Abnormal account checking data processing method and device, equipment, medium and product

Abstract

The invention provides an abnormal reconciliation data processing method and device, equipment, a medium and a product, which can be applied to the field of big data. The method comprises: in response to an account checking request, obtaining a first bill data set and a first flow data set, the first bill data set comprising respective bill attribute data of a plurality of bills, and the first flow data set comprising respective flow attribute data of a plurality of flows; according to a preset first matching rule, matching the respective attribute data of the plurality of bills with the respective attribute data of the plurality of receipts so as to respectively screen out matched bill attribute data and receipt attribute data from the first bill data set and the first receipt data set to obtain a second bill data set and a second receipt data set; according to a preset second matching rule, matching the bill attribute data with the flow attribute data to screen unmatched bills and flows to obtain an abnormal reconciliation data set; and generating a data early warning event based on the abnormal reconciliation data set.

Description

Technical Field

[0001] This application relates to the field of big data, specifically to the field of financial technology, and more specifically to a method, apparatus, device, medium, and product for processing abnormal reconciliation data. Background Technology

[0002] In accounting practice, a large amount of cash flow and invoices need to be matched and verified before reconciliation processing. In related technologies, cash and invoice reconciliation schemes mainly rely on manual review and simple rule matching.

[0003] In realizing the concept disclosed herein, the relevant technologies have at least the following problems: During the manual reconciliation process, due to the presence of human factors and the limitations of rule matching, the existing reconciliation schemes are prone to mismatches or omissions, which affects the compliance and accuracy of fund usage. Summary of the Invention

[0004] In view of the above problems, this application provides a method, apparatus, equipment, medium and program product for processing abnormal reconciliation data to improve reconciliation efficiency.

[0005] According to a first aspect of this application, a method for processing abnormal reconciliation data is provided, comprising: in response to a reconciliation request, obtaining a first invoice dataset and a first transaction dataset, wherein the first invoice dataset includes invoice attribute data of multiple invoices and the first transaction dataset includes transaction attribute data of multiple transactions; matching the attribute data of the multiple invoices and the attribute data of the multiple transactions according to a preset first matching rule, so as to filter out matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset respectively, to obtain a second invoice dataset and a second transaction dataset; matching the invoice attribute data included in the second invoice dataset with the transaction attribute data included in the second transaction dataset according to a preset second matching rule, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, to obtain an abnormal reconciliation dataset; and generating a data warning event based on the abnormal reconciliation dataset.

[0006] According to an embodiment of this application, based on a preset first matching rule, the attribute data of multiple invoices and the attribute data of multiple transaction records are matched to filter out matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset, respectively, to obtain a second invoice dataset and a second transaction dataset. This includes: matching the attribute data of multiple transaction records with the attribute data of multiple invoices to obtain multiple first similarities; determining at least one first target similarity from the multiple first similarities whose similarity is greater than a first threshold; filtering out invoice attribute data related to at least one first target similarity from the first invoice dataset, and filtering out transaction attribute data related to at least one first target similarity from the first transaction dataset, respectively, to obtain a second invoice dataset and a second transaction dataset.

[0007] According to an embodiment of this application, based on a preset second matching rule, the invoice attribute data included in the second invoice dataset is matched with the transaction attribute data included in the second transaction dataset to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset, respectively, to obtain an abnormal reconciliation dataset. This includes: matching the current invoice attribute data in the second invoice dataset with multiple transaction attribute data in the second transaction dataset to obtain a transaction reconciliation dataset related to the current invoice; matching the current transaction attribute data in the second transaction dataset with multiple invoice attribute data in the second invoice dataset to obtain an invoice reconciliation dataset related to the current transaction; and obtaining an abnormal reconciliation dataset based on the transaction reconciliation datasets related to multiple invoices and the invoice set reconciliation datasets related to multiple transactions.

[0008] According to an embodiment of this application, the invoice attribute data includes first invoice data, and the transaction flow attribute data includes first transaction flow data; the current invoice attribute data in the second invoice dataset is matched with multiple transaction flow attribute data in the second transaction flow dataset to obtain a transaction flow reconciliation dataset related to the current invoice; this includes: matching the first invoice data of the current invoice with the first transaction flow data of multiple transactions in the second transaction flow dataset to obtain a first transaction flow reconciliation dataset related to the current invoice; and determining the transaction flow reconciliation dataset related to the current invoice based on the current invoice and the first transaction flow reconciliation dataset.

[0009] According to an embodiment of this application, the first invoice data of the current invoice is matched with the first transaction data of multiple transactions in the second transaction data set to obtain a first transaction reconciliation dataset related to the current invoice; the process includes: matching the first invoice data of the current invoice with the first transaction data of multiple transactions in the second transaction data set to obtain multiple transaction similarities related to the current invoice; determining multiple target transaction similarities with similarities greater than a second threshold from the multiple transaction similarities of the current invoice; and using the transaction attribute data related to the multiple target transaction similarities as the first transaction reconciliation dataset related to the current invoice.

[0010] According to an embodiment of this application, the invoice attribute data further includes second invoice data, and the transaction history attribute data further includes second transaction history data. Based on the current invoice and the first transaction history reconciliation dataset, determining the transaction history reconciliation dataset related to the current invoice includes: randomly selecting at least two second transaction history data from the multiple transaction history attribute data included in the first transaction history reconciliation dataset to form a transaction history reconciliation subset, thereby obtaining multiple transaction history reconciliation subsets related to the current invoice; obtaining subset reconciliation data of the transaction history reconciliation subsets based on the multiple second transaction history data in the transaction history reconciliation subsets; matching the second invoice data of the current invoice with the subset reconciliation data of each of the multiple transaction history reconciliation subsets to obtain transaction history matching results; and determining the transaction history reconciliation dataset related to the current invoice based on the transaction history matching results.

[0011] According to an embodiment of this application, the invoice attribute data includes first invoice data, and the transaction history attribute data includes first transaction history data; the current transaction history attribute data in the second transaction history dataset is matched with multiple invoice attribute data in the second invoice dataset to obtain an invoice reconciliation dataset related to the current transaction history, including: matching the first transaction history data of the current transaction history with the first invoice data of multiple invoices in the second invoice dataset to obtain a first invoice reconciliation dataset related to the current transaction history; and determining the invoice reconciliation dataset related to the current transaction history based on the current transaction history and the first invoice reconciliation dataset.

[0012] According to an embodiment of this application, the first transaction data of the current transaction is matched with the first bill data of multiple bills in the second bill dataset to obtain a first bill reconciliation dataset related to the current transaction; including: matching the first transaction data of the current transaction with the first bill data of multiple bills in the second bill dataset to obtain the similarity of multiple bills related to the current transaction; determining multiple target bill similarities with a similarity greater than a third threshold from the multiple bill similarities of the current transaction; and using the bill attribute data related to the similarity of the multiple target bills as the first bill reconciliation dataset related to the current transaction.

[0013] According to an embodiment of this application, the invoice attribute data further includes second invoice data, and the transaction history attribute data further includes second transaction history data. Based on the current transaction history and the first invoice reconciliation dataset, determining the invoice reconciliation dataset related to the current transaction history includes: randomly selecting at least two second invoice data from the multiple invoice attribute data included in the first invoice reconciliation dataset to form an invoice reconciliation subset, thereby obtaining multiple invoice reconciliation subsets related to the current transaction history; obtaining subset reconciliation data of the invoice reconciliation subsets based on the multiple second invoice data in the invoice reconciliation subsets; matching the second transaction history data of the current transaction history with the subset reconciliation data of each of the multiple invoice reconciliation subsets to obtain invoice matching results; and determining the invoice reconciliation dataset related to the current transaction history based on the invoice matching results.

[0014] According to an embodiment of this application, an abnormal reconciliation dataset is obtained based on a transaction reconciliation dataset associated with multiple invoices and a set of invoices associated with multiple transactions. This includes: filtering out transaction attribute data from the transaction reconciliation data corresponding to each of the multiple invoices from the second transaction dataset to obtain an abnormal transaction dataset; filtering out invoice attribute data from the invoice reconciliation data corresponding to each of the multiple transactions from the second invoice dataset to obtain an abnormal invoice dataset; and using the abnormal transaction dataset and the abnormal invoice dataset as the abnormal reconciliation dataset.

[0015] According to an embodiment of this application, the transaction attribute data includes transaction information subjects, and the invoice attribute data includes invoice information subjects. Based on the abnormal reconciliation dataset, a data warning event is generated, including: based on the transaction information subjects and invoice information subjects in the abnormal reconciliation dataset; obtaining an information subject set; and generating a data warning event based on the information subjects in the information subject set.

[0016] A second aspect of this application provides an abnormal reconciliation data processing apparatus, comprising: an acquisition module, configured to acquire a first invoice dataset and a first transaction dataset in response to a reconciliation request, wherein the first invoice dataset includes invoice attribute data of multiple invoices and the first transaction dataset includes transaction attribute data of multiple transactions.

[0017] The first matching module is used to match the attribute data of multiple invoices and the attribute data of multiple transaction records according to the preset first matching rules, so as to filter out the matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset respectively, and obtain the second invoice dataset and the second transaction dataset.

[0018] The second matching module is used to match the invoice attribute data included in the second invoice dataset with the transaction attribute data included in the second transaction dataset according to the preset second matching rules, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, and obtain the abnormal reconciliation dataset.

[0019] The early warning module is used to generate data early warning events based on abnormal reconciliation datasets.

[0020] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0021] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0022] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0023] According to the embodiments of this disclosure, a first round of matching is used to filter and remove highly correlated invoice attribute data and transaction history attribute data. Then, a second round of matching precisely locates completely uncorrelated invoices and transaction history data, ensuring that the abnormal reconciliation dataset contains only isolated invoice attribute information and transaction history attribute information that are completely uncorrelated. These isolated data are then further flagged for alerts to prevent violations. In this way, not only is matching efficiency improved through multiple rounds of data filtering, but also some complex and special invoice and transaction history correspondences can be identified when filtering abnormal data, further solving the problem of high manpower costs associated with manual identification, making the reconciliation work more accurate and efficient. Attached Figure Description

[0024] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0025] Figure 1 The illustration shows an application scenario diagram of the abnormal reconciliation data processing method, apparatus, device, medium, and program product according to embodiments of this application;

[0026] Figure 2 A flowchart illustrating an abnormal reconciliation data processing method according to an embodiment of this application is shown in the schematic diagram.

[0027] Figure 3 The illustration shows a schematic diagram of the principle of determining the transaction reconciliation dataset according to an embodiment of the present disclosure;

[0028] Figure 4 The illustration shows a schematic diagram of the principle of determining the bill reconciliation dataset according to an embodiment of the present disclosure;

[0029] Figure 5 This schematic diagram illustrates the structural block diagram of an abnormal reconciliation data processing apparatus according to an embodiment of this application;

[0030] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing an abnormal reconciliation data processing method according to an embodiment of this application. Detailed Implementation

[0031] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0032] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0033] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0034] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0035] In the technical solution of this application, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0036] In accounting practice, enterprises need to use legal and valid original vouchers (especially VAT special invoices, ordinary invoices, and bank statements) as the basis for cost and expense recognition and pre-tax deduction. Current accounting standards and tax administration systems both require consistency among the "invoices, accounts, and funds"—that is, the contract flow, invoice flow, and fund flow must maintain a consistent relationship in terms of time, amount, counterparty, and business substance—in order to meet the compliance requirements for pre-tax deduction of corporate income tax and VAT input tax credit.

[0037] Currently, matching of bills and cash flow is mainly done through manual comparison or rule-based matching by the financial system. However, the above methods generally have the following drawbacks.

[0038] (1) Inefficiency: The manual matching of bills and cash flow requires a lot of manpower and time, resulting in low efficiency of the entire reconciliation process.

[0039] (2) Manual and financial system rule matching is difficult to deal with complex situations: For complex transaction and bill data, such as one bill corresponding to multiple funds, or one fund issuing multiple bills, these complex correspondences are normal data, but the existing rule matching methods will identify them as market data, and it is even more difficult to match these data based on manual matching.

[0040] In view of this, embodiments of this application provide a method for processing abnormal reconciliation data, comprising the following steps: in response to a reconciliation request, obtaining a first invoice dataset and a first transaction dataset, wherein the first invoice dataset includes invoice attribute data of multiple invoices and the first transaction dataset includes transaction attribute data of multiple transactions; matching the attribute data of the multiple invoices and the attribute data of the multiple transactions according to a preset first matching rule, so as to filter out matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset respectively, to obtain a second invoice dataset and a second transaction dataset; matching the invoice attribute data included in the second invoice dataset with the transaction attribute data included in the second transaction dataset according to a preset second matching rule, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, to obtain an abnormal reconciliation dataset; and generating a data warning event based on the abnormal reconciliation dataset.

[0041] Figure 1 The illustration shows an application scenario diagram of the abnormal reconciliation data processing method, apparatus, device, medium, and program product according to embodiments of this application;

[0042] like Figure 1As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0043] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0044] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0045] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0046] It should be noted that the abnormal reconciliation data processing method provided in this application embodiment can generally be executed by server 105. Correspondingly, the abnormal reconciliation data processing device provided in this application embodiment can generally be located in server 105. The abnormal reconciliation data processing method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the abnormal reconciliation data processing device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0047] It should be understood that Figure 1The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0048] The following will be based on Figure 1 The described scene, through Figure 2 The method for processing abnormal reconciliation data according to the embodiments of this application will be described in detail.

[0049] Figure 2 A flowchart illustrating an abnormal reconciliation data processing method according to an embodiment of this application is shown.

[0050] like Figure 2 As shown, the abnormal reconciliation data processing method of this embodiment includes operations S210 to S240.

[0051] In operation S210, in response to the reconciliation request, the first invoice dataset and the first transaction dataset are obtained. The first invoice dataset includes the invoice attribute data of multiple invoices, and the first transaction dataset includes the transaction attribute data of multiple transactions.

[0052] In the embodiments disclosed in this application, a bill can refer to a payment voucher issued by a value-added tax taxpayer or tax authority in accordance with the law. This voucher records the legally required elements such as the names of the transacting parties, the name, quantity, amount, tax rate, and tax amount of the goods or taxable services, and serves as an original document with legal validity for pre-tax deduction, input tax credit, and financial accounting. A cash flow statement can refer to a detailed record of all fund inflows and outflows actually occurring within a specific period through a bank account or other payment channels for an enterprise or individual.

[0053] In the embodiments disclosed in this application, the first invoice dataset and the first transaction dataset can be data that can be retrieved at any time by prior system retrieval or manual input and stored in a storage device. The first invoice dataset can include invoice attribute data recorded on multiple invoices, and the first transaction dataset can include transaction attribute data corresponding to multiple transactions.

[0054] In the embodiments disclosed in this application, the bill attribute data can be the header name, tax number, address, telephone number, bank address, account number, bill number, code, type, amount excluding tax, tax amount, and total price including tax recorded on the bill. The transaction history attribute data can be the counterparty's name, counterparty's account, payment amount, income amount, remarks, and summary in the transaction history.

[0055] In operation S220, according to the preset first matching rule, the attribute data of each of the multiple tickets and the attribute data of each of the multiple transactions are matched, so as to filter out the matching ticket attribute data and transaction attribute data from the first ticket dataset and the first transaction dataset respectively, to obtain the second ticket dataset and the second transaction dataset.

[0056] In operation S230, according to the preset second matching rule, the invoice attribute data included in the second invoice dataset is matched with the transaction attribute data included in the second transaction dataset, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, and obtain the abnormal reconciliation dataset.

[0057] In the embodiments disclosed in this application, RNN (Recurrent Neural Network), LSTM (Long Short-Term Memory), or DenseNet (Dense Convolutional Network) models can be used to match invoice attribute data and transaction history attribute data based on a preset first matching rule or a preset second matching rule. This yields matching results for the two sets of attribute data. The matching results can be based on similarity, difference heatmaps, element-level residual matrices, or alignment paths to determine the relationship between invoices and transaction history. The aforementioned matching process can also be any association operation that can establish a correlation between invoice attribute data and transaction history attribute data.

[0058] In the embodiments disclosed in this application, after obtaining related invoices and transaction records in S220, invoice attribute data and transaction record attribute data with strong correlation are obtained. Then, the aforementioned invoice attribute data is filtered out from the first invoice dataset, and the aforementioned transaction record attribute information is filtered out from the first transaction record dataset, resulting in a second invoice dataset and a second transaction record dataset. Therefore, the second invoice dataset and the second transaction record dataset can be invoice attribute data and transaction record attribute data that do not conform to strong correlation obtained after coarse matching.

[0059] In the embodiments disclosed in this application, in S230, the invoice attribute data and transaction attribute data in the second invoice dataset and the second transaction data dataset are further matched. This step can filter out invoice attribute data and transaction attribute data with weak correlation. Then, the invoice attribute data and transaction attribute data with weak correlation are filtered out from the second invoice dataset and the second transaction data dataset, so that invoice attribute data and transaction attribute data with no correlation can be obtained. Finally, these invoice attribute data and transaction attribute data are used as an abnormal reconciliation dataset for subsequent data warning.

[0060] In the embodiments disclosed in this application, the first matching rule can be based on overall matching of subject information, amount, or date. Based on these three attributes, completely matching invoice attribute data and transaction history attribute data can be filtered out. The second matching rule can be a further matching based on amount, for example, multiple transactions correspond to one invoice. The sum of the amounts of each of the multiple transactions can then correspond to this single invoice, thus filtering out complex reconciliation situations that are difficult to filter out using ordinary matching mechanisms. The subject information mentioned above refers to the header information for invoices and the counterparty's account name for transaction history.

[0061] In operation S240, a data alert event is generated based on the abnormal reconciliation dataset.

[0062] Regarding the aforementioned abnormal reconciliation dataset, corresponding early warning events should be generated. Specifically, if an invoice has only been issued to a company but no funds have been transferred, the company should be promptly notified to make payment to prevent violations. If funds have been transferred but no invoice has been issued, the next step should be taken promptly to issue the invoice based on the corresponding inflow record, again to prevent violations.

[0063] In the embodiments disclosed in this application, the method for processing the aforementioned abnormal data firstly uses a first round of matching to filter out and remove invoice attribute data and transaction history attribute data with strong correlation. Then, a second round of matching precisely locates completely uncorrelated invoices and transaction history, ensuring that the abnormal reconciliation dataset contains only isolated invoice attribute information and transaction history attribute information that are completely uncorrelated. These isolated data are then further flagged for warning to prevent violations. In this way, not only is matching efficiency improved through multiple rounds of data filtering, but also some complex and special invoice and transaction history correspondences can be filtered out during the abnormal data filtering process, further saving the significant manpower costs consumed by manual identification and making the reconciliation work more accurate and efficient.

[0064] In the embodiments disclosed in this application, according to a preset first matching rule, the attribute data of multiple invoices and the attribute data of multiple transaction records are matched to filter out matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset, respectively, to obtain a second invoice dataset and a second transaction dataset. This includes: matching the attribute data of multiple transaction records with the attribute data of multiple invoices to obtain multiple first similarities; determining at least one first target similarity from the multiple first similarities whose similarity is greater than a first threshold; filtering out invoice attribute data related to at least one first target similarity from the first invoice dataset, and filtering out transaction attribute data related to at least one first target similarity from the first transaction dataset, respectively, to obtain a second invoice dataset and a second transaction dataset.

[0065] In the embodiments disclosed in this application, the preset first matching rule can be based on the evaluation of multiple attribute data. First, a set of multiple bill attribute data and a set of transaction attribute data from the bills are used. The aforementioned attribute data sets can include multiple dimensions for both the bills and the transactions, such as amount, time, counterparty, and business type. For each dimension, a similarity operator corresponding to that dimension can be called to obtain the sub-similarity of the two attribute information under that dimension. After normalization, each sub-similarity is synthesized into a total similarity according to a preset weight. If the total similarity meets a first threshold condition, i.e., a first target similarity, the bill attribute information and the transaction attribute information corresponding to the first target similarity are marked as matched bills and transactions. The above calculation steps are repeated until all matched bills and transactions are found. Then, the bill attribute data and transaction attribute data corresponding to all matched bills and transactions are removed from the first bill dataset and the first transaction dataset, respectively. The remaining attribute data constitutes the second bill dataset and the second transaction dataset. The preset weights and first threshold can be set manually or adaptively updated using machine learning based on historical matching results to ensure the robustness of the model in different periods or business scenarios.

[0066] In the embodiments disclosed in this application, the matched invoices and transaction records can be one-to-one correspondences, wherein the corresponding invoice attribute data and transaction record attribute data can include subject information, transaction time, transaction amount, and transaction remarks. The subject information and transaction amount of the matched invoices and transaction records must be completely identical; the transaction time of the matched invoices and transaction records must be within a reasonable time range; and the transaction remarks of the matched invoices and transaction records must contain the same contract number, order number, or verification code. Based on the aforementioned identical attribute information, the invoice attribute data and transaction record attribute data can be paired using the corresponding similarity score. When the above information matches, the resulting total similarity score can exceed a preset first threshold.

[0067] In the embodiments disclosed in this application, by performing multi-dimensional matching on various attribute data, one-to-one matching invoice attribute information and transaction history attribute information can be filtered out from the range of corresponding matching invoice attribute information and transaction history attribute information. After removing these one-to-one matching invoice attribute information and transaction history attribute information from the first invoice dataset and the first transaction history dataset, such one-to-one matching invoice attribute information and transaction history attribute information will not interfere with subsequent matching steps, improving the accuracy of abnormal data processing and significantly reducing the rate of missed and mismatched matches. Furthermore, compared with traditional manual matching and reconciliation, it can reduce repetitive manual labor and significantly reduce reconciliation time.

[0068] In the embodiments disclosed in this application, according to a preset second matching rule, the invoice attribute data included in the second invoice dataset is matched with the transaction attribute data included in the second transaction dataset to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset, respectively, to obtain an abnormal reconciliation dataset. This includes: matching the current invoice attribute data in the second invoice dataset with multiple transaction attribute data in the second transaction dataset to obtain a transaction reconciliation dataset related to the current invoice; matching the current transaction attribute data in the second transaction dataset with multiple invoice attribute data in the second invoice dataset to obtain an invoice reconciliation dataset related to the current transaction; and obtaining an abnormal reconciliation dataset based on the transaction reconciliation datasets related to multiple invoices and the invoice set reconciliation datasets related to multiple transactions.

[0069] In the embodiments disclosed in this application, a second round of matching is performed. First, multiple transaction records are matched using a single invoice. Then, multiple invoices are matched using a single transaction. A bidirectional traversal method is used to perform secondary matching on the attribute data in the second invoice dataset and the second transaction record dataset. Specifically, the transactions are first traversed forward using invoices: for each current invoice in the second invoice dataset, multiple transactions with a weak correlation to the current invoice can be matched using similarity or other matching methods. The transaction attribute information of these multiple transactions with a weak correlation to the current invoice is used as the transaction reconciliation dataset related to the current invoice. Then, the invoices are traversed backward using transactions: for each current transaction in the second transaction record dataset, multiple invoices with a weak correlation to the current transaction can also be matched using similarity or other matching methods. The invoice attribute information of these multiple transactions with a weak correlation to the current transaction is used as the invoice reconciliation dataset related to the current transaction. After the traversal is completed, the transaction reconciliation datasets related to the multiple invoices and the invoice reconciliation datasets related to the multiple transactions are calculated. By filtering out the aforementioned multiple invoice reconciliation datasets from the second invoice dataset and the aforementioned multiple transaction reconciliation datasets from the second transaction dataset, abnormal reconciliation data can be obtained.

[0070] In the embodiments disclosed in this application, the aforementioned weaker association can be that the subject information is the same but the amounts are different. That is, the above-mentioned traversal processing method can filter out one invoice corresponding to multiple transaction records, and one transaction record corresponding to multiple invoices.

[0071] In the embodiments disclosed in this application, the invoice attribute information contained in the aforementioned abnormal reconciliation dataset may not have corresponding transaction history attribute information, and similarly, the transaction history attribute information may also not have corresponding transaction history attribute information. These isolated attribute information constitute the filtered abnormal reconciliation data. Furthermore, the invoices corresponding to the abnormal reconciliation data may not have received the corresponding transaction history, and similarly, the transaction history corresponding to the abnormal reconciliation data may be that the customer only generated a transaction history but did not issue an invoice.

[0072] In the embodiments disclosed in this application, the secondary matching is bidirectional, and each direction traverses every invoice or transaction record. Therefore, it can automatically identify the reconciliation invoice attribute information and reconciliation transaction record attribute information that failed to match after the first round of matching. The invoices and transactions corresponding to these reconciliation attribute information are not abnormal data, so these reconcilable data should also be filtered out, resulting in the abnormal reconciliation dataset. The abnormal invoice data and abnormal transaction record data contained in the abnormal reconciliation dataset obtained through two rounds of matching are not only more accurate than the abnormal data obtained by ordinary matching rules, but the reconciliation data obtained from it can also facilitate subsequent accounting for rapid reconciliation, accelerating the reconciliation process from multiple perspectives.

[0073] In the embodiments disclosed in this application, the invoice attribute data includes first invoice data, and the transaction history attribute data includes first transaction history data; the current invoice attribute data in the second invoice dataset is matched with multiple transaction history attribute data in the second transaction history dataset to obtain a transaction history reconciliation dataset related to the current invoice; this includes: matching the first invoice data of the current invoice with the first transaction history data of multiple transactions in the second transaction history dataset to obtain a first transaction history reconciliation dataset related to the current invoice; and determining the transaction history reconciliation dataset related to the current invoice based on the current invoice and the first transaction history reconciliation dataset.

[0074] In the embodiments disclosed in this application, the aforementioned first invoice data can be the invoice header, and the first transaction data can be the counterparty account name of the transaction. During matching, the header of the current invoice is compared with the counterparty account names of multiple transactions. This filters out multiple transactions with the same header as the current invoice. The transaction attribute information of these filtered transactions is then used as the first transaction reconciliation dataset related to the current invoice. Then, with the first transaction reconciliation dataset obtained, other attributes can be further filtered to further determine the transactions that can be reconciled and settled with the current invoice. The transaction attribute information of these transactions that can be reconciled and settled with the current invoice is then used as the transaction reconciliation dataset.

[0075] In the embodiments disclosed in this application, the first invoice data of the current invoice is matched with the first transaction data of multiple transactions in the second transaction data set to obtain a first transaction reconciliation dataset related to the current invoice. This includes: matching the first invoice data of the current invoice with the first transaction data of multiple transactions in the second transaction data set to obtain multiple transaction similarities related to the current invoice; determining multiple target transaction similarities with similarities greater than a second threshold from the multiple transaction similarities of the current invoice; and using the transaction attribute data related to the multiple target transaction similarities as the first transaction reconciliation dataset related to the current invoice.

[0076] In the embodiments disclosed in this application, when using the first invoice data to filter out multiple transactions with the same header as the current invoice, the first invoice data of the current invoice is used to calculate the similarity to multiple first transaction data in the second transaction dataset. Then, based on a preset second threshold, the similarity of multiple target transactions is determined. Since one-to-one corresponding invoices and transactions have been filtered out, what needs to be filtered now is the case of multiple transactions corresponding to the current invoice. Therefore, the second threshold is used to compare and obtain the similarity of multiple transactions, thereby obtaining transaction attribute data related to the similarity of multiple transactions, and using this transaction attribute data as the first transaction reconciliation dataset related to the current invoice.

[0077] In the embodiments disclosed in this application, it is necessary to filter out multiple transactions with the same subject information as the current invoice. These filtered transactions do not necessarily need to be reconciled with the current invoice, but the subject information must be the same so that the one-to-many situation can be further filtered in subsequent steps.

[0078] In the embodiments disclosed in this application, the bill attribute data further includes second bill data, and the transaction history attribute data further includes second transaction history data. Based on the current bill and the first transaction history reconciliation dataset, determining the transaction history reconciliation dataset related to the current bill includes: randomly selecting at least two second transaction history data from the multiple transaction history attribute data included in the first transaction history reconciliation dataset to form a transaction history reconciliation subset, thereby obtaining multiple transaction history reconciliation subsets related to the current bill; obtaining subset reconciliation data of the transaction history reconciliation subsets based on the multiple second transaction history data in the transaction history reconciliation subsets; matching the second bill data of the current bill with the subset reconciliation data of each of the multiple transaction history reconciliation subsets to obtain transaction history matching results; and determining the transaction history reconciliation dataset related to the current bill based on the transaction history matching results.

[0079] In the embodiments disclosed in this application, the aforementioned second invoice data can be the invoice's monetary value, and the second transaction data can be the monetary value corresponding to the transaction. Since the first transaction reconciliation dataset obtained above contains transaction attribute data corresponding to multiple transactions with the same subject information as the current invoice but different amounts, further matching of the amounts is required. Specifically, the monetary value data included in the first transaction reconciliation dataset is randomly combined, and the total combined amount is used as the subset reconciliation data. Further, the amount of the current invoice is compared with the multiple subset reconciliation data. For monetary value matching, the amounts must be completely identical to indicate a one-to-many relationship between the current invoice and the corresponding transactions in the matched transaction reconciliation subset. The aforementioned transaction matching result can be the relationship between the amount of the current invoice and the total amount of the multiple transaction reconciliation subsets.

[0080] Figure 3 The illustration shows a schematic diagram of the principle of determining the transaction reconciliation dataset according to an embodiment of the present disclosure.

[0081] The following is for reference. Figure 3 The method for determining the transaction reconciliation dataset disclosed herein will be further explained in conjunction with specific embodiments.

[0082] In the embodiments disclosed in this application, further, when calculating the transaction reconciliation subset and subset reconciliation data, such as Figure 3 As shown, the first transaction reconciliation dataset 310 may include transaction amount a, transaction amount b, and transaction amount c. Randomly constructing a subset of transaction reconciliation data yields: a first subset 320 including transaction amount a and transaction amount b, a second subset 330 including transaction amount b and transaction amount c, and a third subset 340 including transaction amount a, transaction amount b, and transaction amount c. The subset reconciliation data of the first subset 320 is the first transaction amount a+b, the subset reconciliation data of the second subset 330 is the second transaction amount b+c, and the subset reconciliation data of the third subset 340 is the third transaction amount a+b+c. For the current invoice amount d, comparing d with a+b, b+c, and a+b+c yields the transaction matching result 350. Furthermore, if d = b + c, and transaction amount b corresponds to transaction B, and transaction amount c corresponds to transaction C, then the current invoice can be reconciled and offset against transactions B and C. This means that one invoice corresponds to multiple transactions, which will result in transaction reconciliation dataset 360. Transaction reconciliation dataset 360 includes the transaction attribute data corresponding to transactions B and C.

[0083] In the embodiments disclosed in this application, multiple transactions with the same subject information are filtered based on a second threshold. This allows each unmatched transaction to be filtered into candidate transactions. These candidate transactions are then randomly combined, and the reconciliation data of multiple subsets obtained from the random combination are compared with the second transaction data of the current transaction. If a match is found, multiple transactions corresponding to the current transaction can be accurately located. The above method can accurately locate transactions and transactions with complex correspondences, avoiding mismatches and omissions during manual reconciliation.

[0084] In the embodiments disclosed in this application, the invoice attribute data includes first invoice data, and the transaction history attribute data includes first transaction history data; matching the current transaction history attribute data in the second transaction history dataset with multiple invoice attribute data in the second invoice dataset to obtain an invoice reconciliation dataset related to the current transaction history includes: matching the first transaction history data of the current transaction history with the first invoice data of multiple invoices in the second invoice dataset to obtain a first invoice reconciliation dataset related to the current transaction history; and determining the invoice reconciliation dataset related to the current transaction history based on the current transaction history and the first invoice reconciliation dataset.

[0085] In the embodiments disclosed in this application, during matching, the counterparty account name of the current transaction is compared with the headers of multiple invoices. This filters out multiple invoices with the same counterparty account name as the current transaction. The invoice attribute information of the filtered multiple invoices is then used as the first invoice reconciliation dataset related to the current transaction. Then, with the first invoice reconciliation dataset obtained, other attributes can be further filtered to further determine invoices that can be reconciled and verified with the current transaction. The invoice attribute information of these invoices that can be reconciled and verified with the current transaction is used as the transaction reconciliation dataset.

[0086] In the embodiments disclosed in this application, the first transaction data of the current transaction is matched with the first bill data of multiple bills in the second bill dataset to obtain a first bill reconciliation dataset related to the current transaction. This includes: matching the first transaction data of the current transaction with the first bill data of multiple bills in the second bill dataset to obtain the similarity of multiple bills related to the current transaction; determining multiple target bill similarities with similarities greater than a third threshold from the multiple bill similarities of the current transaction; and using the bill attribute data related to the similarity of the multiple target bills as the first bill reconciliation dataset related to the current transaction.

[0087] In the embodiments disclosed in this application, when using the first transaction data to filter out multiple invoices with the same counterparty account name as the current transaction, the first transaction data of the current transaction is used to calculate the similarity with multiple first invoice data in the second invoice dataset. Then, the similarity of multiple target invoices is determined based on a preset third threshold. Since one-to-one corresponding invoices and transactions have been filtered out, what needs to be filtered now is the case of multiple invoices corresponding to the current transaction. Therefore, the second threshold is used to compare and obtain the similarity of multiple invoices, thereby obtaining invoice attribute data related to the similarity of multiple invoices. This invoice attribute data is then used as the first invoice reconciliation dataset related to the current transaction.

[0088] In the embodiments disclosed in this application, it is necessary to filter multiple invoices with the same subject information as the current transaction. These invoices do not necessarily need to be reconciled with the current transaction, but the subject information must be the same so that the one-to-many situation can be further filtered in subsequent steps.

[0089] In the embodiments disclosed in this application, the bill attribute data further includes second bill data, and the transaction history attribute data further includes second transaction history data. Based on the current transaction history and the first bill reconciliation dataset, determining the bill reconciliation dataset related to the current transaction history includes: randomly selecting at least two second bill data from the multiple bill attribute data included in the first bill reconciliation dataset to form a bill reconciliation subset, thereby obtaining multiple bill reconciliation subsets related to the current transaction history; obtaining subset reconciliation data of the bill reconciliation subsets based on the multiple second bill data in the bill reconciliation subsets; matching the second transaction history data of the current transaction history with the subset reconciliation data of each of the multiple bill reconciliation subsets to obtain bill matching results; and determining the bill reconciliation dataset related to the current transaction history based on the bill matching results.

[0090] In the embodiments disclosed in this application, since the first invoice reconciliation dataset obtained above contains invoice attribute data corresponding to multiple invoices with the same subject information as the current transaction but different amounts, further matching of amounts is required. Specifically, the amount data included in the first invoice reconciliation dataset is randomly combined, and the total amount after combination is used as a subset of reconciliation data. Further, the amount of the current transaction is compared with the multiple subsets of reconciliation data. For amount matching, the amounts must be exactly the same to indicate a one-to-many relationship between the current transaction and the corresponding invoices in the matched subsets of invoice reconciliation. The above invoice matching result can be the relationship between the amount of the current transaction and the total amount of the multiple subsets of invoice reconciliation.

[0091] Figure 4 The illustration shows a schematic diagram of the principle of determining the bill reconciliation dataset according to an embodiment of the present disclosure.

[0092] The following is for reference. Figure 4 The method for determining the bill reconciliation dataset disclosed herein will be further explained in conjunction with specific embodiments.

[0093] In the embodiments disclosed in this application, further, when calculating the transaction reconciliation subset and subset reconciliation data, such as Figure 4 As shown, the first invoice transaction reconciliation dataset 410 may include invoice amount e, invoice amount f, and invoice amount g. Randomly constructing subsets of the invoice reconciliation dataset yields: a first subset 420 including invoice amount e and invoice amount f, a second subset 430 including invoice amount f and invoice amount g, and a third subset 440 including invoice amount e, invoice amount f, and invoice amount g. The subset reconciliation data of the first subset 420 is the first invoice amount e+f, the subset reconciliation data of the second subset 430 is the second invoice amount f+g, and the subset reconciliation data of the third subset 440 is the third invoice amount e+f+g. For the current transaction amount h, comparing h with e+f, f+g, and e+f+g yields the invoice matching result 450. Furthermore, if h = f + g, and the bill amount f corresponds to bill F, and the bill amount g corresponds to bill G, then the current transaction record can be reconciled and offset against bills F and G. This means that one transaction record corresponds to multiple bills, which will result in bill reconciliation dataset 460. Bill reconciliation dataset 460 includes the bill attribute data corresponding to bills F and G.

[0094] In the embodiments disclosed in this application, multiple invoices with the same subject information are filtered based on a third threshold in the current transaction history. This ensures that each unmatched transaction can be filtered to find candidate invoices. These candidate invoices are then randomly combined, and the reconciliation data of multiple subsets obtained from the random combination are compared with the second transaction data of the current transaction history. If a match is found, multiple invoices corresponding to the current transaction history can be accurately located. The above method can accurately locate invoices and transactions with complex correspondences, avoiding mismatches and omissions during manual reconciliation.

[0095] In the embodiments disclosed in this application, an abnormal reconciliation dataset is obtained based on a transaction reconciliation dataset associated with multiple invoices and a set of invoices associated with multiple transactions. This includes: filtering out transaction attribute data from the transaction reconciliation datasets corresponding to each of the multiple invoices from the second transaction dataset to obtain an abnormal transaction dataset; filtering out invoice attribute data from the invoice reconciliation datasets corresponding to each of the multiple transactions from the second invoice dataset to obtain an abnormal invoice dataset; and using the abnormal transaction dataset and the abnormal invoice dataset as the abnormal reconciliation dataset.

[0096] In the embodiments disclosed in this application, the matching situation indicated by the transaction reconciliation datasets corresponding to multiple invoices is that one invoice can be reconciled and offset with multiple transactions, and the matching situation indicated by the invoice reconciliation datasets corresponding to multiple transactions is that one transaction can be reconciled and offset with multiple invoices. Therefore, these invoices and transactions that can be reconciled and offset do not belong to abnormal reconciliation data, so they should also be filtered out from the second transaction dataset and the second invoice dataset. The remaining data are attribute information corresponding to transactions that cannot find corresponding invoices and attribute information corresponding to invoices that cannot find corresponding transactions. By treating these remaining attribute information as abnormal reconciliation data, the identification of abnormal reconciliation data can be made more accurate, complex one-to-many or many-to-one reconciliation situations can be filtered out, reducing the error matching of manual or ordinary machine matching and improving the efficiency of reconciliation matching.

[0097] In the embodiments disclosed in this application, the transaction attribute data includes the transaction information subject, and the invoice attribute data includes the invoice information subject. Based on the abnormal reconciliation dataset, a data warning event is generated, including: based on the transaction information subject and the invoice information subject in the abnormal reconciliation dataset; obtaining an information subject set; and generating a data warning event based on the information subjects in the information subject set.

[0098] In the embodiments disclosed in this application, the filtered abnormal reconciliation dataset includes transaction history subjects and invoice subjects. Specific subject information can represent the true identities of the transacting parties at the legal and financial levels, such as the full company name, individual name, and bank account number on the invoice, and the account name and bank account number on the transaction history. Subject information can also represent the role label of the corresponding subject, such as whether the subject is a buyer or a seller. After obtaining the subject information set, early warning events are generated for the corresponding subjects based on the subject information, such as issuing payment reminders to companies that have not made payments and issuing reminders to companies that have not issued invoices.

[0099] In the embodiments disclosed in this application, data early warning events are generated based on the information subject set. This can quickly issue early warnings to subjects with abnormal situations, prevent violations from occurring, and promptly remind subjects to take measures to deal with the issues. This reduces the financial risks, business interruptions, and other adverse consequences that may be caused by data anomalies, and ensures the reliability of corporate financial data and the stability of business operations.

[0100] Figure 5 A schematic block diagram of an abnormal reconciliation data processing apparatus according to an embodiment of this application is shown.

[0101] like Figure 5 As shown, the abnormal reconciliation data processing device 500 of this embodiment includes an acquisition module 510, a first matching module 520, a second matching module 530, and an early warning module 540.

[0102] The acquisition module 510 is used to obtain a first invoice dataset and a first transaction dataset in response to a reconciliation request. The first invoice dataset includes invoice attribute data for each of multiple invoices, and the first transaction dataset includes transaction attribute data for each of multiple transactions. In one embodiment, the acquisition module 510 can be used to perform the operation S210 described above, which will not be repeated here.

[0103] The first matching module 520 is used to match the attribute data of multiple invoices and the attribute data of multiple transaction records according to a preset first matching rule, so as to filter out matching invoice attribute data and transaction attribute data from the first invoice dataset and the first transaction dataset respectively, to obtain a second invoice dataset and a second transaction dataset. In one embodiment, the first matching module 520 can be used to perform the operation S220 described above, which will not be repeated here.

[0104] The second matching module 530 is used to match the invoice attribute data included in the second invoice dataset with the transaction attribute data included in the second transaction dataset according to a preset second matching rule, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, and obtain an abnormal reconciliation dataset. In one embodiment, the second matching module 530 can be used to perform the operation S230 described above, which will not be repeated here.

[0105] The early warning module 540 is used to generate data early warning events based on the abnormal reconciliation dataset. In one embodiment, the early warning module 540 can be used to perform the operation S230 described above, which will not be repeated here.

[0106] In the embodiments disclosed in this application, the first matching module 520 may include a first sub-matching module, a second matching sub-module, and a third matching sub-module.

[0107] The first sub-matching module is used to match the attribute data of multiple transaction records with the attribute data of multiple tickets to obtain multiple first similarity scores.

[0108] The second matching submodule is used to determine at least one first target similarity from a plurality of first similarities that has a similarity greater than a first threshold.

[0109] The third matching submodule is used to filter out ticket attribute data related to at least one first target similarity from the first ticket dataset and to filter out flow attribute data related to at least one first target similarity from the first flow dataset, so as to obtain the second ticket dataset and the second flow dataset.

[0110] In the embodiments disclosed in this application, the second matching module 530 may include a fourth matching submodule, a fifth matching submodule, and a sixth matching submodule.

[0111] The fourth matching submodule is used to match the current invoice attribute data in the second invoice dataset with multiple transaction attribute data in the second transaction dataset to obtain a transaction reconciliation dataset related to the current invoice.

[0112] The fifth matching submodule is used to match the current transaction attribute data in the second transaction dataset with multiple bill attribute data in the second bill dataset to obtain a bill reconciliation dataset related to the current transaction.

[0113] The sixth matching submodule is used to obtain an abnormal reconciliation dataset based on the transaction reconciliation datasets associated with multiple invoices and the invoice reconciliation datasets associated with multiple transactions.

[0114] In the embodiments disclosed in this application, the fourth matching submodule may include a first matching unit and a second matching unit.

[0115] The first matching unit is used to match the first bill data of the current bill with the first transaction data of multiple transactions in the second transaction dataset to obtain the first transaction reconciliation dataset related to the current bill.

[0116] The second matching unit is used to determine the transaction reconciliation dataset related to the current invoice based on the current invoice and the first transaction reconciliation dataset.

[0117] In the embodiments disclosed in this application, the first matching unit may include a first matching subunit, a second matching subunit, and a third matching subunit.

[0118] The first matching subunit is used to match the first ticket data of the current ticket with multiple first transaction data in the second transaction dataset to obtain multiple transaction similarities related to the current ticket.

[0119] The second matching subunit is used to determine multiple target transaction similarities with a similarity greater than a second threshold from multiple transaction similarities of the current ticket.

[0120] The third matching subunit is used to use the flow attribute data related to the similarity of multiple target flows as the first flow reconciliation dataset related to the current invoice.

[0121] In the embodiments disclosed in this application, the second matching unit may include a fourth matching subunit, a fifth matching subunit, a sixth matching subunit, and a seventh matching subunit.

[0122] The fourth matching subunit is used to randomly select at least two second flow data from the multiple flow attribute data included in the first flow reconciliation dataset to form a flow reconciliation subset, so as to obtain multiple flow reconciliation subsets related to the current invoice.

[0123] The fifth matching subunit is used to obtain subset reconciliation data of the transaction reconciliation subset based on multiple second transaction data in the transaction reconciliation subset.

[0124] The sixth matching subunit is used to match the second document data of the current document with the reconciliation data of each of the multiple transaction reconciliation subsets to obtain the transaction matching result.

[0125] The seventh matching subunit is used to determine the transaction reconciliation dataset related to the current invoice based on the transaction matching results.

[0126] In the embodiments disclosed in this application, the fifth matching submodule may include a third matching unit and a fourth matching unit.

[0127] The third matching unit is used to match the first transaction data of the current transaction with the first bill data of multiple bills in the second bill dataset to obtain the first bill reconciliation dataset related to the current transaction.

[0128] The fourth matching unit is used to determine the invoice reconciliation dataset related to the current transaction based on the current transaction history and the first invoice reconciliation dataset.

[0129] In the embodiments disclosed in this application, the third matching unit may include an eighth matching subunit, a ninth matching subunit, and a tenth matching subunit.

[0130] The eighth matching subunit is used to match the first transaction data of the current transaction with multiple first ticket data in the second ticket dataset to obtain the similarity of multiple tickets related to the current transaction.

[0131] The ninth matching subunit is used to determine the similarity of multiple target tickets with a similarity greater than the third threshold from the multiple ticket similarities in the current flow.

[0132] The tenth matching subunit is used to use the bill attribute data related to the similarity of multiple target bills as the first bill reconciliation dataset related to the current transaction.

[0133] In the embodiments disclosed in this application, the fourth matching unit may include an eleventh matching subunit, a twelfth matching subunit, a thirteenth matching subunit, and a fourteenth matching subunit.

[0134] The eleventh matching subunit is used to randomly select at least two second bill data from multiple bill attribute data included in the first bill reconciliation dataset to form a bill reconciliation subset, so as to obtain multiple bill reconciliation subsets related to the current transaction.

[0135] The twelfth matching subunit is used to obtain subset reconciliation data of the bill reconciliation subset based on multiple second bill data in the bill reconciliation subset.

[0136] The thirteenth matching subunit is used to match the second transaction data of the current transaction with the reconciliation data of each of the multiple invoice reconciliation subsets to obtain the invoice matching result.

[0137] The fourteenth matching subunit is used to determine the invoice reconciliation dataset related to the current transaction based on the invoice matching results.

[0138] In the embodiments disclosed in this application, the sixth matching submodule may include a fifth matching unit, a sixth matching unit, and a seventh matching unit.

[0139] The fifth matching unit is used to filter out the transaction attribute data from the transaction reconciliation dataset corresponding to each of the multiple invoices from the second transaction dataset to obtain the transaction anomaly dataset.

[0140] The sixth matching unit is used to filter out the bill attribute data in the bill reconciliation dataset corresponding to each of the multiple transactions from the second bill dataset to obtain the bill anomaly dataset.

[0141] The seventh matching unit is used to combine the abnormal transaction data set and the abnormal invoice data set into an abnormal reconciliation data set.

[0142] In the embodiments disclosed in this application, the early warning module 540 may include a first early warning submodule and a second early warning submodule.

[0143] The first early warning submodule is used to obtain the information subject set based on the transaction information subject and the invoice information subject in the abnormal reconciliation dataset.

[0144] The second early warning submodule is used to generate data early warning events based on the information subjects in the information subject set.

[0145] According to embodiments of this application, any multiple modules among the acquisition module 510, the first matching module 520, the second matching module 530, and the early warning module 540 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of this application, at least one of the acquisition module 510, the first matching module 520, the second matching module 530, and the early warning module 540 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the acquisition module 510, the first matching module 520, the second matching module 530, and the early warning module 540 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0146] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing an abnormal reconciliation data processing method according to an embodiment of this application.

[0147] like Figure 6 As shown, an electronic device 600 according to an embodiment of this application includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage portion 608 into a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0148] RAM 603 stores various programs and data required for the operation of electronic device 600. Processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Processor 601 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 602 and / or RAM 603. It should be noted that the programs may also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0149] According to embodiments of this application, the electronic device 600 may further include an input / output (I / O) interface 605, which is also connected to a bus 604. The electronic device 600 may also include one or more of the following components connected to the input / output (I / O) interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output (I / O) interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 610 as needed so that computer programs read from it can be installed into the storage section 608 as needed.

[0150] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0151] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 602 and / or RAM 603 and / or one or more memories other than ROM 602 and RAM 603 described above.

[0152] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to enable the computer system to implement the abnormal reconciliation data processing method provided in the embodiments of this application.

[0153] When the computer program is executed by the processor 601, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0154] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 609, and / or installed from the removable medium 611. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0155] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 609, and / or installed from the removable medium 611. When the computer program is executed by the processor 601, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0156] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0157] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0158] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for processing abnormal reconciliation data, characterized in that, include: In response to a reconciliation request, a first invoice dataset and a first transaction dataset are obtained. The first invoice dataset includes invoice attribute data for each of multiple invoices, and the first transaction dataset includes transaction attribute data for each of multiple transactions. According to the preset first matching rule, the attribute data of each of the multiple tickets and the attribute data of each of the multiple transactions are matched to filter out the matching ticket attribute data and transaction attribute data from the first ticket dataset and the first transaction dataset respectively, so as to obtain the second ticket dataset and the second transaction dataset. According to the preset second matching rule, the invoice attribute data included in the second invoice dataset is matched with the transaction attribute data included in the second transaction dataset, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, and obtain the abnormal reconciliation dataset. Based on the aforementioned abnormal reconciliation dataset, a data alert event is generated.

2. The method for processing abnormal reconciliation data according to claim 1, characterized in that, The step of matching the attribute data of multiple tickets and the attribute data of multiple transactions according to a preset first matching rule, so as to filter out matching ticket attribute data and transaction attribute data from the first ticket dataset and the first transaction dataset respectively, to obtain a second ticket dataset and a second transaction dataset, includes: The attribute data of each of the multiple transaction records are matched with the attribute data of each of the multiple tickets to obtain multiple first similarity scores; Determine at least one first target similarity from a plurality of first similarities that has a similarity greater than a first threshold; The second bill dataset and the second transaction dataset are obtained by filtering out bill attribute data related to at least one of the first target similarities from the first bill dataset and filtering out transaction attribute data related to at least one of the first target similarities from the first transaction dataset.

3. The method for processing abnormal reconciliation data according to claim 1, characterized in that, The step involves matching the invoice attribute data in the second invoice dataset with the transaction attribute data in the second transaction data dataset according to a preset second matching rule, thereby filtering out mismatched invoices and transactions from the second invoice dataset and the second transaction data dataset to obtain an abnormal reconciliation dataset, including: The current invoice attribute data in the second invoice dataset is matched with multiple transaction attribute data in the second transaction dataset to obtain a transaction reconciliation dataset related to the current invoice. The current transaction attribute data in the second transaction dataset is matched with multiple bill attribute data in the second bill dataset to obtain a bill reconciliation dataset related to the current transaction. An abnormal reconciliation dataset is obtained based on the transaction reconciliation datasets associated with multiple invoices and the invoice reconciliation datasets associated with multiple transactions.

4. The abnormal reconciliation data processing method according to claim 3, characterized in that, The invoice attribute data includes first invoice data, and the transaction history attribute data includes first transaction history data; The step of matching the current invoice attribute data in the second invoice dataset with multiple transaction attribute data in the second transaction dataset to obtain a transaction reconciliation dataset related to the current invoice includes: The first bill data of the current bill is matched with the first transaction data of multiple transactions in the second transaction data set to obtain the first transaction reconciliation dataset related to the current bill. Based on the current invoice and the first transaction reconciliation dataset, determine the transaction reconciliation dataset associated with the current invoice.

5. The method for processing abnormal reconciliation data according to claim 4, characterized in that, The first invoice data of the current invoice is matched with the first transaction data of multiple transactions in the second transaction dataset to obtain a first transaction reconciliation dataset related to the current invoice, including: The first ticket data of the current ticket is matched with multiple first transaction data in the second transaction dataset to obtain multiple transaction similarities related to the current ticket; From the multiple transaction similarities of the current document, determine multiple target transaction similarities whose similarity is greater than a second threshold; The flow attribute data related to the similarity of multiple target flows are used as the first flow reconciliation dataset associated with the current invoice.

6. The method for processing abnormal reconciliation data according to claim 5, characterized in that, The invoice attribute data also includes second invoice data, and the transaction history attribute data also includes second transaction history data; The step of determining the transaction reconciliation dataset related to the current invoice based on the current invoice and the first transaction reconciliation dataset includes: Randomly select at least two second transaction data from the multiple transaction attribute data included in the first transaction reconciliation dataset to form a transaction reconciliation subset, so as to obtain multiple transaction reconciliation subsets related to the current invoice; Based on multiple second flow data in the flow reconciliation subset, a subset of reconciliation data of the flow reconciliation subset is obtained; The second invoice data of the current invoice is matched with the subset reconciliation data of each of the multiple transaction reconciliation subsets to obtain the transaction matching results; Based on the transaction matching results, determine the transaction reconciliation dataset associated with the current invoice.

7. The method for processing abnormal reconciliation data according to claim 3, characterized in that, The invoice attribute data includes first invoice data, and the transaction history attribute data includes first transaction history data; The step of matching the current transaction attribute data in the second transaction dataset with multiple invoice attribute data in the second invoice dataset to obtain an invoice reconciliation dataset related to the current transaction includes: The first transaction data of the current transaction is matched with the first bill data of multiple bills in the second bill dataset to obtain the first bill reconciliation dataset related to the current transaction. Based on the current transaction history and the first invoice reconciliation dataset, determine the invoice reconciliation dataset related to the current transaction history.

8. The method for processing abnormal reconciliation data according to claim 7, characterized in that, The first transaction data of the current transaction is matched with the first document data of multiple documents in the second document dataset to obtain a first document reconciliation dataset related to the current transaction; including: The first flow data of the current flow is matched with multiple first ticket data in the second ticket dataset to obtain the similarity of multiple tickets related to the current flow. From the multiple similarity scores of the current transaction, determine the similarity scores of multiple target tickets that are greater than a third threshold; The bill attribute data related to the similarity of multiple target bills is used as the first bill reconciliation dataset related to the current transaction flow.

9. The method for processing abnormal reconciliation data according to claim 8, characterized in that, The invoice attribute data also includes second invoice data, and the transaction history attribute data also includes second transaction history data; The step of determining the invoice reconciliation dataset related to the current transaction history based on the current transaction history and the first invoice reconciliation dataset includes: Randomly select at least two second bill data from the multiple bill attribute data included in the first bill reconciliation dataset to form a bill reconciliation subset, so as to obtain multiple bill reconciliation subsets related to the current transaction flow; Based on the multiple second bill data in the bill reconciliation subset, a subset of reconciliation data of the bill reconciliation subset is obtained; The second transaction data of the current transaction is matched with the reconciliation data of each of the multiple invoice reconciliation subsets to obtain the invoice matching results; Based on the invoice matching results, determine the invoice reconciliation dataset related to the current transaction history.

10. The method for processing abnormal reconciliation data according to claim 3, characterized in that, The abnormal reconciliation dataset, obtained based on the transaction reconciliation datasets associated with multiple invoices and the invoice set reconciliation datasets associated with multiple transactions, includes: The transaction attribute data in the transaction reconciliation dataset corresponding to each of the multiple invoices is filtered out from the second transaction dataset to obtain the transaction anomaly dataset; The invoice attribute data in the invoice reconciliation dataset corresponding to each of the multiple transactions is filtered out from the second invoice dataset to obtain an invoice anomaly dataset; The abnormal transaction data set and the abnormal invoice data set are used as the abnormal reconciliation data set.

11. The method for processing abnormal reconciliation data according to claim 1, characterized in that, The transaction history attribute data includes the transaction history information subject, and the invoice attribute data includes the invoice information subject. The generation of data alert events based on the abnormal reconciliation dataset includes: Based on the transaction information subjects and invoice information subjects in the aforementioned abnormal reconciliation dataset, obtain the information subject set; Data early warning events are generated based on the information subjects in the information subject set.

12. An abnormal reconciliation data processing device, characterized in that, The device includes: The acquisition module is used to obtain a first invoice dataset and a first transaction dataset in response to a reconciliation request. The first invoice dataset includes invoice attribute data for each of multiple invoices, and the first transaction dataset includes transaction attribute data for each of multiple transactions. The first matching module is used to match the attribute data of each of the multiple bills and the attribute data of each of the multiple transactions according to the preset first matching rules, so as to filter out the matching bill attribute data and transaction attribute data from the first bill dataset and the first transaction dataset respectively, and obtain the second bill dataset and the second transaction dataset. The second matching module is used to match the invoice attribute data included in the second invoice dataset with the transaction attribute data included in the second transaction dataset according to the preset second matching rules, so as to filter out mismatched invoices and transactions from the second invoice dataset and the second transaction dataset respectively, and obtain an abnormal reconciliation dataset. The early warning module is used to generate data early warning events based on the abnormal reconciliation dataset.

13. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 11.

14. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 11.

15. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 11.

Images