Fund payment auditing method and device for fund account, equipment and storage medium

Abstract

The invention relates to the technical field of logistics fee settlement, in particular to a fund account payment auditing method and device, equipment and a storage medium. The fund account payment auditing method comprises the following steps of obtaining a bill and verification detail data, and judging whether the bill needs to be rejected or not according to the verification detail data so as to obtain a judgment result; if the judgment result is that rejection is not needed, outputting the judgment result that rejection is not needed to obtain a payment instruction; acquiring account balance data according to the payment instruction, and verifying the account balance data according to the verification detail data to obtain a verification result; and if the verification is passed, executing a payment instruction according to a verification passing result to obtain payment success information, and generating applicable invoice information according to the payment success information. According to the fund account payment auditing method, the commission settlement efficiency of the courier can be effectively improved, and the occurrence of artificial auditing errors can be effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics cost settlement, and in particular to a method, device, equipment and storage medium for reviewing payment of a fund account. Background Art

[0002] At present, the two main modes of courier commission payment are direct delivery and outlet payment. Direct delivery adopts a flexible employment model to pay wages. In order to solve the problems of online commission payment and tax invoicing at outlets, logistics and express delivery companies have developed an outlet flexible work system. Outlets can use this system to make payments, import bills, pay amounts, and apply for invoices, solving the problem of amount payment at outlets. After the merchant successfully creates an order, he can initiate payment for orders that have passed the verification in the "Payment Review" module, or review and reject and cancel the payment of this batch.

[0003] Although the system has improved the efficiency of commission payment to a certain extent, it still relies on manual processing in multiple processes. For example, manual review is required after bills are imported, and manual investigation is required for bills rejected due to non-compliance, which results in a long commission payment cycle. It can be seen that the existing technology still needs to be improved and enhanced. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a method, device, equipment and storage medium for reviewing payment of funds to a fund account, aiming to solve the technical problems in the prior art of long courier commission payment cycle and low bill review processing efficiency.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] The first aspect of the present invention provides a method for reviewing payment of a fund account, comprising the following steps: obtaining an invoice and verification detail data, judging whether the invoice needs to be rejected based on the verification detail data to obtain a judgment result; if the judgment result is that it does not need to be rejected, outputting a judgment result that it does not need to be rejected to obtain a payment instruction; obtaining account balance data based on the payment instruction, verifying the account balance data based on the verification detail data to obtain a verification result; if the verification passes, executing the payment instruction based on the verification pass result to obtain payment success information, and generating invoice information that can be applied for based on the payment success information.

[0007] Optionally, in a first implementation method of the first aspect of the present invention, the acquiring of the bill and verification detail data, and judging whether the bill needs to be rejected based on the verification detail data to obtain a judgment result, specifically includes: acquiring the bill and verification detail data, and using an anomaly detection algorithm to judge whether the bill needs to be rejected based on the verification detail data to obtain a judgment result; if the judgment result is rejection, outputting the rejected bill to obtain a data correction bill; and judging the data correction bill based on the verification detail data to obtain the judgment result again.

[0008] Optionally, in a second implementation method of the first aspect of the present invention, the acquiring of bill and verification detail data, using an anomaly detection algorithm, and judging whether the bill needs to be rejected based on the verification detail data to obtain a judgment result specifically includes: acquiring historical bill data, using an isolation forest algorithm as an initial model, and using historical bill data to train the initial model to obtain an anomaly detection model; acquiring abnormal bill data, and optimizing the anomaly detection model based on the abnormal bill data; acquiring bill and verification detail data, using the optimized anomaly detection model, and judging whether the bill needs to be rejected based on the verification detail data to obtain a judgment result.

[0009] Optionally, in a third implementation method of the first aspect of the present invention, the historical billing data is obtained, the isolation forest algorithm is used as the initial model, and the initial model is trained using the historical billing data to obtain an anomaly detection model, specifically including: obtaining historical billing data, the historical billing data including the bill amount, transaction date, user ID, transaction type and geographic location; using the isolation forest algorithm as the initial model, and training the initial model with the bill amount, transaction date, user ID, transaction type and geographic location as training features to obtain training data; determining a threshold based on the anomaly score distribution of the training data, and adjusting the initial model based on the threshold to obtain an anomaly detection model.

[0010] Optionally, in a fourth implementation method of the first aspect of the present invention, if the judgment result is rejection, the rejected bill is output to obtain a data correction bill, specifically including: obtaining historical abnormal bills and their verification details, and formulating abnormal value reference rules based on the historical abnormal bills and their verification details; if the judgment result is rejection, searching for possible abnormal values ​​in the rejected bill according to the abnormal value reference rules; marking possible abnormal values ​​in the rejected bill, and outputting the rejected bill to obtain a data correction bill.

[0011] Optionally, in a fifth implementation of the first aspect of the present invention, the obtaining of historical abnormal bills and their verification details, and formulating abnormal value reference rules based on the historical abnormal bills and their verification details, specifically include: obtaining abnormal bills and their verification details, classifying the abnormal bills to obtain multiple types of abnormal bills; for different types of abnormal bills, respectively extracting key information data from the abnormal bills and their verification details; using a correlation analysis algorithm to calculate the correlation between the key information data and the bill anomaly, obtaining correlation data, and formulating the abnormal value reference rules based on the correlation data.

[0012] Optionally, in a sixth implementation method of the first aspect of the present invention, the account balance data is obtained according to the payment instruction, and the account balance data is verified according to the verification detail data to obtain a verification result, specifically including: obtaining the account balance data according to the payment instruction, extracting the total amount of bill charges in the verification detail data, and determining whether the account balance is greater than the total amount of bill charges; if the account balance ≥ the total amount of bill charges, obtaining a verification pass result; if the account balance < the total amount of bill charges, obtaining a verification fail result, and outputting the account balance recharge information according to the verification fail result.

[0013] The second aspect of the present invention provides a payment review device for a capital account, comprising: a judgment module, used to obtain an invoice and verification detail data, and judge whether the invoice needs to be rejected based on the verification detail data to obtain a judgment result; an output module, used to output a judgment result that does not require rejection if the judgment result is that rejection is not required, so as to obtain a payment instruction; a verification module, used to obtain account balance data according to the payment instruction, and verify the account balance data according to the verification detail data to obtain a verification result; a generation module, used to execute the payment instruction according to the verification result if the verification is passed, so as to obtain payment success information, and generate invoice information that can be applied for according to the payment success information.

[0014] Optionally, in a first implementation method of the second aspect of the present invention, the judgment module includes: a calculation submodule, used to obtain the bill and verification detail data, and adopt an anomaly detection algorithm to judge whether the bill needs to be rejected according to the verification detail data to obtain a judgment result; an output submodule, used to output the rejected bill if the judgment result is rejection, so as to obtain a data correction bill; a judgment submodule, used to judge the data correction bill according to the verification detail data to obtain the judgment result again.

[0015] Optionally, in a second implementation method of the second aspect of the present invention, the computing submodule includes: a construction unit, used to obtain historical billing data, using an isolation forest algorithm as an initial model, and using historical billing data to train the initial model to obtain an anomaly detection model; an optimization unit, used to obtain abnormal billing data, and optimize the anomaly detection model based on the abnormal billing data; a judgment unit, used to obtain bill and verification detail data, using the optimized anomaly detection model, and judging whether the bill needs to be rejected based on the verification detail data to obtain a judgment result.

[0016] Optionally, in a third implementation of the second aspect of the present invention, the construction unit includes: an acquisition subunit, used to acquire historical bill data, the historical bill data including bill amount, transaction date, user ID, transaction type and geographic location; a training subunit, used to use the isolation forest algorithm as the initial model, and the bill amount, transaction date, user ID, transaction type and geographic location as training features to train the initial model to obtain training data; an adjustment subunit, used to determine a threshold based on the anomaly score distribution of the training data, and adjust the initial model based on the threshold to obtain an anomaly detection model.

[0017] Optionally, in a fourth implementation of the second aspect of the present invention, the output submodule includes: a formulation unit, used to obtain historical abnormal bills and their verification details, and formulate abnormal value reference rules based on the historical abnormal bills and their verification details; a search unit, used to search for possible abnormal values ​​in the rejected bills according to the abnormal value reference rules if the judgment result is rejection; a marking unit, used to mark possible abnormal values ​​in the rejected bills, and output the rejected bills to obtain data correction bills.

[0018] Optionally, in a fifth implementation of the second aspect of the present invention, the formulation unit includes: a classification subunit, used to obtain abnormal bills and their verification detail data, and classify the abnormal bills to obtain multiple types of abnormal bills; an extraction subunit, used to extract key information data in the abnormal bills and their verification detail data for different types of abnormal bills; and an association subunit, used to calculate the correlation between the key information data and the bill anomaly using a correlation analysis algorithm, obtain correlation data, and formulate the abnormal value reference rule based on the correlation data.

[0019] Optionally, in a sixth implementation method of the second aspect of the present invention, the verification module includes: a verification unit, used to obtain account balance data according to a payment instruction, extract the total bill charges in the verification detail data, and determine whether the account balance is greater than the total bill charges; a first generation unit, used to obtain a verification pass result if the account balance ≥ the total bill charges; a second generation unit, used to obtain a verification fail result if the account balance < the total bill charges, and output the account balance recharge information based on the verification fail result.

[0020] The third aspect of the present invention provides a fund account payment review device, including a memory and at least one processor, wherein the memory stores computer-readable instructions; the at least one processor calls the computer-readable instructions in the memory to execute the various steps of the fund account payment review method as described above.

[0021] A fourth aspect of the present invention provides a computer-readable storage medium having computer-readable instructions stored thereon, which, when executed by a processor, implement the various steps of the method for reviewing payment of funds to a fund account as described above.

[0022] Beneficial effect: The present invention provides a method for reviewing the payment of a fund account. The method for reviewing the payment of a fund account obtains the bill and verification detail data first, and judges whether the bill needs to be rejected according to the verification detail data to obtain a judgment result. By automatically judging the bill, the efficiency of reviewing the bill is improved; if the judgment result is that it does not need to be rejected, the judgment result that it does not need to be rejected is output to obtain a payment instruction. Through process control, the merchant can pay according to the bill only after the bill is correct, ensuring that the amount of funds is correctly issued; then the account balance data is obtained according to the payment instruction, and the account balance data is verified. If the verification passes, the payment is executed, and finally the invoice information that can be applied is generated according to the successful payment information. The method for reviewing the payment of a fund account of the present invention does not need to use manual review of the bill, and is more efficient, which can effectively shorten the courier commission payment cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A first flow chart of a method for verifying payment of a fund account provided by an embodiment of the present invention;

[0024] Figure 2 A second flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0025] Figure 3 A third flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0026] Figure 4A fourth flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0027] Figure 5 A fifth flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0028] Figure 6 A sixth flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0029] Figure 7 A seventh flow chart of the method for verifying payment of a fund account provided by an embodiment of the present invention;

[0030] Figure 8 A schematic diagram of a structure of a fund account payment review device provided by an embodiment of the present invention;

[0031] Fig. 9 Another structural schematic diagram of the fund account payment review device provided by the embodiment of the present invention;

[0032] Fig.10 A schematic diagram of the structure of a fund account payment review device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The present invention provides a method, device, equipment and storage medium for reviewing the payment of a fund account. The present invention first obtains the bill and verification detail data, and the system automatically determines whether the bill needs to be rejected according to the verification detail data, without manual review, which can effectively improve the review speed and speed up the completion of the process; if the judgment result is that it does not need to be rejected, the judgment result that it does not need to be rejected is output to obtain the payment instruction, ensure that the bill is correct when paying, and avoid the problem of amount disputes in the later stage; then obtain the account balance data according to the payment instruction, and verify the account balance data according to the verification detail data to obtain the verification result; if the verification passes, execute the payment instruction according to the verification result to obtain the payment success information, and generate the invoice information that can be applied according to the payment success information. Through the present invention, in the process of courier salary payment, the payment review speed is faster and the courier needs to wait for a shorter period of time.

[0034] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 The first embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0036] S101. Obtain the bill and verification details, and determine whether the bill needs to be rejected based on the verification details to obtain a judgment result;

[0037] Specifically, the bill usually includes multiple items, such as the name of the payee, the ID number of the payee, the order number, the payee account, the amount paid, the service fee rate, the service fee, the total amount of fees, etc. The bill is provided by the payee, and the verification details are provided by the logistics and express company and uploaded to the system in advance. When the payee needs to pay the bill, the contents of the bill are verified one by one according to the verification details.

[0038] There are many situations in which the application will be rejected after verification, such as: incorrect data format, user has not signed a contract, user name and ID number are inconsistent, the user's payment amount exceeds the limit this month, bills are duplicated, and there are multiple orders in the same batch by the same user.

[0039] S102. If the judgment result is that no rejection is required, output the judgment result that no rejection is required to obtain the payment instruction;

[0040] In this embodiment, when the judgment result is that no rejection is required, the system automatically sends the corresponding result to the client, and the client can see that the verification has passed, and can click on the "Payment" window, and then enter the correct password, and the system will obtain the payment instruction.

[0041] S103. Obtain the account balance data according to the payment instruction, and verify the account balance data according to the verification details to obtain a verification result;

[0042] S104. If the verification is passed, the payment instruction is executed according to the verification result to obtain the payment success information, and the invoice information that can be applied is generated according to the payment success information.

[0043] After the system receives the payment command, it needs to first verify whether the balance in the fund account is sufficient. If the verification fails, the system will generate a reminder of insufficient balance and the payment instruction will be terminated. The customer needs to recharge and re-initiate the payment instruction.

[0044] After the verification is passed, the system will pop up a secondary confirmation window, which displays key information such as the amount issued, service fee, and total fee. After the customer clicks "Confirm", the system executes the payment instruction. After the payment is successful, the system will receive a payment success message, which can be used as the basis for applying for invoices. The client can view the invoice information that can be applied for in the system, and can submit an invoice application to the system by clicking the corresponding application button.

[0045] In order to facilitate users to do data statistics, the system will display the payment information of the invoice that can be applied for and the corresponding bill information when displaying the invoice information that can be applied for.

[0046] In the design of the system display, the page can add total statistics to display the amount data of the full bill, the amount data of the bill that has passed verification, and the amount data of the successful payment, including the total number of transactions, the amount issued, the service fee, and the total amount of fees.

[0047] The present invention automatically verifies the bill through the system, which effectively improves the verification speed and avoids the occurrence of human audit errors. It then verifies the account balance to ensure that the payment instruction can be actually implemented, and then uses the successful payment information as the basis for applying for an invoice, so that the entire payment process has a basis to rely on and is more efficient.

[0048] See also Figure 2 The second embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0049] S201. Obtain bill and verification details, use anomaly detection algorithm, determine whether the bill needs to be rejected based on the verification details, to obtain a judgment result;

[0050] In this embodiment, in the bill verification scenario, the following anomaly detection methods can be used: Isolation Forest, K-means clustering, and density-based detection methods (such as DBSCAN).

[0051] Among them, Isolation Forest is a very effective anomaly detection algorithm, which is particularly suitable for processing high-dimensional data and does not require a large number of abnormal samples for training. Although K-means clustering is usually used for cluster analysis, it can also be used to identify outliers in this field. For example, observe data points far away from other cluster centers, which may be outliers. As for density-based methods (such as DBSCAN), they can detect abnormal areas with complex shapes and different sizes.

[0052] S202. If the judgment result is rejected, output the rejected bill to obtain a data correction bill;

[0053] When a bill is verified and determined to be rejected, the system will output the bill to the client to remind the customer that there is an abnormality in the bill. The customer is required to review and modify the bill and resubmit the data to correct the bill.

[0054] S203. The data correction bill is judged according to the verification detail data to obtain the judgment result again.

[0055] When the system receives a data-corrected bill, it will verify the resubmitted bill again based on the verification details according to the preset process. Only after the verification is passed can it enter the payment stage. Otherwise, the bill will continue to be rejected.

[0056] See also Figure 3 A third embodiment of the method for verifying payment of a fund account in the embodiment of the present invention includes:

[0057] S301. Obtain historical bill data, use the isolation forest algorithm as the initial model, and train the initial model using the historical bill data to obtain an anomaly detection model;

[0058] Isolation forest is an unsupervised learning method used to detect rare abnormal data. Its basic idea is to construct a decision tree by randomly selecting features and randomly selecting split points, and abnormal data will be isolated within a small number of splits. Since bills involve multiple dimensions, such as the name of the payee, the ID number of the payee, the order number, the payee account, the amount issued, the service fee rate, the service fee, the total amount of fees, etc., the isolation forest algorithm is used as the initial model to build the model, and the prediction accuracy of the obtained anomaly detection model will be better.

[0059] S302. Obtain abnormal billing data, and optimize the anomaly detection model according to the abnormal billing data;

[0060] In the daily abnormal bill data collection process, when abnormal bills are found manually, they can be collected for adjustment or optimization of the anomaly detection model parameters, and can also be added to the training set for training the anomaly detection model. If the anomaly detection model also identifies abnormal bills during operation, these abnormal bills will be marked as abnormal and can be copied and stored separately. These abnormal bills automatically identified by the system will be submitted to the manual review team for secondary confirmation, and can be further used to optimize the anomaly detection model based on the results of the manual review, such as updating the training data set or adjusting the model parameters.

[0061] S303. Obtain the bill and verification detail data, use the optimized anomaly detection model, and determine whether the bill needs to be rejected based on the verification detail data to obtain a judgment result.

[0062] Verifying bills through the anomaly detection model running in the system can reduce errors caused by human negligence, greatly shorten processing time, and improve work efficiency. In addition to improving efficiency, it can also reduce risks. For example, it can help detect potential fraudulent behaviors such as false transactions and duplicate payments, protecting users from financial losses.

[0063] See also Figure 4 The fourth embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0064] S401. Obtain historical billing data, the historical billing data including billing amount, transaction date, user ID, transaction type and geographic location;

[0065] S402. Using the isolation forest algorithm as the initial model, the initial model is trained using the bill amount, transaction date, user ID, transaction type and geographic location as training features to obtain training data;

[0066] S403. Determine a threshold value according to the anomaly score distribution of the training data, and adjust the initial model according to the threshold value to obtain an anomaly detection model.

[0067] In the construction of the anomaly detection model, feature engineering needs to be completed first. Among them, the bill amount can be used directly as a continuous variable; the transaction date can be converted into a periodic feature (such as the day of the week, the time period of the day, etc.); the user ID can be one-hot encoded or use user behavior characteristics; the transaction type can be encoded, such as recharge, consumption, etc.; the geographic location can use geographic coordinates or be encoded.

[0068] After completing feature engineering, select the Isolation Forest Algorithm and set the parameters of the Isolation Forest Algorithm, such as determining the number of trees in the Isolation Forest, the maximum number of samples, and other hyperparameters. After the parameters are set, use the historical bill data to train the model.

[0069] When constructing the isolation forest model, each isolated tree is first constructed. The steps include:

[0070] 1. Extract subsamples: extract samples from the training set with replacement as training data for the current tree;

[0071] 2. Build the tree:

[0072] Initialize node: use the extracted subsample as the data set of the root node;

[0073] Randomly select features: randomly select a feature in the data set of the current node;

[0074] Randomly select split point: randomly select a value in the value range of the selected feature as the split point;

[0075] Divide the data set: Divide the data set into two subsets according to the split point and assign them to the left and right child nodes of the current node respectively;

[0076] Recursive partitioning: Repeat the above process for the left and right child nodes until the stopping condition is met (for example, there is only one sample in the child node, or the height of the tree reaches the preset maximum value l).

[0077] Repeat the above steps to build multiple isolated trees to form an isolation forest.

[0078] For each sample in the test set, calculate its path length on each tree (the number of edges from the root node to the leaf node). The anomaly score of the sample is obtained by normalizing the average of its path lengths on all trees. The higher the anomaly score, the more likely the sample is an anomaly. When judging bill anomalies, a threshold can be set based on the anomaly score, and samples above the threshold are considered anomalies. The threshold can also be determined based on business needs, or based on the anomaly score distribution of the training data, for example, the top 1% of anomaly scores are selected as the threshold.

[0079] See also Figure 5 The fifth embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0080] S501. Obtain historical abnormal bills and their verification details, and formulate abnormal value reference rules based on the historical abnormal bills and their verification details;

[0081] S502. If the judgment result is rejection, find possible outliers in the rejected bill according to the outlier reference rule;

[0082] S503. Mark possible outliers in the rejected bill and output the rejected bill to obtain a data correction bill.

[0083] After the abnormal bill is identified by the anomaly detection model, it needs to be rejected and returned to the client (bill upload end) for review. If the system only outputs the judgment result without clarifying the reason for the rejection, the customer will need to spend a lot of time to check the rejected bill item by item, which will greatly slow down the payment process.

[0084] In this embodiment, if the bill is rejected due to problems such as incorrect data format, user not signing a contract, user name and ID number not matching, user's monthly payment amount exceeding the limit, duplicate bills, multiple orders in the same batch by the same user, etc., the system will directly display the corresponding rejection reason. In other more complex situations, such as bill rejection due to multiple data anomalies in the bill, the total amount of payment may involve multiple calculation parameters. If multiple calculation parameters are abnormal, the abnormal data cannot be directly specified. This embodiment formulates abnormal value reference rules by combining historical abnormal bills and their verification details. When a bill is rejected, the system can mark it according to possible abnormal values ​​in the rejected bill, and the output rejected bill will mark the data that may lead to the rejection. When checking, customers can give priority to the data that the system considers to be abnormal according to the prompts, thereby improving the inspection efficiency.

[0085] See also Figure 6 The sixth embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0086] S601. Obtain abnormal bills and their verification details, and classify the abnormal bills to obtain various types of abnormal bills;

[0087] Since the bill involves multiple contents, and most of the contents are unrelated to each other, there are also many types of reasons that lead to bill abnormalities. By classifying the abnormal bill data, the causes of the abnormalities can be better classified and analyzed, thereby improving the accuracy of the prediction.

[0088] S602. For different types of abnormal bills, extract key information data from the abnormal bills and their verification details;

[0089] S603. Use a correlation analysis algorithm to calculate the correlation between the key information data and the bill anomaly, obtain correlation data, and formulate the abnormal value reference rule based on the correlation data.

[0090] By extracting key information data from each type of abnormal bill data in a targeted manner and using the key information data for subsequent correlation analysis, the accuracy of the analysis results can be made higher.

[0091] The bill verification through the anomaly detection model will eventually get a score. When the score falls within the anomaly judgment range, the bill will be rejected. The outlier reference rules formulated based on the correlation data can be used to estimate the features that cause the outliers based on the size of the outliers, the proportion of feature parameters, etc. Based on the correlation data, the features that may cause the outliers can also be sorted by the size of the correlation.

[0092] See also Figure 7 The seventh embodiment of the method for verifying the payment of a fund account in the embodiment of the present invention includes:

[0093] S701. Obtain the account balance data according to the payment instruction, extract the total bill fee in the verification detail data, and determine whether the account balance is greater than the total bill fee;

[0094] S702. If the account balance ≥ the total bill fee, obtain the verification result;

[0095] Specifically, when the balance verification is passed, the system will pop up a page window for secondary confirmation of the payment to ensure that the payment amount, service fee, and total fee are correct, thereby reducing unnecessary subsequent erroneous operations.

[0096] S703. If the account balance is less than the total amount of the bill, a verification failure result is obtained, and the account balance recharge information is output according to the verification failure result.

[0097] If the verification fails, the system will inform the user of insufficient balance through a pop-up window, and guide the user to recharge first and then make a payment. The payment will not be initiated for this batch. The pop-up window will display the payment amount, service fee, total fee, and available balance in the account.

[0098] The above describes the method for verifying the payment of funds to a fund account in an embodiment of the present invention. The following describes the device for verifying the payment of funds to a fund account in an embodiment of the present invention. Figure 8 In one embodiment of the present invention, a payment review device for a fund account includes:

[0099] The judgment module 10 is used to obtain the bill and verification detail data, and judge whether the bill needs to be rejected according to the verification detail data to obtain a judgment result;

[0100] The output module 20 is used for outputting the judgment result that no rejection is required if the judgment result is that no rejection is required, so as to obtain the payment instruction;

[0101] A verification module 30, for obtaining the account balance data according to the payment instruction, and verifying the account balance data according to the verification detail data to obtain a verification result;

[0102] The generation module 40 is used to execute the payment instruction according to the verification result if the verification is passed, so as to obtain the payment success information, and generate the invoice information that can be applied according to the payment success information.

[0103] See also Fig. 9 In one embodiment of the present invention, a payment review device for a fund account includes:

[0104] The judgment module 10 is used to obtain the bill and verification detail data, and judge whether the bill needs to be rejected according to the verification detail data to obtain a judgment result;

[0105] The output module 20 is used for outputting the judgment result that no rejection is required if the judgment result is that no rejection is required, so as to obtain the payment instruction;

[0106] A verification module 30, for obtaining the account balance data according to the payment instruction, and verifying the account balance data according to the verification detail data to obtain a verification result;

[0107] A generating module 40 is used to execute the payment instruction according to the verification result if the verification is passed, so as to obtain the payment success information, and generate the invoice information that can be applied according to the payment success information;

[0108] In this embodiment, the judging module 10 includes:

[0109] The calculation submodule 11 is used to obtain the bill and verification detail data, and use an abnormality detection algorithm to determine whether the bill needs to be rejected according to the verification detail data to obtain a judgment result;

[0110] The output submodule 12 is used to output the rejected bill if the judgment result is rejection, so as to obtain a data-corrected bill;

[0111] The judgment submodule 13 is used to judge the data correction bill according to the verification detail data to obtain the judgment result again;

[0112] In this embodiment, the calculation submodule 11 includes:

[0113] A construction unit 111 is used to obtain historical bill data, use an isolation forest algorithm as an initial model, and use the historical bill data to train the initial model to obtain an anomaly detection model;

[0114] An optimization unit 112, configured to obtain abnormal billing data and optimize an abnormality detection model according to the abnormal billing data;

[0115] The judgment unit 113 is used to obtain the bill and verification detail data, and use the optimized anomaly detection model to judge whether the bill needs to be rejected according to the verification detail data to obtain a judgment result;

[0116] In this embodiment, the construction unit 111 includes:

[0117] The acquisition subunit 1111 is used to acquire historical bill data, where the historical bill data includes bill amount, transaction date, user ID, transaction type and geographic location;

[0118] The training subunit 1112 is used to train the initial model using the isolation forest algorithm as an initial model and the bill amount, transaction date, user ID, transaction type and geographic location as training features to obtain training data;

[0119] An adjustment subunit 1113 is used to determine a threshold value according to the abnormal score distribution of the training data, and adjust the initial model according to the threshold value to obtain an abnormality detection model;

[0120] In this embodiment, the output submodule 12 includes:

[0121] A formulation unit 121 is used to obtain historical abnormal bills and verification details thereof, and formulate abnormal value reference rules according to the historical abnormal bills and verification details thereof;

[0122] A searching unit 122, configured to search for possible abnormal values ​​in the rejected bill according to the abnormal value reference rule if the judgment result is rejection;

[0123] A marking unit 123, used for marking possible abnormal values ​​in the rejected bill, outputting the rejected bill, so as to obtain a data correction bill;

[0124] In this embodiment, the formulation unit 121 includes:

[0125] The classification subunit 1211 is used to obtain abnormal bills and their verification details, and classify the abnormal bills to obtain multiple types of abnormal bills;

[0126] The extraction subunit 1212 is used to extract key information data in the abnormal bills and their verification details for different types of abnormal bills;

[0127] The association subunit 1213 is used to calculate the association between the key information data and the bill anomaly by using a correlation analysis algorithm, obtain the correlation data, and formulate the abnormal value reference rule according to the correlation data;

[0128] In this embodiment, the verification module 30 includes:

[0129] The verification unit 31 is used to obtain the account balance data according to the payment instruction, extract the total amount of the bill fee in the verification detail data, and determine whether the account balance is greater than the total amount of the bill fee;

[0130] The first generating unit 32 is used to obtain a verification result if the account balance ≥ the total amount of the bill;

[0131] The second generating unit 33 is used to obtain a verification failure result if the account balance is less than the total bill charge, and output the account balance recharging information according to the verification failure result.

[0132] The payment review device for a fund account provided by the present invention determines whether the bill needs to be rejected based on the verification details after obtaining the bill. If it needs to be rejected, the rejected bill is returned to the customer for review and modification, and the modified and submitted bill is generated for the second time; if it is determined that the bill does not need to be rejected, the account balance is verified based on the bill details. If the total amount of the bill is greater than the available balance in the account, the user is notified to recharge the account balance; finally, based on the successful payment information, an invoice batch that can be applied is generated. Through this device, the bill payment efficiency of the logistics account can be effectively improved, and the verification accuracy of abnormal bills can be improved.

[0133] The above is a detailed description of the payment review device for the capital account in the embodiment of the present invention from the perspective of modular functional entities. The following is a detailed description of the payment review device for the capital account in the embodiment of the present invention from the perspective of hardware processing.

[0134] Fig.10 A schematic diagram of the structure of a fund account payment review device provided in an embodiment of the present invention, the fund account payment review device 900 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 910 (for example, one or more processors) and a memory 920, and one or more storage media 930 (for example, one or more mass storage devices) storing application programs 933 or data 932. Among them, the memory 920 and the storage medium 930 can be temporary storage or permanent storage. The program stored in the storage medium 930 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the fund account payment review device 900. Furthermore, the processor 910 can be configured to communicate with the storage medium 930, and execute a series of instruction operations in the storage medium 930 on the fund account payment review device 900 to implement the steps of the fund account payment review method provided in the above-mentioned method embodiments.

[0135] The fund account payment audit device 900 may also include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input and output interfaces 960, and / or one or more operating systems 931, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Fig.10 The structure of the fund account payment review device shown does not constitute a limitation on the fund account payment review device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0136] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of a method for reviewing payment of a fund account.

[0137] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the equipment or device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0139] It is understandable that those skilled in the art can make equivalent substitutions or changes based on the technical solution and inventive concept of the present invention, and all these changes or substitutions should fall within the protection scope of the claims attached to the present invention.

Claims

1. A method for verifying payment of a fund account, characterized in that: The steps include: Obtaining bill and verification detail data, and judging whether the bill needs to be rejected according to the verification detail data to obtain a judgment result; If the judgment result is that no rejection is required, output the judgment result that no rejection is required to obtain the payment instruction; Acquire the account balance data according to the payment instruction, and verify the account balance data according to the verification detail data to obtain a verification result; If the verification is passed, the payment instruction is executed according to the verification result to obtain the payment success information, and the invoice information that can be applied is generated according to the payment success information.

2. The method for verifying the payment of a fund account according to claim 1, characterized in that: The obtaining of the bill and the verification detail data, and judging whether the bill needs to be rejected according to the verification detail data to obtain a judgment result specifically includes: Obtain the bill and verification detail data, and use an anomaly detection algorithm to determine whether the bill needs to be rejected based on the verification detail data to obtain a judgment result; If the judgment result is rejection, the rejected bill is output to obtain a data correction bill; The data correction bill is judged according to the verification detail data to regain the judgment result.

3. The method for verifying the payment of a fund account according to claim 2, characterized in that: The obtaining of the bill and verification detail data, using an anomaly detection algorithm, and judging whether the bill needs to be rejected according to the verification detail data to obtain a judgment result specifically includes: Obtain historical bill data, use the isolation forest algorithm as the initial model, and use the historical bill data to train the initial model to obtain an anomaly detection model; Acquire abnormal billing data, and optimize the anomaly detection model according to the abnormal billing data; Obtain the bill and verification detail data, use the optimized anomaly detection model, and determine whether the bill needs to be rejected based on the verification detail data to obtain a judgment result.

4. The method for verifying the payment of a fund account according to claim 3, characterized in that: The acquisition of historical bill data, taking the isolation forest algorithm as the initial model, and training the initial model with the historical bill data to obtain an anomaly detection model specifically includes: Obtain historical billing data, including billing amount, transaction date, user ID, transaction type, and geographic location; Using the isolation forest algorithm as an initial model, and using the bill amount, transaction date, user ID, transaction type, and geographic location as training features to train the initial model to obtain training data; A threshold is determined according to the anomaly score distribution of the training data, and the initial model is adjusted according to the threshold to obtain an anomaly detection model.

5. The method for verifying the payment of funds to a capital account according to claim 4, characterized in that: If the judgment result is rejection, outputting the rejected bill to obtain a data correction bill specifically includes: Obtain historical abnormal bills and their verification details, and formulate abnormal value reference rules based on the historical abnormal bills and their verification details; If the judgment result is rejection, possible abnormal values ​​in the rejected bill are searched according to the abnormal value reference rule; Possible outliers in the rejection bill are marked and the rejection bill is output to obtain a data correction bill.

6. The method for verifying the payment of funds to a capital account according to claim 5, characterized in that: The obtaining of historical abnormal bills and verification details thereof, and formulating abnormal value reference rules according to the historical abnormal bills and verification details thereof, specifically includes: Acquire abnormal bills and verification details thereof, and classify the abnormal bills to obtain various types of abnormal bills; For different types of abnormal bills, extract key information data from the abnormal bills and their verification details; A correlation analysis algorithm is used to calculate the correlation between the key information data and the bill anomaly, to obtain correlation data, and the abnormal value reference rule is formulated according to the correlation data.

7. The method for verifying the payment of funds to a capital account according to claim 1, characterized in that: The acquiring of the account balance data according to the payment instruction and the verification of the account balance data according to the verification detail data to obtain the verification result specifically include: Obtaining account balance data according to the payment instruction, extracting the total amount of bill charges in the verification detail data, and determining whether the account balance is greater than the total amount of bill charges; If the account balance is ≥ the total amount of bill charges, the verification result is obtained; If the account balance is less than the total amount of the bill charges, a verification failure result is obtained, and the account balance recharge information is output according to the verification failure result.

8. A fund account payment review device, characterized in that: include: A judgment module is used to obtain the bill and verification detail data, and judge whether the bill needs to be rejected according to the verification detail data to obtain a judgment result; An output module is used for outputting the judgment result that no rejection is required if the judgment result is that no rejection is required, so as to obtain a payment instruction; A verification module, used to obtain the account balance data according to the payment instruction, and verify the account balance data according to the verification detail data to obtain a verification result; The generation module is used to execute the payment instruction according to the verification result if the verification is passed, so as to obtain the payment success information, and generate the invoice information that can be applied for according to the payment success information.

9. A fund account payment review device, characterized in that: comprising a memory and at least one processor, wherein the memory has computer-readable instructions stored therein; The at least one processor calls the computer-readable instructions in the memory to execute the various steps of the method for reviewing payment of a fund account as described in any one of claims 1-7.

10. A computer-readable storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by the processor, the various steps of the method for reviewing payment to a fund account as described in any one of claims 1-7 are implemented.

Images