079] (1) If a≥A and b≤B, it indicates that the real-time expenditure curve of the user is quite different from the historical typical expenditure curve of the user, that is, the current expenditure behavior of the user has changed significantly compared with the user's historical behavior habits, and there is an abnormal risk. And the correlation between the real-time expenditure curve of the user and the historical typical expenditure curve of the user group is relatively small, that is, the current expenditure behavior of the user does not conform to the expenditure behavior habits of the user group. In this case, the expenditure behavior of this user can be determined as an abnormal application and should be monitored and given early warnings.

[0080] (2) If aB, it indicates that the real-time expenditure curve of the user is similar to the historical typical expenditure curve of the user, that is, the current expenditure behavior of the user is consistent with the user's historical behavior habits. And the correlation between the real-time expenditure curve of the user and the historical typical expenditure curve of the user group is relatively high, that is, the current expenditure behavior of the user conforms to the expenditure behavior habits of the user group. In this case, the expenditure behavior of this user can be determined as a normal application, conforming to the historical expenditure habits of the user himself and the user group.

[0081] (3) If a ≥ A and b > B, it indicates that the user's real-time usage curve is significantly different from the user's historical typical usage curve, that is, the user's current usage behavior has changed compared with the user's historical behavior habits. However, the user's real-time usage curve has a relatively high correlation with the historical typical usage curve of the user group, that is, the user's current usage behavior conforms to the usage behavior habits of the user group. In this case, although the user's own usage behavior has changed, it conforms to the habit trend of the user group, indicating that the change is within the expected range. Therefore, the user's usage behavior can also be determined as a normal application.

[0082] (4) If a < A and b ≤ B, it indicates that the user's real-time usage curve is similar to the user's historical typical usage curve, that is, the user's current usage behavior is consistent with the user's historical behavior habits. However, the user's real-time usage curve has a relatively small correlation with the historical typical usage curve of the user group, that is, the user's current usage behavior is different from the usage behavior habits of the user group. In this case, although the user's usage behavior does not conform to the overall trend of the user group, it is consistent with the user's own historical usage habits. Therefore, the user's usage behavior can also be determined as a normal application.

[0083] The embodiment of the present application also provides a monitoring device for the use of funds for consumer loans. It should be noted that the monitoring device for the use of funds for consumer loans in the embodiment of the present application can be used to execute the method for monitoring the use of funds for consumer loans provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0084] The following introduces the monitoring device for the use of funds for consumer loans provided in the embodiment of the present application.

[0085] Figure 4 It is a structural block diagram of the monitoring device for the use of funds for consumer loans according to the embodiment of the present application. As Figure 4As shown in the figure, the device includes an acquisition unit 10, a first determination unit 20, a similarity comparison unit 30, a trend comparison unit 40, and a comparison unit 50. Among them, the acquisition unit is used to acquire the user quota utilization data of multiple users, and perform clustering analysis on the user quota utilization data of multiple users using the first clustering algorithm, and extract multiple typical daily utilization curves of each user in different time periods. The time periods include a time period composed of multiple working days within a preset time period, and a time period composed of multiple holidays within the preset time period. The determination unit is used to determine the typical daily utilization curve of the user group according to the typical daily utilization curves of all the above users. The similarity comparison unit is used to compare the real-time daily utilization curve of each user on the day of applying for utilization with the multiple typical daily utilization curves of the user to obtain multiple similarity comparison results, and select the minimum similarity comparison result. The minimum similarity comparison result is the minimum value among the multiple similarity comparison results. The trend comparison unit is used to compare the trend of the real-time daily utilization curve of each user with the typical daily utilization curve of the user group to obtain multiple trend comparison results, and select the maximum trend comparison result. The maximum trend comparison result is the maximum value among the multiple trend comparison results. The comparison unit is used to compare the minimum similarity comparison result and the maximum trend comparison result with the corresponding preset thresholds respectively to monitor whether there is an abnormality in the utilization behavior of the user.

[0086] Specifically, the acquisition unit is used to acquire the user quota utilization data of multiple users from the relevant database, and perform clustering analysis on the user quota utilization data of multiple users using the first clustering algorithm. In this embodiment, the first clustering algorithm refers to the K-means++ clustering algorithm, and its purpose is to group the utilization behaviors of users and analyze the utilization habits of each user in different time periods. The clustering analysis is based on the characteristics of the user utilization data, such as the utilization time, utilization amount, etc., and assigns the data points to different clusters (or groups). Each cluster represents a pattern of utilization behavior. Through clustering, the consumption habits and patterns of each user in different time periods can be revealed, providing a basis for subsequent risk assessment.

[0087] Through cluster analysis, curves representing the average spending behavior of users within each cluster are extracted, namely, typical daily spending curves. The typical daily spending curves reflect the average spending patterns of users in different time periods (weekdays, holidays). Specifically, since the consumption patterns on holidays are quite different from those on weekdays, by performing cluster analysis on the historical quota spending data of users, clusters with similar spending habits can be identified from numerous historical spending data, and then the cluster center of each cluster is calculated. The cluster center of each cluster is a typical daily spending curve. Users' spending habits vary on weekdays and holidays. Among them, weekdays can be further divided into the beginning of the month, the middle of the month, the end of the month, etc., and holidays can be divided into weekends, long holidays, short holidays, shopping festivals, etc. The quota spending data of multiple users in the past year can be obtained, which includes various scenarios of weekdays and holidays during this period. By performing cluster analysis on these users' quota spending data, typical daily spending curves that can represent the spending behavior characteristics of users in different time periods are automatically identified and extracted, that is, these typical daily spending curves are automatically generated according to the clustering patterns naturally formed in the users' historical spending data. For example, the cluster analysis results will reveal that one curve particularly conforms to the spending behavior of users on most weekdays; another curve represents the spending behavior of users on weekends or public holidays; and there is also a curve reflecting the spending behavior of users during shopping festivals. By clustering to obtain the typical daily spending curves of users, the spending behavior characteristics of users in different time periods can be captured more accurately, and then potential abnormal spending behaviors can be better identified.

[0088] The determination unit is used to further extract the typical daily spending curve of the user group from the typical daily spending curves of all users. In the acquisition unit, cluster analysis has been performed on the quota spending data of each user, so as to extract multiple typical daily spending curves of each user in different time periods. These typical daily spending curves reflect the consumption patterns and habits of a single user based on their historical data. Further, in order to more comprehensively evaluate whether there are abnormalities in the spending behavior of users, it is not enough to only consider the typical daily spending curves of a single user. The behavior of a wider user group also needs to be considered. Therefore, according to the typical daily spending curves of all users, the typical daily spending curve of the user group needs to be extracted. By dividing users into different groups and further analyzing the typical daily spending curves of all users within each group, a curve that can represent the consumption pattern of the group is obtained, that is, the typical daily spending curve of the user group. The typical daily spending curve of each group reflects the common consumption habits of the users in that group. In short, the purpose of extracting the typical daily spending curve of the user group is to establish a reference benchmark, give the normal consumption pattern of the group to which a single user belongs, and provide a comparison for the subsequent judgment of abnormal spending behavior of users.

[0089] The similarity comparison unit is used to compare the real-time daily expenditure curve of a user with each typical daily expenditure curve of the user, obtaining multiple similarity comparison results to quantify the similarity degree between the real-time daily expenditure curve of the user and each typical daily expenditure curve of the user. The minimum similarity comparison result is selected from the obtained multiple similarity comparison results. The minimum similarity comparison result means that the difference between the real-time daily expenditure curve of the user and a certain typical daily expenditure curve of the user is the smallest, that is, the real-time expenditure behavior of the user is closest to the typical expenditure behavior pattern of the user. By selecting the minimum similarity comparison result, a quantitative index can be provided for the subsequent judgment of user behavior anomalies.

[0090] The trend comparison unit is used to compare the trends of the real-time daily expenditure curve of each user with multiple typical daily expenditure curves of the user group determined previously, obtaining multiple trend comparison results. These trend comparison results reflect the differences in trends between the consumption behavior of the user on the day of applying for expenditure and the consumption behavior of the group to which the user belongs. The maximum trend comparison result is selected from the obtained multiple trend comparison results. The maximum trend comparison result is an indication of the highest correlation, that is, the trend of the real-time daily expenditure curve of the user is closest to the trend of a certain typical daily expenditure curve of the user group, that is, the situation where the user's real-time behavior is most consistent with the historical behavior of the group to which the user belongs. By selecting the maximum trend comparison result, a quantitative index can be provided for the subsequent judgment of user behavior anomalies.

[0091] The comparison unit is used to compare the maximum trend comparison result with the corresponding preset threshold. The preset threshold is obtained through a large amount of actual data for simulation training and is used to judge whether the trend of the user's behavior is normal. If the maximum trend comparison result is lower than the preset threshold, it indicates that there is a significant difference between the current expenditure trend of the user and the group behavior, indicating that there is an anomaly in the user's expenditure behavior; similarly, the minimum similarity comparison result is compared with the corresponding preset threshold. The preset threshold is also obtained through a large amount of actual data for simulation training. If the minimum similarity comparison result exceeds the preset threshold, it indicates that the current expenditure behavior of the user deviates greatly from the previous typical expenditure pattern, which also indicates that there is an anomaly in the expenditure behavior.

[0092] By comparing the minimum similarity comparison result and the maximum trend comparison result with the corresponding preset thresholds respectively, it is possible to judge whether there is an anomaly in the user's expenditure behavior. Among them, the selection of the preset threshold is very crucial. It is set based on a large amount of historical data and in-depth understanding of the user group, aiming to ensure that both abnormal expenditure behaviors can be captured and normal expenditure activities will not be misjudged. Using the minimum similarity comparison result and the maximum trend comparison result, combined with the preset threshold, it is possible to monitor the anomalies in the user's expenditure behavior in real time, identify potential risks in a timely manner, and thus improve the risk identification ability of fund use.

[0093] Through this embodiment, the real-time fund-using application data of the user is immediately subjected to risk assessment on the day of applying for fund use, overcoming the lag of traditional post-loan fund monitoring in the prior art. At the same time, based on a large amount of historical user quota fund-using data, the fund-using behavior habit curves of the user and the user group are extracted, making the extracted reference curves more referenceable and enabling more accurate prediction of whether there is an abnormality in the user's current fund use, thereby solving the problem of limited ability to identify the risk of abnormal fund use in the consumer loan fund-using application stage in the prior art.

[0094] In the specific implementation process, the above-mentioned determination unit includes a clustering analysis module and a first determination module. Among them, the clustering analysis module is used to perform clustering analysis on the above-mentioned typical daily fund-using curves of all the above-mentioned users by using a second clustering algorithm to obtain multiple clusters of typical daily fund-using curves; the first determination module is used to determine the cluster center curve of the multiple above-mentioned clusters of typical daily fund-using curves as the typical daily fund-using curve of the above-mentioned user group. Among them, the clustering analysis module includes a calculation sub-module and an assignment sub-module. The calculation sub-module is used to take the above-mentioned typical daily fund-using curve of each above-mentioned user as the initial centroid for each above-mentioned user, calculate the Euclidean distance between the above-mentioned typical daily fund-using curves of all other users except this user and the above-mentioned initial centroid to obtain multiple Euclidean distances of each above-mentioned user; the assignment sub-module is used to assign the above-mentioned typical daily fund-using curve of each above-mentioned user to the cluster where the initial centroid corresponding to the minimum Euclidean distance among the multiple above-mentioned Euclidean distances of this user is located, and recalculate the centroid of each above-mentioned cluster until a preset stop condition is reached to obtain multiple above-mentioned clusters of typical daily fund-using curves of each above-mentioned user.

[0095] Specifically, the clustering analysis module is used to cluster all the above-mentioned typical daily expenditure curves of all users using the second clustering algorithm after obtaining them. The second clustering algorithm in this embodiment is the K-means++ clustering algorithm. The purpose of clustering is to gather curves with similar consumption behavior patterns together to form multiple clusters of typical daily expenditure curves. The clustering algorithm automatically groups the curves based on the similarity or difference between the curves. The curves within each group have a high degree of similarity in expenditure behavior, while the curves between different groups have significant differences in behavior patterns. The calculation sub-module is used to use the typical daily expenditure curve of each user as the initial centroid, that is, the typical daily expenditure curve of each user will be regarded as a clustering center, which helps to ensure that the algorithm can fully consider the historical behavior patterns of each user. For example, when determining the user group where a certain user belongs, the typical daily expenditure curve of this user is used as the initial centroid. Then, calculate the Euclidean distance between each typical daily expenditure curve of other users and the initial centroid of this user to obtain multiple Euclidean distances of each user. The Euclidean distance is a method to measure the spatial distance between two points and is used in this embodiment to quantify the difference between curves. The smaller the distance, the more similar the two curves are. The assignment sub-module is used to assign the typical daily expenditure curve of each user to the cluster where the initial centroid with the smallest Euclidean distance corresponding to it is located. In this way, similar curves will be gathered together to form a preliminary cluster. This process is essentially an automatic classification of the historical expenditure behaviors of the user group, and each classification represents the consumption habits of the user group in a certain specific period.

[0096] Whenever a typical daily expenditure curve of a user is analyzed and added, it will be assigned to the closest cluster according to its similarity to the curves in the existing clusters, and the centroid of the assigned cluster will be recalculated. The new centroid can be the average value of all the curves in the cluster. After that, the whole process (calculating the Euclidean distance, assigning curves, and updating the centroid) is repeated until a preset stop condition is reached. The preset stop condition in this embodiment can be that the centroid no longer changes, or the change in the centroid is less than a certain threshold, or a preset number of iterations is reached, indicating that the division of the cluster has stabilized and no further adjustment is required.

[0097] Through the above process, multiple typical daily expenditure curve clusters are finally obtained. Each cluster contains a set of curves with similar expenditure behaviors. The cluster center curve reflects the typical expenditure behavior of the user group where the user belongs in a certain time period. The determination of the cluster center curve enables the clear identification of the expenditure behavior patterns of the user group where the user belongs in different time periods, providing a benchmark for subsequent comparison of group behavior trends. The first determination module is used to determine the cluster center curve of each user's typical daily expenditure curve cluster as the typical daily expenditure curve of the user group to which the user belongs. Each typical daily expenditure curve of the user group represents the typical consumption behavior of a type of user within a specific time period. These typical daily expenditure curves of the user groups will be used to compare trends with the real-time daily expenditure curve of a single user on the day of applying for expenditure, so as to evaluate whether the user's behavior deviates from the normal mode of their group and identify potential abnormal expenditure behaviors.

[0098] In some embodiments, the above device includes a second determination unit and a third determination unit. Among them, the second determination unit is used to determine the number of initial cluster centers using the elbow method before performing cluster analysis on the above typical daily expenditure curves of all the above users, and compare the number of the above initial cluster centers with the number of the above typical daily expenditure curves of all the above users; the third determination unit is used to determine the number of cluster centers as the number of initial cluster centers when the number of the above initial cluster centers is greater than or equal to the number of the above typical daily expenditure curves of all the above users, and determine the number of the above typical daily expenditure curves of all the above users as the number of the above cluster centers when the number of the above initial cluster centers is less than the number of the above typical daily expenditure curves of all the above users.

[0099] Specifically, the second determination unit is used to determine a reasonable number of initial cluster centers using the elbow method before performing analysis with the clustering algorithm. The third determination unit is used to compare the number of initial cluster centers obtained through the elbow method with the number of typical daily expenditure curves of all users after obtaining the number of initial cluster centers through the elbow method. The number of typical daily expenditure curves reflects the number of historical behavior data of all users clustered into typical curves. When the number of initial cluster centers determined by the elbow method is greater than or equal to the number of typical daily expenditure curves of all users, the number of initial cluster centers is used as the number of cluster centers because a sufficient number of cluster centers can ensure that each typical daily expenditure curve belongs to at least one cluster, avoiding the absence of clusters. If the number of initial cluster centers is less than the number of typical daily expenditure curves of all users, the number of typical daily expenditure curves of all users is used as the number of cluster centers. This step ensures that the typical daily expenditure curves of each user can be taken into account, avoiding the omission of some user behavior patterns due to insufficient number of cluster centers.

[0100] Through the above process, before starting the clustering analysis, it is possible to select the number of clustering centers that is neither too cumbersome nor too simplified based on the elbow method and the number of typical daily expenditure curves of all users, so as to achieve the best clustering effect. This not only helps to improve the efficiency and accuracy of clustering, ensures that the historical expenditure patterns of each user can be properly represented and analyzed, but also avoids the waste of computing resources caused by too many clustering centers or the omission of expenditure patterns caused by too few clustering centers.

[0101] In some embodiments, the above similarity comparison unit includes an acquisition module, a similarity comparison module, and a second determination module. Among them, the acquisition module is used to acquire the above real-time daily expenditure curves of each of the above users; the similarity comparison module is used to compare the similarity between the above real-time daily expenditure curves of each of the above users and the above multiple typical daily expenditure curves of the above users by using the Euclidean distance method, and obtain multiple similarity comparison results; the second determination module is used to determine the minimum value among the multiple above similarity comparison results as the above minimum similarity comparison result.

[0102] Specifically, the acquisition module is used to collect the real-time daily expenditure data of each user on the day of applying for expenditure, and construct a real-time daily expenditure curve, which reflects the actual expenditure situation of the user on the application day, including information such as the amount of expenditure and the timestamp. The similarity comparison module is used to use the Euclidean distance method to measure the similarity between the real-time daily expenditure curve of the user and the multiple typical daily expenditure curves of the user. The Euclidean distance method is a method for measuring the distance between two points (in this embodiment, it is for two sets of curve data points) in a multi-dimensional space, which can intuitively reflect the closeness of two curves numerically. Calculate the Euclidean distance between the real-time daily expenditure curve of the user and each of the typical daily expenditure curves of the user to obtain multiple similarity comparison results. The second determination module is used to select the minimum value among the multiple similarity comparison results after obtaining the multiple similarity comparison results, that is, the minimum similarity comparison result. This minimum similarity comparison result indicates that the Euclidean distance between the real-time daily expenditure curve of the user and a certain typical daily expenditure curve is the smallest, that is, the real-time expenditure behavior of the user is closest to the typical expenditure behavior of the user in a certain time period. The determination of the minimum similarity comparison result can help to understand whether the user's expenditure behavior on that day is consistent with the past behavior pattern. If the minimum similarity comparison result is relatively large, it means that the user's current expenditure behavior is abnormal and does not conform to the previous behavior pattern, and attention needs to be paid.

[0103] The above-mentioned trend comparison unit includes a trend comparison module and a third determination module. Among them, the trend comparison module is used to compare the real-time daily expenditure curves of each of the above-mentioned users with the typical daily expenditure curve of the above-mentioned user group using the Pearson correlation coefficient, and obtain multiple above-mentioned trend comparison results; the third determination module is used to determine the maximum value among the multiple above-mentioned trend comparison results as the above-mentioned maximum trend comparison result.

[0104] Specifically, the comparison module is used to calculate the Pearson correlation coefficient between the real-time daily expenditure curve on the day of application for expenditure of each user and the typical daily expenditure curve of the user group determined previously. This means that for each real-time daily expenditure curve, a trend comparison result with each typical daily expenditure curve of the user group will be calculated, and these trend comparison results reflect the matching degree between the current expenditure trend of the user and the typical expenditure trend in the history of the group.

[0105] The third determination module is used to determine the maximum value among these trend comparison results, that is, the maximum trend comparison result. This maximum trend comparison result represents the highest degree of correlation between the real-time daily expenditure curve on the day of the user's application for expenditure and a certain typical daily expenditure curve of the user group. The core of the above calculation process lies in using the Pearson correlation coefficient as a trend comparison tool, which can effectively measure the similarity in trends between two curves, even if there are differences in the values of the two curves. By comparing the trends of the user's real-time daily expenditure curve and the typical daily expenditure curve of the user group, it is possible to find out whether the user's current expenditure behavior conforms to the behavior pattern of their group. Especially when the maximum trend comparison result is low, it indicates that the user's current expenditure behavior deviates significantly from the historical behavior of the group, that is, there is a risk of abnormal expenditure.

[0106] The above-mentioned comparison unit includes a fourth determination module, a fifth determination module, a sixth determination module, and a seventh determination module. Among them, the fourth determination module is used to determine that the expenditure behavior of the above-mentioned user is abnormal when the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold after obtaining the minimum similarity comparison result and the maximum trend comparison result; the fifth determination module is used to determine that the expenditure behavior of the above-mentioned user is not abnormal when the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is greater than the second preset threshold; the sixth determination module is used to determine that the expenditure behavior of the above-mentioned user is not abnormal when the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is greater than the second preset threshold; the seventh determination module is used to determine that the expenditure behavior of the above-mentioned user is not abnormal when the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold.

[0107] Specifically, when the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold, it is determined that the user has an abnormal spending behavior. This indicates that the user's current spending behavior has changed significantly compared to the user's historical behavior habits, there is an abnormal risk, and the correlation between the user's real-time spending curve and the historical typical spending curve of the user group is small, that is, the user's current spending behavior does not conform to the spending behavior habits of the user group. In this case, the user's spending behavior can be determined as an abnormal application and should be monitored and warned. When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that the user's spending behavior is not abnormal. This indicates that the user's real-time spending curve is similar to the user's historical typical spending curve, that is, the user's current spending behavior is consistent with the user's historical behavior habits, and the correlation between the user's real-time spending curve and the historical typical spending curve of the user group is high, that is, the user's current spending behavior conforms to the spending behavior habits of the user group. In this case, the user's spending behavior can be determined as a normal application and conforms to the historical spending habits of the user and the user group. When the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that the user's spending behavior is not abnormal. It shows that the user's real-time spending curve is quite different from the user's historical typical spending curve, that is, the user's current spending behavior has changed compared to the user's historical behavior habits, but the correlation between the user's real-time spending curve and the historical typical spending curve of the user group is high, that is, the user's current spending behavior conforms to the spending behavior habits of the user group. In this case, although the user's own spending behavior has changed but conforms to the habit trend of the user group, it shows that the change is within the expected range, so the user's spending behavior can also be determined as a normal application. When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold, it is determined that the user's spending behavior is not abnormal. It shows that the user's real-time spending curve is similar to the user's historical typical spending curve, that is, the user's current spending behavior is consistent with the user's historical behavior habits, but the correlation between the user's real-time spending curve and the historical typical spending curve of the user group is small, that is, the user's current spending behavior is different from the spending behavior habits of the user group. In this case, although the user's spending behavior does not conform to the overall trend of the user group, it is consistent with the user's own historical spending habits, so the user's spending behavior can also be determined as a normal application.

[0108] Through the above logic, it is possible to conduct multi-dimensional risk assessments based on the user's real-time and historical fund usage data, combined with group behavior patterns, timely detect and alert potential abnormal behaviors, thereby effectively managing the risk of fund usage for consumer loans. The advantage of this method is that it not only considers the behavioral changes of individual users but also the behavioral consistency between users and their groups, achieving comprehensiveness and forward-looking in risk assessment during the consumer loan application stage.

[0109] The above-mentioned monitoring device for the fund usage of consumer loans includes a processor and a memory. The above-mentioned acquisition unit, first determination unit, similarity comparison unit, trend comparison unit, comparison unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions. The above-mentioned modules are all located in the same processor; alternatively, the above-mentioned each module is located in different processors in any combined form.

[0110] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0111] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program. When the above-mentioned program runs, it controls the device where the computer-readable storage medium is located to execute the above-mentioned method for monitoring the fund usage of consumer loans.

[0112] An embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps of the above-mentioned method for monitoring the fund usage of consumer loans.

[0113] This application also provides a computer program product, which is suitable for executing a program initialized with the steps of the above-mentioned method for monitoring the fund usage of consumer loans when executed on a data processing device.

[0114] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. They can be implemented by program codes executable by the computing device, so that they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described herein can be executed in a different order, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0115] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0116] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0117] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0119] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0120] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash RAM. The memory is an example of computer-readable media.

[0121] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes or other magnetic storage devices, or any other non-transitory media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0122] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0123] The above are only the preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A method for monitoring the use of funds for consumer loans, characterized in that: include: Acquire user credit limit expenditure data of multiple users, and use a first clustering algorithm to perform cluster analysis on the multiple user credit limit expenditure data, and extract multiple typical daily expenditure curves of each user in different time periods, wherein the time period includes a time period consisting of multiple working days within a preset time period and a time period consisting of multiple holidays within the preset time period; Determining a typical daily expenditure curve for a user group based on the typical daily expenditure curves of all the users; Compare the real-time daily expenditure curve of each user on the day of application for expenditure with the multiple typical daily expenditure curves of the user to obtain multiple similarity comparison results, and select the minimum similarity comparison result, which is the minimum value among the multiple similarity comparison results; Performing a trend comparison between the real-time daily expenditure curve of each user and the typical daily expenditure curve of the user group to obtain multiple trend comparison results, and selecting a maximum trend comparison result, wherein the maximum trend comparison result is the maximum value among the multiple trend comparison results; The minimum similarity comparison result and the maximum trend comparison result are respectively compared with corresponding preset thresholds to monitor whether there is any abnormality in the spending behavior of the user.

2. The method according to claim 1, characterized in that Determining a typical daily expenditure curve of a user group according to the typical daily expenditure curves of all the users includes: Using a second clustering algorithm to perform cluster analysis on the typical daily expenditure curves of all the users to obtain a plurality of typical daily expenditure curve clusters; A cluster center curve of a plurality of the typical daily expenditure curve clusters is determined as the typical daily expenditure curve of the user group.

3. The method according to claim 2, characterized in that A second clustering algorithm is used to perform cluster analysis on the typical daily expenditure curves of all the users to obtain a plurality of typical daily expenditure curve clusters, including: Using the second clustering algorithm, for each of the users, taking the typical daily expenditure curve of each of the users as an initial centroid, calculating the Euclidean distance between the typical daily expenditure curves of all other users except the user and the initial centroid, and obtaining a plurality of Euclidean distances of each of the users; The typical daily expenditure curve of each user is assigned to the cluster where the initial centroid corresponding to the minimum Euclidean distance among the multiple Euclidean distances of the user is located, and the centroid of each cluster is recalculated until a preset stopping condition is reached, thereby obtaining multiple clusters of the typical daily expenditure curves of each user.

4. The method according to claim 2, characterized in that: Before using a second clustering algorithm to perform cluster analysis on the typical daily expenditure curves of all the users, the method includes: Determine the number of initial cluster centers using the elbow rule, and compare the number of initial cluster centers with the number of typical daily expenditure curves of all the users; When the number of the initial cluster centers is greater than or equal to the number of the typical daily expenditure curves of all the users, the number of the initial cluster centers is determined as the number of cluster centers; when the number of the initial cluster centers is less than the number of the typical daily expenditure curves of all the users, the number of the typical daily expenditure curves of all the users is determined as the number of cluster centers.

5. The method according to claim 1, characterized in that: The real-time daily expenditure curve of each user on the day of application for expenditure is compared with the multiple typical daily expenditure curves of the user to obtain multiple similarity comparison results, and the minimum similarity comparison result is selected, and the minimum similarity comparison result is the minimum value among the multiple similarity comparison results, including: Obtaining the real-time daily expenditure curve of each user; Using the Euclidean distance method, the real-time daily expenditure curve of each user is compared with the multiple typical daily expenditure curves of the user for similarity, to obtain multiple similarity comparison results; The minimum value among the plurality of similarity comparison results is determined as the minimum similarity comparison result.

6. The method according to claim 1, characterized in that Performing a trend comparison between the real-time daily expenditure curve of each user and the typical daily expenditure curve of the user group, obtaining multiple trend comparison results, and selecting a maximum trend comparison result, the maximum trend comparison result being the maximum value among the multiple trend comparison results, including: Using the Pearson correlation coefficient, the real-time daily expenditure curve of each user is compared with the typical daily expenditure curve of the user group to obtain a plurality of trend comparison results; The maximum value among the plurality of trend comparison results is determined as the maximum trend comparison result.

7. The method according to claim 1, characterized in that Comparing the minimum similarity comparison result and the maximum trend comparison result with corresponding preset thresholds respectively, and monitoring whether the spending behavior of the user is abnormal, including: When the minimum similarity comparison result is greater than or equal to a first preset threshold and the maximum trend comparison result is less than or equal to a second preset threshold, it is determined that the user's spending behavior is abnormal; When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that there is no abnormality in the spending behavior of the user; When the minimum similarity comparison result is greater than or equal to the first preset threshold and the maximum trend comparison result is greater than the second preset threshold, it is determined that there is no abnormality in the user's spending behavior; When the minimum similarity comparison result is less than the first preset threshold and the maximum trend comparison result is less than or equal to the second preset threshold, it is determined that there is no abnormality in the user's spending behavior.

8. A device for monitoring the use of funds for consumer loans, characterized in that: include: an acquisition unit, configured to acquire user credit limit expenditure data of a plurality of users, and perform cluster analysis on the plurality of user credit limit expenditure data using a first clustering algorithm, and extract a plurality of typical daily expenditure curves of each of the users in different time periods, wherein the time period includes a time period consisting of a plurality of working days within a preset time period and a time period consisting of a plurality of holidays within the preset time period; A first determining unit, configured to determine a typical daily expenditure curve of a user group according to the typical daily expenditure curves of all the users; A similarity comparison unit is used to compare the real-time daily expenditure curve of each user on the day of application for expenditure with the multiple typical daily expenditure curves of the user to obtain multiple similarity comparison results, and select the minimum similarity comparison result, which is the minimum value among the multiple similarity comparison results; A trend comparison unit, used for performing trend comparison between the real-time daily expenditure curve of each user and the typical daily expenditure curve of the user group, obtaining multiple trend comparison results, and selecting a maximum trend comparison result, wherein the maximum trend comparison result is the maximum value among the multiple trend comparison results; The comparison unit is used to compare the minimum similarity comparison result and the maximum trend comparison result with corresponding preset thresholds respectively, so as to monitor whether there is any abnormality in the spending behavior of the user.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for monitoring the use of funds for consumer loans as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for monitoring the use of funds for executing a consumer loan as described in any one of claims 1 to 7.

Images