CN-122021649-A - Resource data processing method, apparatus, computer device, computer readable storage medium and computer program product

CN122021649ACN 122021649 ACN122021649 ACN 122021649ACN-122021649-A

Abstract

The present application relates to a resource data processing method, apparatus, computer device, computer readable storage medium and computer program product. The method comprises the steps of determining a reference service system corresponding to a target service system, obtaining reference subject table data of the reference service system, generating an initial subject table corresponding to the target service system according to a clustering analysis result of the reference subject table data and a preset resource data recording rule, responding to a to-be-recorded service document of the target service system, carrying out natural language analysis on the to-be-recorded service document to obtain service characteristic information of the to-be-recorded service document, determining a target subject corresponding to the to-be-recorded service document from the plurality of selectable subjects according to the service characteristic information through a pre-trained recommendation model, and carrying out resource information recording on the to-be-recorded service document according to the target subject. The application can improve the efficiency and accuracy of resource data processing.

Inventors

LI YINGCHEN
DAI YUWEI
CHENG MIN
LUO CHANGSHENG
LIU JINSHAN

Assignees

金蝶软件（中国）有限公司

Dates

Publication Date: 20260512
Application Date: 20260206

Claims (10)

1. A method of resource data processing, the method comprising: Determining a reference service system corresponding to a target service system, and acquiring reference subject table data of the reference service system; Generating an initial subject table corresponding to the target service system according to the clustering analysis result of the reference subject table data and a preset resource data recording rule, wherein the initial subject table comprises a plurality of selectable subjects; responding to a business document to be recorded of the target business system, and carrying out natural language analysis on the business document to be recorded to obtain business characteristic information of the business document to be recorded; determining a target subject corresponding to the business document to be input from the plurality of selectable subjects according to the business feature information through a pre-trained recommendation model; and inputting the resource information of the business document to be input according to the target subjects.
2. The method of claim 1, wherein the generating the initial subject table corresponding to the target service system according to the result of the cluster analysis of the reference subject table data and a preset resource data recording rule includes: removing invalid subject items from the reference subject table data to obtain cleaned data; performing cluster analysis on the cleaned data to obtain a common subject structure of the reference subject table data; Determining a target resource data recording rule corresponding to the target service system from a plurality of resource data recording rules according to the portrait information of the target service system; Correcting the common subject structure according to the target resource data recording rule to obtain a subject hierarchical structure corresponding to the target service system and subject coding and naming rules; And generating the initial subject table according to the subject hierarchical structure and subject coding and naming rules.
3. The method according to claim 1, wherein the responding to the business document to be entered of the target business system performs natural language analysis on the business document to be entered to obtain the business feature information of the business document to be entered, and includes: identifying the business document to be recorded through a preset natural language model to obtain a business entity and a business action corresponding to the business document to be recorded; and mapping the business entity and the business action to corresponding accounting element categories based on a preset domain knowledge mapping rule, and taking the accounting element categories as the business characteristic information.
4. The method according to claim 1, wherein the determining, by the pre-trained recommendation model, the target subject corresponding to the business document to be entered from the plurality of selectable subjects according to the business feature information includes: carrying out reasoning calculation based on semantic association degrees between the business feature information and the plurality of elective course subjects through the recommendation model to obtain at least one candidate subject corresponding to the business document to be input and the confidence degree of the at least one candidate subject; Checking subjects with confidence degrees meeting a preset threshold value in the candidate subjects according to the target resource data recording rule; and obtaining the candidate subjects passing the verification as the target subjects, wherein at least one of correction, replacement or marking processing is carried out on the subjects which do not accord with the resource data recording rule.
5. The method of claim 4, wherein after verifying that the confidence level of one or more of the candidate subjects meets a preset threshold according to the target resource data recording rule, the method further comprises: if the recommendation model does not output candidate subjects with confidence degree meeting a preset threshold value and/or all candidate subjects do not pass verification, searching parent subject nodes with the correlation degree meeting the preset threshold value with the service characteristic information in the initial subject table according to the accounting element category of the service document to be input; generating a new subject code associated with the parent subject node according to the subject code and naming rule corresponding to the initial subject table; And generating a recommended name of the new subject code based on the service characteristic information, and adding a new subject item containing the new subject code and the recommended name to the initial subject table as the target subject.
6. The method according to any one of claims 1-5, wherein before the determining, by the pre-trained recommendation model, a target subject corresponding to the business document to be entered from the plurality of selectable subjects according to the business feature information, the method comprises: According to the historical service bill collection and the verification subjects corresponding to each historical service bill, a training sample is constructed; carrying out the natural language analysis on the historical service bill to obtain historical service characteristic information; taking the historical service characteristic information as input, taking the corresponding subjects as training targets, and performing supervision training on a machine learning model to obtain the recommendation model; recording the service characteristic information of the service document to be recorded, candidate subjects output by the recommendation model and actual subjects aiming at the document; and if the actual subjects are inconsistent with the candidate subjects output by the recommendation model, performing incremental training on the recommendation model by taking the actual subjects as incremental training samples.
7. A resource data processing apparatus, the apparatus comprising: The first determining module is used for determining a reference service system corresponding to the target service system and acquiring reference subject table data of the reference service system; The generation module is used for generating an initial subject table corresponding to the target service system according to the clustering analysis result of the reference subject table data and a preset resource data recording rule, wherein the initial subject table comprises a plurality of selectable subjects; The analysis module is used for responding to the business document to be recorded of the target business system, and carrying out natural language analysis on the business document to be recorded to obtain business characteristic information of the business document to be recorded; The second determining module is used for determining a target subject corresponding to the business document to be input from the plurality of selectable subjects according to the business characteristic information through a pre-trained recommending model; and the recording module is used for recording the resource information of the business document to be recorded according to the target subjects.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

Description

Resource data processing method, apparatus, computer device, computer readable storage medium and computer program product Technical Field The present application relates to the field of computer data processing technology, and in particular, to a resource data processing method, apparatus, computer device, computer readable storage medium and computer program product. Background With the penetration of enterprise digital transformation and the wide application of enterprise resource planning systems, the automation and intelligence requirements for enterprise resource data processing are becoming increasingly stringent. In this context, financial data processing techniques based on rules engines and automation scripts are emerging. The technology performs automatic accounting processing on the structured business bill through a preset business logic and accounting subject mapping rule, and aims to reduce manual operation and improve accounting efficiency. In the related technology, in the traditional resource data processing mode, a set of recording tables of resource data are manually designed and configured by experts according to industry experience and enterprise scale, so that a framework is provided for subsequent resource accounting. In daily operation, when business receipts are generated, a system or financial staff relies on preset hard coding rules based on keywords or receipt types to match the receipts to specific subjects in a subject table, and then generation of accounting vouchers is completed. The whole process is highly dependent on the completeness and accuracy of expert experience and manual configuration, and the coverage degree of a rule base on a service scene, so that the problems that the process is time-consuming and missing or non-compliant setting is easy to occur easily, and the matching accuracy is obviously reduced when complex, fuzzy or brand-new service description is faced by automatic accounting based on fixed rules, and frequent manual intervention is often needed for correction. Disclosure of Invention In view of the foregoing, it is desirable to provide a resource data processing method, apparatus, computer device, computer readable storage medium, and computer program product that can improve processing efficiency and accuracy. In a first aspect, the present application provides a resource data processing method, the method including: Determining a reference service system corresponding to a target service system, and acquiring reference subject table data of the reference service system; Generating an initial subject table corresponding to the target service system according to the clustering analysis result of the reference subject table data and a preset resource data recording rule, wherein the initial subject table comprises a plurality of selectable subjects; responding to a business document to be recorded of the target business system, and carrying out natural language analysis on the business document to be recorded to obtain business characteristic information of the business document to be recorded; determining a target subject corresponding to the business document to be input from the plurality of selectable subjects according to the business feature information through a pre-trained recommendation model; and inputting the resource information of the business document to be input according to the target subjects. In a second aspect, the present application also provides a resource data processing apparatus, the apparatus comprising: The first determining module is used for determining a reference service system corresponding to the target service system and acquiring reference subject table data of the reference service system; The generation module is used for generating an initial subject table corresponding to the target service system according to the clustering analysis result of the reference subject table data and a preset resource data recording rule, wherein the initial subject table comprises a plurality of selectable subjects; The analysis module is used for responding to the business document to be recorded of the target business system, and carrying out natural language analysis on the business document to be recorded to obtain business characteristic information of the business document to be recorded; The second determining module is used for determining a target subject corresponding to the business document to be input from the plurality of selectable subjects according to the business characteristic information through a pre-trained recommending model; and the recording module is used for recording the resource information of the business document to be recorded according to the target subjects. In a third aspect, the present application also provides a computer device comprising a memory storing a computer program and a processor implementing the steps comprised by any of the method embodiments described above when the computer program