CN-122023035-A - Account data processing method, device, computer equipment and storage medium

CN122023035ACN 122023035 ACN122023035 ACN 122023035ACN-122023035-A

Abstract

The embodiment of the application discloses an account data processing method, an account data processing device, computer equipment and a storage medium. The method comprises the steps of obtaining accounting data, processing the accounting data into data nodes in a standardized format, processing the data nodes into combined clusters and isolated nodes through a clustering algorithm, carrying out balancing processing on the combined clusters and the isolated nodes to obtain associated accounting groups and independent accounts, regarding the combined clusters and the isolated nodes as independent individuals, carrying out balancing processing on the combined clusters and the isolated nodes to obtain the associated accounting groups and the independent accounts, and generating and outputting a reconciliation report according to the associated accounting groups and the independent accounts. Therefore, the application can combine and cluster the accounting data with similar accounting characteristics, reduce the calculation complexity and then carry out balancing processing, and finally generate the accounting report covering the associated accounting group and the isolated accounts, thereby facilitating the rapid positioning of the abnormality of the user, completing the checking and improving the accuracy and the automation level of the financial processing.

Inventors

ZHU JUN
ZHANG FENGXIA
Liu Quliang
WU DI
HU SHALI
LIU YAN
ZHANG DONG
LIU CHENGZHEN
YAO BINHUI
WANG ZENG

Assignees

浙江吉利控股集团有限公司

Dates

Publication Date: 20260512
Application Date: 20251209

Claims (10)

1. The accounting data processing method is characterized by comprising the following steps: Acquiring accounting data, and processing the accounting data into data nodes in a standardized format; Processing the data nodes into a merged cluster and an isolated node through a clustering algorithm, wherein the merged cluster comprises a plurality of data nodes, and the isolated node corresponds to one data node; The method comprises the steps of regarding the combined clusters and the isolated nodes as independent individuals, carrying out balancing treatment on the combined clusters and the isolated nodes to obtain an associated accounting group and independent accounts, wherein the associated accounting group is composed of a plurality of the combined clusters and/or the isolated nodes, and the combined clusters and/or the isolated nodes in the associated accounting group have association relations; And generating and outputting a reconciliation report according to the associated accounting group and the independent accounts.
2. The method of processing accounting data as claimed in claim 1, wherein said processing said accounting data into a standardized format data node comprises: extracting key features from the accounting data through a large language model and/or regular expression; generating a structured field according to the key characteristics and a preset rule; and mapping the structured fields into data nodes with unified semantics.
3. The method for processing account data according to claim 1, wherein the processing the data node into the merged cluster and the orphaned node by a preset clustering algorithm comprises: According to a preset similarity calculation rule, aiming at the similarity of each data node and all the rest data nodes; Acquiring a preset similarity threshold; If the similarity among the plurality of data nodes is greater than or equal to the similarity threshold, dividing the plurality of data nodes into the same merging cluster, wherein the merging cluster comprises the plurality of data nodes; And if the similarity of the data nodes and all the rest data nodes is smaller than the similarity threshold value, marking the data nodes as the isolated nodes.
4. The accounting data processing method as set forth in claim 3, wherein said calculating a rule according to a preset similarity includes, for each of said data nodes, remaining similarities of all of said data nodes: acquiring a feature vector of each data node; inputting the feature vector into a similarity calculation model which is built and trained in advance; and calculating the similarity between every two data nodes according to the similarity calculation model.
5. The method for processing accounting data as claimed in claim 1, wherein said balancing said merged clusters and said isolated nodes to obtain associated accounting groups and independent accounts comprises: the merging clusters and the isolated nodes are respectively regarded as mutually independent nodes to be processed; determining the account amount of each node to be processed according to the account data corresponding to the node to be processed; Traversing all the nodes to be processed through a preset algorithm to determine zero and a subset; Summarizing the data nodes included in the zero and subset into the associated financial group; Marking the remaining data nodes except the associated accounting group as independent accounts.
6. The method of claim 5, wherein traversing all of the nodes to be processed by a predetermined algorithm to determine zeros and subsets comprises: setting the subset scale as a preset scale lower limit, and arbitrarily combining the nodes to be processed according to the subset scale to obtain a plurality of pre-group subsets; when the sum of the account amounts of all the nodes to be processed in the pre-group subset is zero, marking the pre-group subset as the zero sum subset; removing the nodes to be processed corresponding to the data nodes included in the zero and the subset; And gradually increasing the scale of the subset, and combining the rest nodes to be processed to find out all zeros and subsets until the scale of the subset is equal to a preset upper scale limit.
7. The accounting data processing method of claim 1, wherein said generating and outputting a reconciliation report from said associated accounting group and said independent accounting comprises: combining the accounting data corresponding to all the data nodes in one associated accounting group into a first accounting; Determining accounting data corresponding to the independent accounts and marking the accounting data as second accounts; and generating the reconciliation report according to the first accounting and the second accounting, and outputting the reconciliation report.
8. An account data processing apparatus, comprising: the acquisition module is used for acquiring the accounting data and processing the accounting data into data nodes in a standardized format; The clustering module is used for processing the data nodes into a merged cluster and an isolated node through a clustering algorithm, wherein the merged cluster comprises a plurality of data nodes, and the isolated node corresponds to one data node; The balancing module is used for regarding the combined clusters and the isolated nodes as independent individuals and balancing the combined clusters and the isolated nodes to obtain an associated accounting group and independent accounts, wherein the associated accounting group is composed of a plurality of the combined clusters and/or the isolated nodes, and the combined clusters and/or the isolated nodes in the associated accounting group have association relations; And the output module is used for generating and outputting a reconciliation report according to the associated accounting group and the independent accounts.
9. A computer device comprising a processor and a memory; The processor is configured to execute a computer program stored in the memory to implement the method of any one of claims 1 to 7.
10. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program which, when executed by a processor, implements the method according to any of claims 1 to 7.

Description

Account data processing method, device, computer equipment and storage medium Technical Field The present application relates to the technical field of accounting data, and in particular, to an accounting data processing method, an accounting data processing device, a computer device, and a computer readable storage medium. Background Financial related transaction reconciliation refers to the process of reconciling and validating economic transactions (purchases, sales, borrowings, vouchers, services, charge allocations, etc.) that have occurred between related enterprises (e.g., parent and subsidiary, different subsidiary within the same group, related investors) on financial accounts. I.e. checking whether the transaction amount, direction and time recorded on each other are consistent or not, otherwise, the group merger report is "on not". Manual reconciliation has been a "nightmare" for financial staff. The existing reconciliation has the problems of multiple parties, such as multiple numbers of associated parties, frequent transactions (such as fund allocation, commodity purchase and sale, service provision and the like among parent-subsidiary companies and subsidiary companies in a group), manual check of Excel forms per month, communication difference through mails/telephones, time and labor consumption and easy error. The related transaction processing flows and the confirmation standards of different subsidiaries and different service lines are different, so that the data caliber is disordered, and the aggregation and management are difficult. Similarly, most enterprises still have a large amount of data after systemizing the account checking capability of the related transactions, and thus the standardization and the unification cannot be achieved. The account checking efficiency is low, the automation degree is insufficient, abnormal differences are difficult to locate and process in time, and the accuracy and timeliness of the financial report are affected. How to increase the rate of accounting for a large amount of accounting data is a technical problem to be solved by those skilled in the art. The foregoing description is provided for general background information and does not necessarily constitute prior art. Disclosure of Invention Based on this, it is necessary to address the above-described problems, and an accounting data processing method, an accounting data processing device, a computer device, and a computer-readable storage medium are proposed that can efficiently find accounts that match each other from a large amount of accounting data. The application solves the technical problems by adopting the following technical scheme: The application provides an accounting data processing method, which specifically comprises the steps of obtaining accounting data, processing the accounting data into data nodes in a standardized format, processing the data nodes into merged clusters and isolated nodes through a clustering algorithm, wherein the merged clusters comprise a plurality of data nodes corresponding to one data node, the merged clusters and the isolated nodes are regarded as independent individuals, balancing processing is carried out on the merged clusters and the isolated nodes to obtain an associated accounting group and independent accounts, the associated accounting group is formed by a plurality of merged clusters and/or isolated nodes, the merged clusters and/or the isolated nodes in the associated accounting group have an association relationship, the independent accounts are formed by one merged cluster or isolated node, the independent accounts do not have an association relationship with other associated accounting groups or independent accounts, and a reconciliation report is generated and output according to the associated accounting group and the independent accounts. In an alternative embodiment of the application, the processing of the accounting data into the data nodes in the standardized format comprises the steps of extracting key features from the accounting data through a large language model and/or a regular expression, generating structured fields according to preset rules according to the key features, and mapping the structured fields into the data nodes with unified semantics. In an optional embodiment of the application, the data nodes are processed into merged clusters and isolated nodes through a preset clustering algorithm, and the method comprises the steps of aiming at the similarity between each data node and all other data nodes according to a preset similarity calculation rule, obtaining a preset similarity threshold value, dividing the data nodes into the same merged clusters if the similarity between the data nodes is greater than or equal to the similarity threshold value, wherein the merged clusters comprise the data nodes, and marking the data nodes as the isolated nodes if the similarity between the data nodes and all other data nodes is les