CN117455699A - Automatic identification billing algorithm model based on machine learning - Google Patents
Automatic identification billing algorithm model based on machine learning Download PDFInfo
- Publication number
- CN117455699A CN117455699A CN202311453325.0A CN202311453325A CN117455699A CN 117455699 A CN117455699 A CN 117455699A CN 202311453325 A CN202311453325 A CN 202311453325A CN 117455699 A CN117455699 A CN 117455699A
- Authority
- CN
- China
- Prior art keywords
- billing
- bill
- automatic identification
- model
- identification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 33
- 238000010801 machine learning Methods 0.000 title claims abstract description 33
- 230000006870 function Effects 0.000 claims description 38
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000000034 method Methods 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 9
- 238000012795 verification Methods 0.000 claims description 9
- 230000003993 interaction Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000004519 manufacturing process Methods 0.000 claims description 4
- 238000003066 decision tree Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000013480 data collection Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000446 fuel Substances 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/12—Accounting
- G06Q40/125—Finance or payroll
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Finance (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Computational Linguistics (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Databases & Information Systems (AREA)
- Technology Law (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of automatic identification and accounting, and discloses an automatic identification and accounting algorithm model based on machine learning.
Description
Technical Field
The invention relates to the technical field of automatic identification and billing, in particular to an automatic identification and billing algorithm model based on machine learning.
Background
With the development of scientific technology, the identification billing converts text or image information provided by a user into structured billing data, and the text or the provided bill pictures input by the user are analyzed and identified through text analysis, pattern matching, computer vision and image processing technologies, so that the workload of manual billing is reduced, the billing efficiency is improved, and human errors and omission are reduced.
The application publication number CN111428599A discloses a bill identification method, a bill identification device and bill identification equipment, and the bill identification method in the scheme comprises the following steps: acquiring an image of a bill to be identified; inputting the image of the bill to be identified into a pre-trained bill classification model to obtain the type of the bill to be identified; acquiring a corresponding relation between the type of at least one bill stored in advance and a corresponding bill model; searching a bill model corresponding to the type of the bill to be identified from the corresponding relation; identifying the image of the bill to be identified based on the found bill model to obtain bill information of the bill to be identified; the method avoids possible billing errors caused by manual billing, saves the input time and improves the accuracy and the working efficiency.
The application publication number CN115578192a discloses a refund billing method, a processor, a computer device and a machine-readable storage medium, the method comprising: acquiring fund information of a client remittance; determining a target client to which the funds belong according to the fund information; obtaining a repayment model of a target customer; determining an order repayment plan, order overdue data and order repayment priority of a target customer according to the repayment model; distributing and confirming remittance funds of the target clients according to the order repayment plan, the order overdue data and the order repayment priority, and generating a repayment decomposition list; generating a financial accounting voucher according to the refund decomposition list to finish refund accounting; the method for billing the refund is systematic, online, automatic and transparent; the fund can automatically identify the affiliated clients, so that the fund recognition efficiency is improved; financial accounting is carried out automatically, so that the working efficiency is improved; and the fund use and the optimal allocation detail are automatically determined through the repayment model, so that manual checking operation is reduced, and the working quality is improved.
The problems presented in the background art exist in the above patents: only the diversity of bill identification is researched, the data dependency in the identification process is not considered, and the constructed various model decision processes are difficult to explain, so that the trust degree of users on the results is reduced. In order to solve the problem, the invention provides an automatic identification billing algorithm model based on machine learning.
Disclosure of Invention
This section is intended to outline some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section as well as in the description summary and in the title of the application, to avoid obscuring the purpose of this section, the description summary and the title of the invention, which should not be used to limit the scope of the invention.
The present invention has been made in view of the above-mentioned problems with the prior art implementation of automatic recognition billing algorithm model based on machine learning.
It is therefore an object of the present invention to provide a machine learning based implementation of an automatic identification billing algorithm model.
In order to solve the technical problems, the invention provides the following technical scheme: collecting and collating billing data;
extracting features from the billing data;
stage classification is carried out on the characteristics, and an automatic identification billing model is constructed;
evaluating the automatically identified billing model;
perfecting the automatic identification billing model;
the automatic identification billing model is monitored periodically.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: the billing process interaction flow of the billing data is as follows:
uploading the bills singly or in batches, and carrying out compliance examination on the bills;
if the bill is not compliant, returning the bill;
if the bill is compliant, carrying out authenticity verification;
if the bill information is not complete, prompting filling information, and carrying out authenticity verification after compliance;
if the bill is not true, returning the bill;
if the bill is true, checking the bill again;
if the bill is repeated, returning the bill;
if the bill is not repeated, comparing the bill with a bank statement;
if the bank statement has a balance record, carrying out grading calculation according to the record of the statement;
if the bank has no balance record on the bill, the bank automatically recognizes the bill to generate the balance record, and then carries out grading calculation according to the record of the bill.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: and extracting different features according to the data name, wherein the extraction rule is as follows:
if the data name is a bank transaction class, extracting the data name as a bank statement;
if the data name is a value-added tax invoice class, extracting the data name as a value-added tax invoice;
and if the data name is other special bill types, extracting the data name as other bill types.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: the rules of the phase classification are as follows:
if the characteristic is a value-added tax invoice, direction one identification, bill name identification and semantic one identification are carried out;
if the characteristics are other types of notes, direction one recognition and note name recognition are carried out;
wherein the direction one identification comprises an entry bill and a pin bill, and the semantic one identification comprises a name and a remark;
if the direction I is identified as an entry bill, department identification and enterprise type identification are carried out;
if the direction I is identified as the sales item bill, carrying out enterprise type identification;
the department identification comprises a production department, a technical department, a management department and a business department, and the enterprise type identification comprises a general tax payer and a small-scale tax payer;
if the characteristic is a bank statement, performing direction two recognition and semantic two recognition;
the second direction identification comprises a borrower and a lender, and the second semantic identification comprises a purpose, a abstract and a remark.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: according to the stage classification, the constructed automatic identification billing model is as follows:
wherein KA represents an automatic recognition billing model function, c represents a constant, h n A fitting function representing the phase classification, λ representing the optimal coefficients of the fitting function, N representing the billing data, N representing the total number of billing data;
wherein, the function expression of the fitting function is as follows:
h n ＝min[min∑(n-a 1 ) 2 +min∑(n-a 2 ) 2 +min∑(n-a 3 ) 2 ]；
in the formula, h n Fitting functions representing the phase classifications, n representing the billing data, a 1 Representing the value added tax invoice, a 2 Representing said other types of notes, a 3 Representing the bank statement.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: and carrying out evaluation calculation on the automatic identification billing model, wherein the calculation formula is as follows:
wherein AS represents the automatic identification billing modelEvaluation result, a 1 Representing the value added tax invoice, a 2 Representing said other types of notes, a 3 Representing the bank statement, KA represents an automatically identifying billing model function, and n represents the billing data.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: and perfecting the automatic identification billing model to enhance the interpretability of the model, wherein the perfected function expression is as follows:
ξ(n)＝∑κ n (KA-KA * ) 2 ；
where ζ (n) represents the refined result of the auto-id billing model, n represents the billing data, κ n Indicating the proximity of the classification result and the actual result of the billing data stage, KA indicates an automatic identification billing model function, KA * An interpretable auto-id billing model function is represented.
As a preferred solution for implementing the automatic identification billing algorithm model based on machine learning according to the present invention, the preferred solution is as follows: monitoring the automatic identification billing model, wherein the monitored function expression is as follows:
M＝argmax[ρ(KA)]；
wherein M represents the monitoring result of the automatic identification billing model, ρ (KA) represents the online learning function of the automatic identification billing model, and KA represents the automatic identification billing model function.
A computer device comprising, a memory for storing instructions; and a processor for executing the instructions to cause the device to perform implementing an automatic identification billing algorithm model based on machine learning.
A computer readable storage medium having stored thereon a computer program which, when executed by the processor, implements a machine learning based implementation of an automatic identification billing algorithm model.
The invention has the beneficial effects that: according to the invention, by collecting and arranging the billing data, solving the data dependency, ensuring the effectiveness of data collection, extracting the characteristics of the billing data, classifying the characteristics in stages, constructing the automatic identification billing model, realizing the accurate identification and automatic identification billing of the billing data, evaluating the automatic identification billing model, verifying the generalization capability of the automatic identification billing model, improving the performance of the automatic identification billing model, perfecting the automatic identification billing model, enhancing the interpretability of the automatic identification billing model, periodically monitoring the automatic identification billing model, finding out the condition of the performance degradation of the automatic identification billing model in time and making corrective measures in time.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. Wherein:
FIG. 1 is a diagram of steps in a method for implementing an automatic identification billing algorithm model based on machine learning in accordance with the present invention;
FIG. 2 is a flow chart of the billing process interaction described by the automatic identification billing algorithm model based on machine learning implementation of the present invention.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.
Further, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic can be included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
Further, in describing the embodiments of the present invention in detail, the cross-sectional view of the device structure is not partially enlarged to a general scale for convenience of description, and the schematic is only an example, which should not limit the scope of protection of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in actual fabrication.
Example 1
In this embodiment, an overall structure schematic diagram of an automatic identification billing algorithm model based on machine learning is provided, as shown in fig. 1, and the automatic identification billing algorithm model based on machine learning includes:
s1, collecting and sorting billing data.
Collecting and sorting data comprising bank statements and notes, ensuring the quality and integrity of the data according to an accounting processing interaction flow, solving the data dependence, ensuring the effectiveness of data collection, wherein the accounting processing interaction flow is shown in fig. 2, and the specific flow steps are as follows:
uploading the bills singly or in batches, and carrying out compliance examination on the bills;
if the bill is not compliant, returning the bill;
if the bill is qualified, carrying out authenticity verification;
if the bill information is not complete, prompting filling information, and carrying out authenticity verification after compliance;
if the bill is not true, returning the bill;
if the bill is true, checking the bill again;
if the bill is repeated, returning the bill;
if the bill is not repeated, comparing the bill with the bank statement;
if the bank has a balance record on the bill, carrying out grading calculation according to the record of the bill;
if the bank has no balance record on the bill, firstly automatically identifying the bill to generate the balance record, and then carrying out grading calculation according to the record of the bill;
the compliance examination is to check whether the ticket information is complete, whether the ticket name is consistent with the enterprise name, whether the invoice is provided with a seal, whether the invoice is out of date, and whether the invoice is not aligned; the authenticity check refers to calling an invoice verification API interface, checking whether the ticket information is consistent with verification data, and returning inconsistent, and not allowing accounting; checking the bill according to four factors of bill code, bill number and amount according to the date of making out the bill, checking whether all the bills in the database have repeated account, and accumulating automatically when the bills are not repeated but are the same as a plurality of bills in a unit; the bank account bill comparison means that all bank transaction records of a database are checked according to three elements of the name of an issuing unit, the name of a reimbursement account person and the name of an item, and whether the bank account bill has the same unit, the same name and the same item name; the hierarchical calculation means calculation according to different levels, the levels including that the total amount of the ticket is equal to the pre-receipt prepaid balance, the total amount of the ticket is greater than the pre-receipt prepaid balance, and the total amount of the ticket is less than the pre-receipt prepaid balance; the automatic identification and accounting means that natural language processing identification is carried out according to five major elements of direction, bill name, semantics, department and special case;
in the specific application, 2000 pieces of data are collected and arranged, compliance examination, authenticity verification, bill check and bank account checking comparison are respectively carried out on the 2000 pieces of data according to the accounting processing interaction flow, the 23 pieces of data with incomplete information are manually filled in, and 6 pieces of data of the same company are automatically accumulated.
S2, extracting features of the billing data.
And extracting different characteristics according to the data name, wherein the extraction rule is as follows:
if the data name is a bank transaction class, extracting the data name as a bank statement;
if the data name is a value-added tax invoice, extracting the data name as a value-added tax invoice;
if the data name is other special bill types, extracting the data name as other bill types;
extracting features from the data, wherein all field information of each banking transaction is one piece of data, all field information of each bill is one piece of data, and each piece of data is bound with a data name;
the data names are bank transaction types including bank statement collection, bank statement payment, initial balance of the home period and end balance of the home period, value tax invoice types including value tax general invoice and value tax special invoice, and other special bill types including plane ticket, train ticket, bus ticket, ship ticket, taxi ticket, road bridge ticket, fuel ticket, business acceptance draft, bank acceptance draft, rating invoice, administrative department legal ticket, administrative department commodity inspection receipt, administrative department fine receipt, donation unified ticket, administrative department non-tax admission ticket, administrative department unified charge receipt, administrative department school education charge receipt, administrative department tax receipt, administrative hospital charge ticket, property maintenance fund special ticket, administrative fund bill, motor vehicle sales unified invoice, second hand vehicle sales unified invoice, insurance fund special invoice, utility ticket, scenic region entrance ticket and business sales invoice;
in the specific application, 2000 pieces of data are subjected to feature extraction, wherein 1160 pieces of banking transaction data, 480 pieces of value-added tax invoice and 360 pieces of other special notes are subjected to conversion according to the requirements, and the final required result is output.
S3, classifying the characteristics in stages, and constructing an automatic identification billing model.
The rules of the phase classification are as follows:
if the characteristic is a value-added tax invoice, direction one identification, bill name identification and semantic one identification are carried out;
if the characteristics are other types of notes, direction one recognition and note name recognition are carried out;
the direction one identification comprises an entry bill and a sales bill, and the semantic one identification comprises a name and remarks;
if the direction one is identified as an entry bill, department identification and enterprise type identification are carried out;
if the direction one is identified as the sales item bill, carrying out enterprise type identification;
the department identification comprises a production department, a technical department, a management department and a business department, and the enterprise type identification comprises a general tax payer and a small-scale tax payer;
if the characteristics are bank statement, direction two recognition and semantic two recognition are carried out;
the direction two identification comprises a borrower and a lender, and the semantic two identification comprises a use, a abstract and a remark;
the ticket name identification comprises the classification of the data names, and is sequentially value-added tax common invoice, value-added tax special invoice, plane ticket, train ticket, automobile ticket, ship ticket, taxi ticket, road and bridge ticket, fuel ticket, business acceptance draft, bank acceptance draft, rating invoice, administrative court litigation ticket, administrative commodity inspection receipt, administrative fine non-fine receipt, donation unified ticket, administrative non-tax income ticket, administrative unified charge receipt, administrative school education charge receipt, administrative tax fine receipt, administrative hospital charge ticket, property maintenance fund special ticket, administrative fund exchange ticket, motor vehicle sales unified invoice, second hand vehicle sales unified invoice, insurance fund special invoice, public utility payment ticket, scenic spot ticket and business sales invoice;
according to the stage classification, the automatic identification billing model is constructed as follows:
wherein KA represents an automatic recognition billing model function, c represents a constant, h n A fitting function representing the phase classification, λ representing the optimal coefficient of the fitting function, N representing billing data, N representing the total number of billing data;
wherein, the function expression of the fitting function is as follows:
h n ＝min[min∑(n-a 1 ) 2 +min∑(n-a 2 ) 2 +min∑(n-a 3 ) 2 ]；
in the formula, h n Fitting function representing phase classification, n representing billing data, a 1 Representing value-added tax invoice, a 2 Representing other kinds of notes, a 3 Representing a bank statement;
according to the stage classification, the billing data are classified, a decision tree is constructed, and accurate identification and automatic identification billing of the billing data are realized by using a powerful classifier of the decision tree.
In the specific application, there are 1160 bank statements, 480 value-added tax invoices and 360 other kinds of invoices, a decision tree is constructed through classification rules, and other billing data are automatically identified and billed by using the decision tree.
S4, evaluating an automatic identification billing model.
And (3) carrying out evaluation calculation on the automatic identification billing model, wherein the calculation formula is as follows:
wherein AS represents the evaluation result of the automatic identification billing model, a 1 Representing value-added tax invoice, a 2 Representing other kinds of notes, a 3 Representing a bank statement, KA represents an automatic identification billing model function, and n represents billing data;
after the constructed automatic identification billing model is trained by utilizing a large amount of data, model evaluation is needed, so that the effects of verifying the generalization capability of the model, improving the model performance and feeding back the data quality can be achieved, the evaluation result of the automatic identification billing model is calculated, the value range is [0,1], and the larger the value is, the more the classification result is matched with the real situation;
in the specific application, the evaluation result of the automatic identification billing model is calculated to be 0.92, and the classification result is proved to be very consistent with the real situation, and meanwhile, the classification result is checked, so that the real situation is basically consistent.
S5, perfecting an automatic identification billing model.
The automatic identification billing model is perfected, the interpretability of the model is enhanced, and a perfected functional expression is as follows:
ξ(n)＝∑κ n (KA-KA * ) 2 ；
where ζ (n) represents the perfected result of the automatic identification billing model, n represents billing data, κ n Indicating the proximity of the classification result to the actual result in the billing data stage, KA indicating an automatic recognition billing model function, KA * Representing an interpretable auto-recognition billing model function;
the interpretation and the credibility of the automatic identification billing model need to be interpreted by using a related algorithm, the automatic identification billing model is interpreted by using the formula, the interpretation and the credibility of the automatic identification billing model are enhanced, the decision reasons of stage classification are given for each billing data, the acceptability and the applicability of the model are enhanced, and the effect and the sustainable development of machine learning in practical application are improved;
s6, periodically monitoring and automatically identifying an accounting model.
Monitoring the automatic identification billing model, wherein the monitored function expression is as follows:
M＝argmax[ρ(KA)]；
wherein M represents a monitoring result of the automatic identification billing model, ρ (KA) represents an online learning function of the automatic identification billing model, and KA represents an automatic identification billing model function;
the performance of the automatic identification billing model in actual application changes along with the change of time, the condition of the performance reduction of the automatic identification billing model can be timely found by periodically monitoring the automatic identification billing model, corresponding measures are taken to correct and optimize the model, feedback information of a user can be collected, the problem of the automatic identification billing model under specific conditions is found, timely repair and improvement are carried out, the accuracy and stability of the automatic identification billing model are improved, and meanwhile, the automatic identification billing model can be analyzed for controllability and interpretability;
in the specific application, when the evaluation result of the automatic identification billing model is lower than 0.7, a manager is reminded to correct and optimize the automatic identification billing model, meanwhile, the automatic identification billing model is interpreted by combining the perfection result of the automatic identification billing model, and the interpretability and the credibility of the automatic identification billing model are enhanced.
Example 2
In this embodiment, a computer device is provided that includes a memory for storing instructions and a processor for executing the instructions to cause the computer device to perform steps for implementing an automatic identification billing algorithm model based on machine learning as described above.
Example 3
In this embodiment, a computer readable storage medium is provided, on which a computer program is stored, which when executed, implements a method for implementing an automatic identification billing algorithm model based on machine learning as described above.
The computer readable storage medium may include: various media capable of storing program codes, such as a U disk, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk.
It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.
Claims (10)
1. An automatic identification billing algorithm model based on machine learning is realized, which is characterized in that: comprising the steps of (a) a step of,
collecting and collating billing data;
extracting features from the billing data;
stage classification is carried out on the characteristics, and an automatic identification billing model is constructed;
evaluating the automatically identified billing model;
perfecting the automatic identification billing model;
the automatic identification billing model is monitored periodically.
2. The machine-learning based implementation of an automatic identification billing algorithm model of claim 1 wherein: the billing process interaction flow of the billing data is as follows:
uploading the bills singly or in batches, and carrying out compliance examination on the bills;
if the bill is not compliant, returning the bill;
if the bill is compliant, carrying out authenticity verification;
if the bill information is not complete, prompting filling information, and carrying out authenticity verification after compliance;
if the bill is not true, returning the bill;
if the bill is true, checking the bill again;
if the bill is repeated, returning the bill;
if the bill is not repeated, comparing the bill with a bank statement;
if the bank statement has a balance record, carrying out grading calculation according to the record of the statement;
if the bank has no balance record on the bill, the bank automatically recognizes the bill to generate the balance record, and then carries out grading calculation according to the record of the bill.
3. The machine-learning based implementation of an automatic identification billing algorithm model of claim 2 wherein: and extracting different features according to the data name, wherein the extraction rule is as follows:
if the data name is a bank transaction class, extracting the data name as a bank statement;
if the data name is a value-added tax invoice class, extracting the data name as a value-added tax invoice;
and if the data name is other special bill types, extracting the data name as other bill types.
4. A machine learning based implementation of an automatic identification billing algorithm model of claim 3 wherein: the rules of the phase classification are as follows:
if the characteristic is a value-added tax invoice, direction one identification, bill name identification and semantic one identification are carried out;
if the characteristics are other types of notes, direction one recognition and note name recognition are carried out;
wherein the direction one identification comprises an entry bill and a pin bill, and the semantic one identification comprises a name and a remark;
if the direction I is identified as an entry bill, department identification and enterprise type identification are carried out;
if the direction I is identified as the sales item bill, carrying out enterprise type identification;
the department identification comprises a production department, a technical department, a management department and a business department, and the enterprise type identification comprises a general tax payer and a small-scale tax payer;
if the characteristic is a bank statement, performing direction two recognition and semantic two recognition;
the second direction identification comprises a borrower and a lender, and the second semantic identification comprises a purpose, a abstract and a remark.
5. The machine-learning based implementation of an automatic identification billing algorithm model of claim 4 wherein: according to the stage classification, the constructed automatic identification billing model is as follows:
wherein KA represents an automatic recognition billing model function, c represents a constant, h n A fitting function representing the phase classification, λ representing the optimal coefficients of the fitting function, N representing the billing data, N representing the total number of billing data;
wherein, the function expression of the fitting function is as follows:
h n ＝min[min∑(n-a 1 ) 2 +min∑(n-a 2 ) 2 +min∑(n-a 3 ) 2 ]；
in the formula, h n Fitting functions representing the phase classifications, n representing the billing data, a 1 Representing the value added tax invoice, a 2 Representing said other types of notes, a 3 Representing the bank statement.
6. The machine-learning based implementation of an automatic identification billing algorithm model of claim 5 wherein: and carrying out evaluation calculation on the automatic identification billing model, wherein the calculation formula is as follows:
wherein AS represents the evaluation result of the automatic identification billing model, a 1 Representing the value added tax invoice, a 2 Representing said other types of notes, a 3 Representing the bank statement, KA represents an automatically identifying billing model function, and n represents the billing data.
7. The machine-learning based implementation of an automatic identification billing algorithm model of claim 6 wherein: and perfecting the automatic identification billing model to enhance the interpretability of the model, wherein the perfected function expression is as follows:
ξ(n)＝∑κ n (KA-KA * ) 2 ；
where ζ (n) represents the refined result of the auto-id billing model, n represents the billing data, κ n Indicating the proximity of the classification result and the actual result of the billing data stage, KA indicates an automatic identification billing model function, KA * An interpretable auto-id billing model function is represented.
8. The machine-learning based implementation of an automatic identification billing algorithm model of claim 7 wherein: monitoring the automatic identification billing model, wherein the monitored function expression is as follows:
M＝argmax[ρ(KA)]；
wherein M represents the monitoring result of the automatic identification billing model, ρ (KA) represents the online learning function of the automatic identification billing model, and KA represents the automatic identification billing model function.
9. A computer device, characterized by: comprising the steps of (a) a step of,
a memory for storing instructions;
a processor for executing the instructions to cause the device to perform the steps of implementing a machine learning based implementation of an automatic identification billing algorithm model as claimed in any of claims 1 to 8.
10. A computer-readable storage medium having stored thereon a computer program, characterized by: the computer program, when executed by said processor, implements a method for implementing an automatic identification billing algorithm model based on machine learning as claimed in any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311453325.0A CN117455699A (en) | 2023-11-02 | 2023-11-02 | Automatic identification billing algorithm model based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311453325.0A CN117455699A (en) | 2023-11-02 | 2023-11-02 | Automatic identification billing algorithm model based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117455699A true CN117455699A (en) | 2024-01-26 |
Family
ID=89583255
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311453325.0A Pending CN117455699A (en) | 2023-11-02 | 2023-11-02 | Automatic identification billing algorithm model based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117455699A (en) |
-
2023
- 2023-11-02 CN CN202311453325.0A patent/CN117455699A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bellini | IFRS 9 and CECL Credit Risk Modelling and Validation: A Practical Guide with Examples Worked in R and SAS | |
| Finlay | Credit scoring, response modeling, and insurance rating: a practical guide to forecasting consumer behavior | |
| US20120011041A1 (en) | Post bankruptcy pattern and transaction detection and recovery apparatus and method | |
| CN109739844B (en) | Data classification method based on attenuation weight | |
| KR102246782B1 (en) | System for Debt Repayment Capability Evaluation Of Corporation Including Evaluation Model Based On AI Using Commerce Data | |
| Cao et al. | Bond rating using support vector machine | |
| Ruyu et al. | A comparison of credit rating classification models based on spark-evidence from lending-club | |
| Barua et al. | Swindle: Predicting the probability of loan defaults using catboost algorithm | |
| CN112037039A (en) | Loan assessment method and device | |
| Ahmed | Robust drivers of Bitcoin price movements: an extreme bounds analysis | |
| Hu | Predicting and improving invoice-to-cash collection through machine learning | |
| CN117114812A (en) | Financial product recommendation method and device for enterprises | |
| CN117455699A (en) | Automatic identification billing algorithm model based on machine learning | |
| CN115827798A (en) | Bill management method established based on industry characteristics | |
| Del Pilar et al. | Towards the Improvement of Credit Card Approval Process Using Classification Algorithm | |
| JP2015125534A (en) | Transaction history data analysis system and transaction history data analysis method | |
| CN114331105A (en) | Electronic draft processing system, method, electronic device and storage medium | |
| Zand | Towards intelligent risk-based customer segmentation in banking | |
| US20180137578A1 (en) | System and method for prediction of deduction claim success based on an analysis of electronic documents | |
| CN111932368B (en) | Credit card issuing system and construction method and device thereof | |
| KR102249015B1 (en) | Calculation System for Corporate Debt Payment Capability | |
| Zaytsev | Selection and evaluation of relevant predictors for credit scoring in peer-to-peer lending with random forest based methods | |
| CN113888143B (en) | Bill data processing method and device and storage medium | |
| Budd | Modelling credit card usage for individual card-holders | |
| Plaksiy et al. | The National Research Nuclear University MEPHI (Moscow Engineering Physics Institute), 31 Kashirskoye shosse, Moscow, Russia kirillplaksiy@ gmail. com, andreinikiforov993@ gmail. com, NGMiloslavskaya@ mephi. ru |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |