Cost settlement method and system based on public-private separation mode, and medium
通过建立费用识别规则库和使用大数据、机器学习算法,解决了现有技术中公共费用与私人费用难以识别的问题,实现了精准的费用结算和审核。
Patent Information
- Application Number
- CN202510368799.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-27
- Publication Date
- 2025-07-11
AI Technical Summary
The existing fee settlement methods are difficult to accurately identify public and private expenses, resulting in settlement difficulties and affecting audit accuracy and financial management efficiency.
By establishing a fee identification rule library, the fee data is classified and analyzed based on big data and machine learning algorithms, public expenses and private expenses are identified, and individual settlement and audits are carried out.
It realizes accurate identification and separate settlement of public and private expenses, improving the accuracy of audit of fee settlement and transparency of financial management.
Smart Images
Figure CN120297978A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of expense settlement. Specifically, it relates to an expense settlement method, system, and medium based on a public-private separation model. Background Art
[0002] The public-private separation model aims to clearly divide public expenses and private expenses, achieve precision and standardization in expense settlement, solve the settlement problems caused by expense confusion, and improve the transparency and efficiency of financial management. Under this model, public expenses refer to the expenditures borne by a collective or organization for public affairs or services, such as enterprise office space rental, water and electricity bills, etc.; private expenses are the expenses incurred by individuals for personal reasons, such as business trips and shopping expenses incurred by employees for personal matters. In existing expense settlement methods, it is difficult to accurately identify public expense data and private expense data from the collected expense data through expense identification rules, so it is impossible to settle public expenses and private expenses separately. During the audit of expense settlement, it is difficult to accurately review, which affects the accuracy of expense reimbursement review. Summary of the Invention
[0003] The purpose of the embodiments of this application is to provide an expense settlement method, system, and medium based on a public-private separation model. By identifying and analyzing the collected expense data through expense identification rules, public expenses and private expenses can be accurately obtained, and then public expenses and private expenses can be settled separately, and the expense settlement can be better audited.
[0004] The embodiments of this application also provide an expense settlement method based on a public-private separation model, including:
[0005] Establish an expense identification rule library based on big data, and formulate detailed expense identification rules based on the expense identification rule library;
[0006] Collect expense data from multiple channels, classify and process the expense data based on the expense identification rules, and obtain expense data of multiple categories;
[0007] Analyze the expense data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public expenses and private expenses;
[0008] Audit based on public expenses to obtain public expense settlement information, and audit based on private expenses to obtain private expense settlement information;
[0009] Transmit the public expense settlement information and the private expense settlement information to the terminal in a set manner.
[0010] Optionally, in the cost settlement method based on the public-private separation mode described in the embodiments of the present application, a cost identification rule library is established based on big data, and detailed cost identification rules are formulated based on the cost identification rule library, specifically including:
[0011] A cost identification rule library is established based on big data, and the cost identification rule library includes cost generation scenarios, payment entities, and purposes.
[0012] Analyze the cost occurrence scenarios and purposes based on the cost generation scenarios to obtain scenario rules, and the cost occurrence scenarios include business entertainment and meetings.
[0013] Analyze the cost payment methods based on the payment entities to obtain payment rules, and the payment methods include bank transfer payment and electronic platform payment.
[0014] Analyze the reasons for cost use based on the purposes to obtain purpose rules.
[0015] Perform fusion processing on the scenario rules, payment rules, and purpose rules to obtain cost identification rules.
[0016] Optionally, in the cost settlement method based on the public-private separation mode described in the embodiments of the present application, cost data is collected from multiple channels, and the cost data is classified based on the cost identification rules to obtain cost data of multiple categories, specifically including:
[0017] Collect multi-channel data from the enterprise financial system, payment platform, and reimbursement management system to obtain cost data.
[0018] Obtain cost payment records and reimbursement application information in real time based on the cost data, and analyze the data formats of the payment records and reimbursement application information.
[0019] Standardize the data in different formats to obtain a unified data structure, and perform cost correlation analysis on the cost data after format processing based on the cost identification rules to obtain a correlation result.
[0020] Analyze the correlation degree based on the correlation result, and compare the correlation degree with multiple set correlation threshold intervals to obtain an analysis result.
[0021] Classify the cost data based on the analysis result to obtain cost data of multiple categories.
[0022] Optionally, in the cost settlement method based on the public-private separation mode described in the embodiments of the present application, analyze the cost data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public costs and private costs, specifically including:
[0023] Select a machine learning classification algorithm, use the expense data labeled by the natural language processing algorithm as training samples, and train a classification model;
[0024] Perform feature processing on the data based on the language processing algorithm to obtain training data;
[0025] Analyze the expense amount, expense occurrence time, and expense description text features based on the training data to obtain the input features of the model;
[0026] Input the input features of the model into the classification model for training to obtain training results;
[0027] Verify the model parameters based on the training results to obtain verification results, and adjust the hyperparameters of the analysis model based on the verification results to obtain an adjusted classification model;
[0028] Analyze the expense data based on the adjusted classification model to obtain public expenses and private expenses.
[0029] Optionally, in the expense settlement method based on the public-private separation mode described in the embodiments of the present application, audit the public expenses to obtain public expense settlement information, and audit the private expenses to obtain private expense settlement information, which specifically includes:
[0030] Obtain public expenses, calculate the public expenses to obtain a first expense calculation result;
[0031] Analyze whether the calculation result is correct based on the first expense calculation result;
[0032] If it is correct, generate public expense settlement information, settle the public expense accounting information to obtain public expense settlement information. If it is incorrect, re-calculate the public expenses twice;
[0033] Obtain private expenses, calculate the private expenses based on the private expenses to obtain a second expense calculation result;
[0034] Analyze whether the calculation result is correct based on the second expense calculation result;
[0035] If it is correct, generate private expense accounting information, audit and settle the private expenses to obtain private expense settlement information; if it is incorrect, re-calculate the private expenses twice.
[0036] Optionally, in the expense settlement method based on the public-private separation mode described in the embodiments of the present application, transmit the public expense settlement information and the private expense settlement information to the terminal in a set manner, which specifically includes:
[0037] Obtain public expense settlement information and private expense settlement information, and perform encryption processing on the public expense settlement information and the private expense settlement information based on an encryption protocol to obtain ciphertext data;
[0038] Perform compression processing on the ciphertext data based on a data compression algorithm to obtain compressed data;
[0039] Analyze the data integrity of the compressed data to obtain missing data;
[0040] Fuse the missing data into the compressed data to obtain complete data, and analyze the transmission condition information of the complete data;
[0041] Match a corresponding transmission channel based on the transmission condition information, and transmit the complete data to the terminal based on the transmission channel.
[0042] In a second aspect, an embodiment of the present application provides an expense settlement system based on a public-private separation mode. The system includes: a memory and a processor. The memory includes a program of an expense settlement method based on a public-private separation mode. When the program of the expense settlement method based on the public-private separation mode is executed by the processor, the following steps are implemented:
[0043] Establish an expense identification rule library based on big data, and formulate detailed expense identification rules based on the expense identification rule library;
[0044] Collect expense data from multiple channels, and classify the expense data based on the expense identification rules to obtain expense data of multiple categories;
[0045] Analyze the expense data of multiple categories based on a language processing algorithm and a machine learning algorithm to obtain public expenses and private expenses;
[0046] Audit based on public expenses to obtain public expense settlement information, and audit based on private expenses to obtain private expense settlement information;
[0047] Transmit the public expense settlement information and the private expense settlement information to the terminal in a set manner.
[0048] Optionally, in the expense settlement system based on a public-private separation mode described in the embodiment of the present application, establishing an expense identification rule library based on big data and formulating detailed expense identification rules based on the expense identification rule library specifically include:
[0049] Establish an expense identification rule library based on big data, where the expense identification rule library includes expense generation scenarios, payment subjects, and purposes;
[0050] Analyze the expense occurrence scenarios and purposes based on the expense generation scenarios to obtain scenario rules, where the expense occurrence scenarios include business entertainment and meetings;
[0051] Analyze the expense payment method based on the payment entity to obtain the payment rule. The payment methods include bank transfer payment and electronic platform payment;
[0052] Analyze the reason for expense usage based on the purpose to obtain the purpose rule;
[0053] Perform a fusion process based on the scenario rule, payment rule, and purpose rule to obtain the expense identification rule.
[0054] Optionally, in the expense settlement system based on the public-private separation mode described in the embodiments of this application, collect expense data through multiple channels, and classify the expense data based on the expense identification rule to obtain expense data of multiple categories, specifically including:
[0055] Collect multi-channel data based on the enterprise financial system, payment platform, and reimbursement management system to obtain expense data;
[0056] Obtain the expense payment record and reimbursement application information in real time based on the expense data, and analyze the data formats of the payment record and reimbursement application information;
[0057] Standardize the data in different formats to obtain a unified data structure, and perform expense correlation analysis on the expense data after format processing based on the expense identification rule to obtain the correlation result;
[0058] Analyze the correlation degree based on the correlation result, and compare the correlation degree with multiple set correlation threshold intervals to obtain the analysis result;
[0059] Classify the expense data based on the analysis result to obtain expense data of multiple categories.
[0060] In a third aspect, the embodiments of this application also provide a computer-readable storage medium, which includes a program for the expense settlement method based on the public-private separation mode. When the program for the expense settlement method based on the public-private separation mode is executed by a processor, the steps of the expense settlement method based on the public-private separation mode as described in any one of the above are implemented.
[0061] As can be seen from the above, a cost settlement method, system and medium based on the public-private separation mode provided by the embodiments of the present application establish a cost identification rule library based on big data, and formulate detailed cost identification rules based on the cost identification rule library; collect cost data through multiple channels, classify and process the cost data based on the cost identification rules to obtain cost data of multiple categories; analyze the cost data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public costs and private costs; audit based on the public costs to obtain public cost settlement information, audit based on the private costs to obtain private cost settlement information; transmit the public cost settlement information and the private cost settlement information to the terminal in a set manner; identify and analyze the collected cost data through the cost identification rules, so as to accurately obtain public costs and private costs, and then separately settle the public costs and private costs, and better audit the cost settlement. BRIEF DESCRIPTION OF THE DRAWINGS
[0062] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative efforts.
[0063] Figure 1 It is a flowchart of the cost settlement method based on the public-private separation mode provided by the embodiments of the present application;
[0064] Figure 2 It is a flowchart of the cost identification rule generation method of the cost settlement method based on the public-private separation mode provided by the embodiments of the present application;
[0065] Figure 3 It is a flowchart of the cost data classification method of the cost settlement method based on the public-private separation mode provided by the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all of them. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application to be protected, but only represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.
[0067] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, terms such as "first" and "second" are only used for distinguishing descriptions and cannot be construed as indicating or implying relative importance.
[0068] Please refer to Figure 1 , Figure 1 which is a flowchart of a cost settlement method based on a public-private separation mode in some embodiments of the present application. The cost settlement method based on the public-private separation mode is used in a terminal device. The cost settlement method based on the public-private separation mode includes the following steps:
[0069] S101, establish a cost identification rule library based on big data, and formulate detailed cost identification rules based on the cost identification rule library;
[0070] S102, collect cost data based on multiple channels, classify the cost data based on the cost identification rules, and obtain cost data of multiple categories;
[0071] S103, analyze the cost data of multiple categories based on a language processing algorithm and a machine learning algorithm to obtain public costs and private costs;
[0072] S104, audit based on the public costs to obtain public cost settlement information, and audit based on the private costs to obtain private cost settlement information;
[0073] S105, transmit the public cost settlement information and the private cost settlement information to the terminal in a set manner.
[0074] It should be noted that by classifying the collected cost data according to the on-demand identification rules, the public costs and private costs are identified, and the public costs and private costs are separately audited to achieve separate settlement of public costs and private costs, providing an effective basis for subsequent cost settlement audits.
[0075] Please refer to Figure 2 , Figure 2 which is a flowchart of a method for generating cost identification rules of a cost settlement method based on a public-private separation mode in some embodiments of the present application. According to an embodiment of the present invention, a cost identification rule library is established based on big data, and detailed cost identification rules are formulated based on the cost identification rule library, specifically including:
[0076] S201, establish a cost identification rule library based on big data, where the cost identification rule library includes cost generation scenarios, payment subjects, and purposes;
[0077] S202. Analyze the expense occurrence scenarios and purposes based on the expense generation scenarios to obtain scenario rules. The expense occurrence scenarios include business entertainment and meetings;
[0078] S203. Analyze the expense payment methods based on the payment entities to obtain payment rules. The payment methods include bank transfer payment and electronic platform payment;
[0079] S204. Analyze the reasons for expense usage based on the purposes to obtain purpose rules;
[0080] S205. Perform a fusion process based on the scenario rules, payment rules, and purpose rules to obtain expense recognition rules.
[0081] It should be noted that corresponding recognition rules are formulated according to different business trip modes obtained through cluster analysis. For short - distance business trips, it is stipulated that transportation expenses mainly consist of local public transportation or online car - hailing, and the reimbursement standard is set according to the local transportation expense level; accommodation expenses are set with a reasonable reimbursement ceiling based on the hotel price range at the business trip destination and in combination with the company's travel standards. For long - distance project - related business trips, the selection range and reimbursement standard of transportation methods (such as airplanes, trains, etc.) are determined according to the travel distance and project cycle, and at the same time, the accommodation and meal subsidy standards at the project location are considered.
[0082] Based on the relationships between office supply procurement expenses discovered by association rule mining and factors such as the number of employees and departmental needs, recognition rules are formulated. For example, according to the number of employees and the business characteristics of the department, the monthly office supply procurement budget for each department is set; for the procurement of special office supplies, detailed procurement applications and usage instructions are required to ensure the rationality and necessity of the procurement.
[0083] Please refer to Figure 3 , Figure 3 is a flowchart of a method for classifying expense data in a cost settlement method based on a public - private separation model in some embodiments of the present application. According to the embodiments of the present invention, expense data is collected through multiple channels, and the expense data is classified and processed based on the expense recognition rules to obtain expense data of multiple categories, specifically including:
[0084] S301. Collect multi - channel data based on the enterprise financial system, payment platform, and reimbursement management system to obtain expense data;
[0085] S302. Obtain the expense payment records and reimbursement application information in real - time based on the expense data, and analyze the data formats of the payment records and reimbursement application information;
[0086] S303. Standardize the data in different formats to obtain a unified data structure, and perform expense correlation analysis on the expense data after format processing based on the expense recognition rules to obtain correlation results;
[0087] S304. Analyze the degree of association based on the association result, compare the degree of association with multiple set association threshold intervals, and obtain the analysis result;
[0088] S305. Classify the expense data based on the analysis result to obtain expense data of multiple categories.
[0089] It should be noted that data from multiple channels such as the enterprise financial system, payment platform, and reimbursement management system are integrated to ensure the comprehensiveness and accuracy of the expense data. Through data interface docking, information such as expense payment records and reimbursement applications is obtained in real time, and data in different formats is standardized to unify the data structure, facilitating subsequent expense classification and settlement.
[0090] According to the embodiments of the present invention, public expenses and private expenses are obtained by analyzing expense data of multiple categories based on language processing algorithms and machine learning algorithms, specifically including:
[0091] Select a machine learning classification algorithm, and use the expense data labeled by the natural language processing algorithm as training samples to train the classification model;
[0092] Perform feature processing on the data based on the language processing algorithm to obtain training data;
[0093] Analyze the expense amount, expense occurrence time, and expense description text features based on the training data to obtain the input features of the model;
[0094] Input the input features of the model into the classification model for training to obtain the training result;
[0095] Verify the model parameters based on the training result to obtain the verification result, and adjust the hyperparameters of the analysis model based on the verification result to obtain the adjusted classification model;
[0096] Analyze the expense data based on the adjusted classification model to obtain public expenses and private expenses.
[0097] It should be noted that the training samples are divided into a training set, a validation set, and a test set, usually in a ratio of 70%, 15%, and 15%. The training set data is used to train the selected machine learning model, and the model parameters are optimized through cross-validation (such as K-fold cross-validation) to improve the model performance. The validation set is used to adjust the model hyperparameters to avoid overfitting. The model performance is evaluated on the test set using metrics such as accuracy, recall rate, and F1 value. The accuracy measures the proportion of correctly classified samples by the model; the recall rate reflects the ability of the model to identify a certain type of sample; the F1 value comprehensively considers the accuracy and recall rate to more comprehensively evaluate the model performance.
[0098] To improve classification accuracy, model fusion techniques can be adopted. For example, the prediction results of logistic regression, decision tree, and support vector machine models are weighted and averaged to construct an ensemble model. The weights are determined according to the performance of each model on the validation set. Models with better performance have higher weights, enabling the ensemble model to integrate the advantages of each model and enhance the expense classification effect.
[0099] According to an embodiment of the present invention, auditing is performed based on public expenses to obtain public expense settlement information, and auditing is performed based on private expenses to obtain private expense settlement information, which specifically includes:
[0100] Obtain public expenses, conduct accounting for the public expenses to obtain a first expense accounting result;
[0101] Analyze whether the accounting result is correct based on the first expense accounting result;
[0102] If it is correct, generate public expense settlement information, and conduct settlement for the public expense accounting information to obtain public expense settlement information. If it is incorrect, re-account the public expenses twice;
[0103] Obtain private expenses, conduct accounting for the private expenses based on the private expenses to obtain a second expense accounting result;
[0104] Analyze whether the accounting result is correct based on the second expense accounting result;
[0105] If it is correct, generate private expense accounting information, conduct audit settlement for the private expenses to obtain private expense settlement information; if it is incorrect, conduct re-accounting for the private expenses.
[0106] It should be noted that after the expenses occur, the business department submits an expense payment application, specifying details such as the purpose of the expense, amount, payee, etc. The application is reviewed by the department head. After confirming the reasonableness and necessity of the expenses, it is submitted to the finance department for review. The finance department reviews the compliance of the expenses, authenticity of the invoices, etc. After passing the review, payment is arranged. The finance department selects a suitable payment method (such as bank transfer, electronic payment) for payment according to the approved application. After the payment is completed, accounting treatment is promptly carried out, the expenses are recorded in the corresponding accounting accounts, and financial statements are generated to enable the management to grasp the public expense expenditure situation. Employees submit private expense reimbursement applications through the reimbursement management system, upload expense vouchers (such as invoices, receipts), and fill in details such as expense details and reimbursement reasons. The system automatically conducts a preliminary review of the expense vouchers, checking the voucher format, content integrity, etc. After passing the review, the application is transferred to the direct superior for manual review to confirm the authenticity and reasonableness of the expenses.
[0107] The approved reimbursement application enters the payment link. After the finance department checks the reimbursement information, the reimbursement amount is paid to the employee's designated account. After the payment is completed, the system updates the reimbursement status, and employees can query the reimbursement progress and results at any time.
[0108] According to an embodiment of the present invention, the public expense settlement information and the private expense settlement information are transmitted to the terminal in a set manner, specifically including:
[0109] Obtain the public expense settlement information and the private expense settlement information, and perform encryption processing on the public expense settlement information and the private expense settlement information based on an encryption protocol to obtain ciphertext data;
[0110] Perform compression processing on the ciphertext data based on a data compression algorithm to obtain compressed data;
[0111] Analyze the data integrity of the compressed data to obtain lost data;
[0112] Fuse the lost data into the compressed data to obtain complete data, and analyze the transmission condition information of the complete data;
[0113] Match a corresponding transmission channel based on the transmission condition information, and transmit the complete data to the terminal based on the transmission channel.
[0114] It should be noted that whether it is transmitted through a network interface or a message queue, the security of the expense settlement information must be ensured. The SSL / TLS encryption protocol is used to encrypt the transmitted data to prevent the data from being stolen or tampered with during the transmission process. In HTTP / HTTPS transmission, the server and the terminal device establish a secure connection through the SSL / TLS handshake protocol to encrypt and decrypt the transmitted expense settlement information. In message queue transmission, the message content is also encrypted, and only the terminal device with the correct key can decrypt and obtain the expense settlement information.
[0115] According to an embodiment of the present invention, it further includes: setting budget quotas for various public expenses according to the enterprise's annual budget plan. Monitor the public expense expenditures every month. When the expenses are close to the budget quota, the system automatically gives an alarm to remind the relevant departments to control the expenditures. For example, the budget for the procurement of office equipment is 50,000 yuan per month. When the expenditure reaches 40,000 yuan, the system sends an alarm to the procurement department.
[0116] In a second aspect, an embodiment of the present application provides an expense settlement system based on a public-private separation mode. The system includes: a memory and a processor. The memory includes a program of an expense settlement method based on a public-private separation mode. When the program of the expense settlement method based on a public-private separation mode is executed by the processor, the following steps are implemented:
[0117] Establish an expense identification rule library based on big data, and formulate detailed expense identification rules based on the expense identification rule library;
[0118] Collect expense data from multiple channels, and classify the expense data based on the expense identification rules to obtain expense data of multiple categories;
[0119] Analyze the cost data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public costs and private costs;
[0120] Conduct audits based on public costs to obtain public cost settlement information, and conduct audits based on private costs to obtain private cost settlement information;
[0121] Transmit the public cost settlement information and the private cost settlement information to the terminal in a set manner.
[0122] It should be noted that by using the recognition rules for the collected cost data to classify the categories, the public costs and private costs are identified, and the public costs and private costs are separately audited to achieve separate settlement of public costs and private costs, providing an effective basis for subsequent cost settlement audits.
[0123] According to the embodiments of the present invention, a cost recognition rule library is established based on big data, and detailed cost recognition rules are formulated based on the cost recognition rule library, specifically including:
[0124] Establish a cost recognition rule library based on big data. The cost recognition rule library includes cost generation scenarios, payment entities, and purposes;
[0125] Analyze the cost occurrence scenarios and purposes based on the cost generation scenarios to obtain scenario rules. The cost occurrence scenarios include business entertainment and meetings;
[0126] Analyze the cost payment methods based on the payment entities to obtain payment rules. The payment methods include bank transfer payments and electronic platform payments;
[0127] Analyze the reasons for cost use based on the purposes to obtain purpose rules;
[0128] Perform fusion processing based on the scenario rules, payment rules, and purpose rules to obtain cost recognition rules.
[0129] It should be noted that corresponding recognition rules are formulated according to different business trip modes obtained through cluster analysis. For short-distance business trips, it is stipulated that transportation costs are mainly local public transportation or online car-hailing, and the reimbursement standard is set according to the local transportation cost level; the accommodation cost is set with a reasonable reimbursement ceiling according to the hotel price range at the business trip destination and in combination with the enterprise's business trip standards. For long-distance project business trips, the selection range and reimbursement standard of transportation methods (such as airplanes, trains, etc.) are determined according to the travel distance and project cycle, and at the same time, the accommodation and catering subsidy standards at the project location are considered.
[0130] Based on the relationships between office supply procurement costs discovered through association rule mining and factors such as the number of employees and departmental requirements, identification rules are formulated. For example, based on the number of employees and the business characteristics of each department, set the monthly office supply procurement budget for each department; for the procurement of special office supplies, require detailed procurement applications and usage instructions to ensure the rationality and necessity of the procurement.
[0131] According to an embodiment of the present invention, based on multi-channel collection of expense data, the expense data is classified and processed based on expense identification rules to obtain expense data of multiple categories, specifically including:
[0132] Perform multi-channel data collection based on the enterprise financial system, payment platform, and reimbursement management system to obtain expense data;
[0133] Based on the expense data, obtain expense payment records and reimbursement application information in real time, and analyze the data formats of the payment records and reimbursement application information;
[0134] Standardize data of different formats to obtain a unified data structure, and perform expense correlation analysis on the expense data after format processing based on expense identification rules to obtain a correlation result;
[0135] Analyze the degree of association based on the correlation result, and compare the degree of association with multiple set association threshold intervals to obtain an analysis result;
[0136] Classify the expense data based on the analysis result to obtain expense data of multiple categories.
[0137] It should be noted that integrating multi-channel data such as the enterprise financial system, payment platform, and reimbursement management system ensures the comprehensiveness and accuracy of expense data. Through data interface docking, obtain information such as expense payment records and reimbursement applications in real time, standardize data of different formats, and unify the data structure to facilitate subsequent expense classification and settlement.
[0138] According to an embodiment of the present invention, based on language processing algorithms and machine learning algorithms, analyze expense data of multiple categories to obtain public expenses and private expenses, specifically including:
[0139] Select a machine learning classification algorithm, and use the expense data labeled by natural language processing algorithms as training samples to train a classification model;
[0140] Perform feature processing on the data based on language processing algorithms to obtain training data;
[0141] Based on the training data, analyze the expense amount, expense occurrence time, and expense description text features to obtain the input features of the model;
[0142] Input the input features of the model into the classification model for training to obtain a training result;
[0143] Verify the model parameters based on the training results to obtain verification results, and adjust the hyperparameters of the analysis model based on the verification results to obtain an adjusted classification model;
[0144] Analyze the expense data based on the adjusted classification model to obtain public expenses and private expenses.
[0145] It should be noted that the training samples are divided into a training set, a validation set, and a test set, usually in a ratio of 70%, 15%, and 15%. Use the training set data to train the selected machine learning model, optimize the model parameters through cross-validation (such as K-fold cross-validation) to improve the model performance. Use the validation set to adjust the model hyperparameters to avoid overfitting. Evaluate the model performance on the test set using metrics such as accuracy, recall rate, and F1 value. The accuracy measures the proportion of correctly classified samples by the model; the recall rate reflects the ability of the model to identify a certain type of sample; the F1 value comprehensively considers the accuracy and recall rate to more comprehensively evaluate the model performance.
[0146] To improve the classification accuracy, a model fusion technique can be adopted. For example, perform a weighted average of the prediction results of the logistic regression, decision tree, and support vector machine models to construct an ensemble model. Determine the weights according to the performance of each model on the validation set. The model with better performance has a higher weight, so that the ensemble model combines the advantages of each model and improves the expense classification effect.
[0147] According to the embodiments of the present invention, conduct an audit based on the public expenses to obtain public expense settlement information, and conduct an audit based on the private expenses to obtain private expense settlement information, specifically including:
[0148] Obtain the public expenses, conduct an accounting of the public expenses to obtain a first expense accounting result;
[0149] Analyze whether the accounting result is correct based on the first expense accounting result;
[0150] If it is correct, generate public expense settlement information, conduct a settlement of the public expense accounting information to obtain public expense settlement information. If it is incorrect, re-conduct a secondary accounting of the public expenses;
[0151] Obtain the private expenses, conduct an accounting of the private expenses based on which to obtain a second expense accounting result;
[0152] Analyze whether the accounting result is correct based on the second expense accounting result;
[0153] If it is correct, generate private expense accounting information, conduct an audit settlement of the private expenses to obtain private expense settlement information; if it is incorrect, conduct a secondary accounting of the private expenses.
[0154] It should be noted that after the expenses occur, the business department submits an expense payment application, specifying details such as the purpose of the expense, the amount, and the payee. After the application is reviewed by the department head and the reasonableness and necessity of the expenses are confirmed, it is submitted to the finance department for re-verification. The finance department reviews the compliance of the expenses, the authenticity of the invoices, etc. After passing the review, it arranges for payment. The finance department selects a suitable payment method (such as bank transfer, electronic payment) for payment according to the approved application. After the payment is completed, the accounting treatment is carried out in a timely manner, the expenses are recorded in the corresponding accounting accounts, and financial statements are generated so that the management can grasp the public expense expenditures. Employees submit private expense reimbursement applications through the reimbursement management system, upload expense vouchers (such as invoices, receipts), and fill in details such as expense details and reimbursement reasons. The system automatically conducts a preliminary review of the expense vouchers, checking the voucher format, content integrity, etc. After passing the review, the application is transferred to the direct superior for manual review to confirm the authenticity and reasonableness of the expenses.
[0155] The approved reimbursement application enters the payment process. After the finance department verifies the reimbursement information, it pays the reimbursement amount to the employee's designated account. After the payment is completed, the system updates the reimbursement status, and employees can query the reimbursement progress and results at any time.
[0156] According to the embodiments of the present invention, the public expense settlement information and the private expense settlement information are transmitted to the terminal in a set manner, specifically including:
[0157] Obtain the public expense settlement information and the private expense settlement information, and perform encryption processing on the public expense settlement information and the private expense settlement information based on an encryption protocol to obtain ciphertext data;
[0158] Perform compression processing on the ciphertext data based on a data compression algorithm to obtain compressed data;
[0159] Analyze the data integrity of the compressed data to obtain lost data;
[0160] Fuse the lost data into the compressed data to obtain complete data, and analyze the transmission condition information of the complete data;
[0161] Match the corresponding transmission channel based on the transmission condition information, and transmit the complete data to the terminal based on the transmission channel.
[0162] It should be noted that whether it is transmitted through a network interface or a message queue, the security of the fee settlement information must be ensured. The SSL / TLS encryption protocol is used to encrypt the transmitted data to prevent the data from being stolen or tampered with during transmission. In HTTP / HTTPS transmission, the server and the terminal device establish a secure connection through the SSL / TLS handshake protocol to encrypt and decrypt the transmitted fee settlement information. In message queue transmission, the message content is also encrypted, and only the terminal device with the correct key can decrypt and obtain the fee settlement information.
[0163] According to an embodiment of the present invention, it further includes: setting budget limits for various public expenses according to the enterprise's annual budget plan. Monitor the public expense expenditures monthly. When the expenses are close to the budget limit, the system automatically issues a warning to remind the relevant departments to control the expenditures. For example, the budget for office equipment procurement expenses is 50,000 yuan per month. When the expenditure reaches 40,000 yuan, the system sends a warning to the procurement department.
[0164] A third aspect of the present invention provides a computer-readable storage medium. The readable storage medium includes a fee settlement method program based on the public-private separation mode. When the fee settlement method program based on the public-private separation mode is executed by a processor, the steps of the fee settlement method based on the public-private separation mode as described in any one of the above are implemented.
[0165] A fee settlement method, system and medium based on the public-private separation mode disclosed by the present invention. By establishing a fee identification rule library based on big data, detailed fee identification rules are formulated based on the fee identification rule library; fee data is collected through multiple channels, and the fee data is classified and processed based on the fee identification rules to obtain fee data of multiple categories; based on language processing algorithms and machine learning algorithms, the fee data of multiple categories is analyzed to obtain public expenses and private expenses; the public expenses are audited to obtain public expense settlement information, and the private expenses are audited to obtain private expense settlement information; the public expense settlement information and the private expense settlement information are transmitted to the terminal in a set manner; the collected fee data is identified and analyzed through the fee identification rules, so as to accurately obtain public expenses and private expenses, and then the public expenses and private expenses are settled separately, and the fee settlement is better audited.
[0166] In several embodiments provided by this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the displayed or discussed components can be through some interfaces. The indirect coupling or communication connection of devices or units can be electrical, mechanical, or other forms.
[0167] The units described above as separate components may or may not be physically separated. The components displayed as units may or may not be physical units; they can be located in one place or distributed to multiple network units; some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0168] In addition, each functional unit in the embodiments of the present invention can be fully integrated into a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated into one unit; the above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.
[0169] Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a readable storage medium. When the program is executed, it executes the steps including the above method embodiments; and the foregoing storage medium includes: removable storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks or optical disks and other various media that can store program codes.
[0170] Or, if the above-mentioned integrated units of the present invention are implemented in the form of software function modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on such an understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. And the foregoing storage medium includes: removable storage devices, ROM, RAM, magnetic disks or optical disks and other various media that can store program codes.
Claims
1. A fee settlement method based on a public-private separation model, characterized in that Including: Establish a cost identification rule base based on big data, and formulate detailed cost identification rules based on the cost identification rule base; Collect cost data from multiple channels, classify and process the cost data based on the cost identification rules to obtain cost data of multiple categories; Analyze the cost data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public costs and private costs; Audit based on public costs to obtain public cost settlement information, and audit based on private costs to obtain private cost settlement information; Transmit the public cost settlement information and the private cost settlement information to the terminal in a set manner.
2. The fee settlement method based on the public-private separation mode according to claim 1, characterized in that, Establish a cost identification rule base based on big data, and formulate detailed cost identification rules based on the cost identification rule base, specifically including: Establish a cost identification rule base based on big data, and the cost identification rule base includes cost generation scenarios, payment entities, and purposes; Analyze the cost occurrence scenarios and purposes based on the cost generation scenarios to obtain scenario rules, and the cost occurrence scenarios include business entertainment and meetings; Analyze the cost payment methods based on the payment entities to obtain payment rules, and the payment methods include bank transfer payments and electronic platform payments; Analyze the reasons for cost use based on the purposes to obtain purpose rules; Perform fusion processing on the scenario rules, payment rules, and purpose rules to obtain cost identification rules.
3. The cost settlement method based on the public-private separation mode according to claim 2, characterized in that Collect cost data from multiple channels, classify and process the cost data based on the cost identification rules to obtain cost data of multiple categories, specifically including: Collect multi-channel data based on the enterprise financial system, payment platform, and reimbursement management system to obtain cost data; Obtain cost payment records and reimbursement application information in real time based on the cost data, and analyze the data formats of the payment records and reimbursement application information; Standardize the data in different formats to obtain a unified data structure, and perform cost correlation analysis on the cost data after format processing based on the cost identification rules to obtain correlation results; Analyze the correlation degree based on the correlation results, compare the correlation degree with multiple set correlation threshold intervals to obtain analysis results; Classify the cost data based on the analysis results to obtain cost data of multiple categories.
4. The fee settlement method based on the public-private separation mode according to claim 3, wherein, Analyze the cost data of multiple categories based on language processing algorithms and machine learning algorithms to obtain public costs and private costs, specifically including: Select a machine learning classification algorithm, and use the cost data labeled by the natural language processing algorithm as training samples to train a classification model; Perform feature processing on the data based on the language processing algorithm to obtain training data; Analyze the cost amount, cost occurrence time, and cost description text features based on the training data to obtain the input features of the model; Input the input features of the model into the classification model for training to obtain training results; Verify the model parameters based on the training results to obtain verification results, and adjust the hyperparameters of the analysis model based on the verification results to obtain an adjusted classification model; Analyze the cost data based on the adjusted classification model to obtain public costs and private costs.
5. The fee settlement method based on the public-private separation mode according to claim 4, wherein Audit based on public costs to obtain public cost settlement information, and audit based on private costs to obtain private cost settlement information, specifically including: Obtain public expenses, conduct accounting on the public expenses, and obtain the first expense accounting result; Analyze whether the accounting result is correct based on the first expense accounting result; If it is correct, generate public expense settlement information, settle the public expense accounting information, and obtain the public expense settlement information. If it is incorrect, re-account the public expenses; Obtain private expenses, conduct accounting on the private expenses, and obtain the second expense accounting result; Analyze whether the accounting result is correct based on the second expense accounting result; If it is correct, generate private expense accounting information, audit and settle the private expenses, and obtain the private expense settlement information. If it is incorrect, conduct secondary accounting on the private expenses; 6. The cost settlement method based on the public-private separation mode according to claim 5, wherein Transmit the public expense settlement information and the private expense settlement information to the terminal in a set manner, specifically including: Obtain the public expense settlement information and the private expense settlement information, and perform encryption processing on the public expense settlement information and the private expense settlement information based on the encryption protocol to obtain ciphertext data; Perform compression processing on the ciphertext data based on the data compression algorithm to obtain compressed data; Analyze the data integrity of the compressed data to obtain lost data; Fuse the lost data into the compressed data to obtain complete data, and analyze the transmission condition information of the complete data; Match the corresponding transmission channel based on the transmission condition information, and transmit the complete data to the terminal based on the transmission channel.
7. A fee settlement system based on a public-private separation model, characterized in that, The system includes: a memory and a processor. The memory includes a program for the expense settlement method based on the public-private separation mode. When the program for the expense settlement method based on the public-private separation mode is executed by the processor, the following steps are implemented: Establish an expense identification rule library based on big data, and formulate detailed expense identification rules based on the expense identification rule library; Collect expense data from multiple channels, and classify the expense data based on the expense identification rules to obtain expense data of multiple categories; Analyze the expense data of multiple categories based on the language processing algorithm and the machine learning algorithm to obtain public expenses and private expenses; Conduct audits based on the public expenses to obtain public expense settlement information, and conduct audits based on the private expenses to obtain private expense settlement information; Transmit the public expense settlement information and the private expense settlement information to the terminal in a set manner.
8. The fee settlement system based on the public-private separation mode according to claim 7, wherein Establish an expense identification rule library based on big data, and formulate detailed expense identification rules based on the expense identification rule library, specifically including: Establish an expense identification rule library based on big data. The expense identification rule library includes expense generation scenarios, payment entities, and purposes; Analyze the expense occurrence scenarios and purposes based on the expense generation scenarios to obtain scenario rules. The expense occurrence scenarios include business entertainment and meetings; Analyze the expense payment methods based on the payment entities to obtain payment rules. The payment methods include bank transfer payment and electronic platform payment; Analyze the reasons for expense use based on the purposes to obtain purpose rules; Perform fusion processing on the scenario rules, payment rules, and purpose rules to obtain expense identification rules.
9. The cost settlement system based on the public-private separation mode according to claim 8, wherein Collect expense data from multiple channels, and classify the expense data based on the expense identification rules to obtain expense data of multiple categories, specifically including: Multi-channel data collection is carried out based on the enterprise financial system, payment platform and reimbursement management system to obtain expense data; Based on the expense data, expense payment records and reimbursement application information are obtained in real time, and the data formats of the payment records and reimbursement application information are analyzed; Standardize the data in different formats to obtain a unified data structure, and perform expense correlation analysis on the expense data after format processing based on the expense recognition rules to obtain the correlation result; Analyze the correlation degree based on the correlation result, and compare the correlation degree with multiple set correlation threshold intervals to obtain the analysis result; Classify the expense data based on the analysis result to obtain expense data in multiple categories.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a program for the expense settlement method based on the public-private separation mode. When the program for the expense settlement method based on the public-private separation mode is executed by a processor, the steps of the expense settlement method based on the public-private separation mode as described in any one of claims 1 to 6 are implemented.