CN117934158B - Credit data automatic examination method based on RPA - Google Patents
Credit data automatic examination method based on RPA Download PDFInfo
- Publication number
- CN117934158B CN117934158B CN202410281954.8A CN202410281954A CN117934158B CN 117934158 B CN117934158 B CN 117934158B CN 202410281954 A CN202410281954 A CN 202410281954A CN 117934158 B CN117934158 B CN 117934158B
- Authority
- CN
- China
- Prior art keywords
- information
- variance
- preset
- data
- receptive field
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000012549 training Methods 0.000 claims abstract description 33
- 238000013480 data collection Methods 0.000 claims abstract description 28
- 230000008859 change Effects 0.000 claims abstract description 23
- 230000008569 process Effects 0.000 claims description 27
- 238000000605 extraction Methods 0.000 claims description 23
- 238000007689 inspection Methods 0.000 claims description 11
- 238000012552 review Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000006872 improvement Effects 0.000 abstract description 3
- 230000010365 information processing Effects 0.000 abstract description 2
- 230000001105 regulatory effect Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 101000872084 Danio rerio Delta-like protein B Proteins 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004801 process automation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 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
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/18—Extraction of features or characteristics of the image
- G06V30/186—Extraction of features or characteristics of the image by deriving mathematical or geometrical properties from the whole image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/42—Document-oriented image-based pattern recognition based on the type of document
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Mathematical Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Multimedia (AREA)
- Computational Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Software Systems (AREA)
- Operations Research (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Databases & Information Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of information processing, in particular to an automatic credit information examination method based on RPA, which comprises the following steps: step S1, RPA examines the credit information; step S2, counting the number of error information in a plurality of examination periods, and calculating the variance of the information error rate according to the number counting result; step S3, reconfiguring the number of data collection sources based on the variance of the information error rate, or redetermining the receptive field based on the average change amount of the image noise area to output a first corresponding receptive field; step S4, redetermining the first corresponding receptive field based on the increasing speed of the error quantity of the training set so as to output a second corresponding receptive field; and step S5, adjusting the frequency of the data backup based on the loss rate of the stored data. The invention realizes the improvement of the efficiency and accuracy of data examination.
Description
Technical Field
The invention relates to the technical field of information processing, in particular to an automatic credit information examination method based on RPA.
Background
When a financial institution performs loan business, it is necessary to examine the image data of a loan client to ensure the integrity and accuracy of the data. However, conventional manual inspection methods are inefficient and prone to error, resulting in long processing times, high costs, and risks. Thus, there is a need for an automated post-credit data review method that increases efficiency and reduces errors.
The patent application with the publication number of CN111523996A discloses an approval method and a system, wherein the method comprises the following steps: basic approval information of a user to be approved is obtained, wherein the basic approval information comprises a client application form, a credit investigation report, operation data, bank running water, real estate information, client three-party behavior data information, telephone meeting information, external visit information, repayment information, early warning information and collection urging information; inputting the basic approval information into an approval decision model, and acquiring a risk rule examination result and credit scores of the users to be approved, wherein the credit score model is obtained after training through a sample and a label; and according to the risk rule examination result and the credit score of the user to be examined, giving proposal of the user to be examined in the aspects of admittance, credit giving and risk pricing. Therefore, the approval method and the approval system have the following problems: inaccuracy in the stability determination of the information acquisition due to the reflection of the variance of the information error rate and inaccuracy in the adjustment of the receptive field according to the average amount of change in the noise area of the several areas result in reduced accuracy in the examination.
Disclosure of Invention
Therefore, the invention provides an RPA-based credit data automatic examination method, which is used for solving the problems that in the prior art, the stability judgment of information acquisition reflected by the variance of an information error rate is inaccurate, and the examination accuracy is reduced due to inaccurate adjustment of a receptive field according to the average variation of the noise areas of a plurality of areas.
In order to achieve the above object, the present invention provides an automatic credit information auditing method based on RPA, comprising: step S1, an RPA logs in an image system to acquire a credit information image of a user, carries out information identification on the credit information image to output to-be-inspected credit data of the user, and compares the to-be-inspected credit data of the user with user data input into the RPA to finish credit information inspection; step S2, counting the number of error information in a plurality of examination periods, and calculating the variance of the information error rate according to the number counting result; step S3, if the stability of the information acquisition is lower than the allowable range based on the variance of the information error rate, the number of the data collection sources is reconfigured, or the receptive field is redetermined based on the average change amount of the image noise area so as to output a first corresponding receptive field; step S4, training the credit data image after outputting the first corresponding receptive field, and if the image acquisition effectiveness is judged to be lower than the allowable range based on the increasing speed of the error quantity of the training set, re-determining the first corresponding receptive field to output a second corresponding receptive field; and S5, if the information leakage risk is judged to be out of the allowable range based on the loss rate of the storage data after the number of the data collection sources is reconfigured, the frequency of the data backup is adjusted.
Further, the determining process of the stability of the information acquisition includes determining that the stability of the information acquisition is lower than an allowable range if the variance of the information error rate satisfies a preset first variance condition or satisfies a preset second variance condition;
if the variance of the information error rate only meets a preset second variance condition, reconfiguring the number of data collection sources;
And if the variance of the information error rate only meets a preset first variance condition, primarily judging that the comprehensiveness of the image feature extraction is lower than the allowable range, and secondarily judging the comprehensiveness of the image feature extraction according to the average change quantity of the image noise area.
The preset first variance condition is that the variance of the information error rate is larger than the preset first variance and smaller than or equal to the preset second variance; the preset second variance condition is that the variance of the information error rate is larger than the preset second variance; the preset first variance is smaller than the preset second variance.
Further, the calculation formula of the average variation of the image noise area is as follows: wherein B is the average variation of the image noise area, (B i-bi-1) is the difference between the i-th area noise area and the i-1-th area noise area, n is the number of image dividing areas, and n is a natural number greater than or equal to 2.
Further, the process of reconfiguring the number of data collection sources includes raising the number of data collection sources according to a difference between a variance of the information error rate and a preset second variance.
Further, the process of performing secondary judgment on the comprehensiveness of the image feature extraction comprises the step of determining again the receptive field if the average change amount of the image noise area is larger than the preset change amount and the comprehensiveness of the image feature extraction is lower than the allowable range.
Further, the process of redetermining the receptive field includes reducing the receptive field according to a difference between an average change amount of the image noise area and a preset change amount to output a first corresponding receptive field.
Further, in the step S4, the determining process for the validity of the image acquisition includes: and acquiring the error quantity of the training set and calculating the increasing speed of the error quantity of the training set when the first corresponding receptive field is output, and if the increasing speed of the error quantity of the training set is larger than the preset speed, judging that the image acquisition effectiveness is lower than the allowable range, and re-determining the first corresponding receptive field.
Further, the process of redefining the first corresponding receptive field includes increasing the first corresponding receptive field according to a difference between an increasing speed of the number of training set errors and a preset speed to output a second corresponding receptive field.
Further, the specific steps of the step S5 include:
step S51, when the reconfiguration of the number of the data collection sources is completed, calculating the loss rate of the stored data, and judging whether the information leakage risk is within an allowable range according to the loss rate of the stored data;
And step S52, if the information leakage risk exceeds the allowable range, adjusting the frequency of the data backup.
Further, the process of adjusting the data backup includes adjusting a frequency of the data backup according to a difference between a loss rate of the stored data and a preset loss rate, wherein the frequency of the data backup is inversely proportional to the difference between the loss rate of the stored data and the preset loss rate.
Compared with the prior art, the method has the beneficial effects that the method automatically examines the credit data based on RPA by setting steps S1-S5, reduces the time and effort of manual operation, judges the stability of information acquisition reflected by the variance of the information error rate and reconfigures the number of data collection sources in the examination process, reduces the influence on the information credibility caused by single collection channel, or reduces the loss of information on the premise of ensuring the image characteristic extraction effect on the redetermination of the receptive field, reduces the frequency of data backup, reduces the influence on the examination process caused by data loss, and realizes the improvement of the efficiency and accuracy of data examination.
Further, in the process of checking the credit information, because of the problem of image quality of the acquired credit image or more image noise caused by interference of different signal sources in the transmission process, if the image features are still extracted according to the original receptive field at the moment, the error information is extracted, so that the checking accuracy is reduced, and the receptive field is reduced, so that the capturing effect of the local features is improved, and the efficiency and accuracy of checking the information are further improved by setting a preset variation difference value, a preset first adjustment coefficient and a preset second adjustment coefficient.
Further, the method trains the credit data images after outputting the first corresponding receptive field, but the extraction area is too small, the extraction of part of characteristic information cannot reach the effective range, so that the training set is continuously error, the first corresponding receptive field is redetermined by setting a preset growth speed difference value, a preset third regulating coefficient and a preset fourth regulating coefficient, more original information is reserved on the premise of ensuring the extraction precision, the loss of information is reduced, and the efficiency and the accuracy of data examination are further improved.
Furthermore, the system of the invention reconfigures the data collection sources, and improves the risk of information exposure, which can lead to the increase of attack degree when effective information is extracted by different collection sources, so that the efficiency of examination is reduced due to the loss of stored information and the decrease of information integrity, and the efficiency and accuracy of data examination are further improved by reducing the influence of data loss on the examination process by reducing the frequency of data backup.
Drawings
FIG. 1 is a flowchart illustrating an exemplary RPA-based credit data automatic review method according to an embodiment of the present invention;
FIG. 2 is a flowchart showing a step S3 of an RPA-based credit data automatic review method according to an embodiment of the present invention;
FIG. 3 is a flowchart showing a step S4 of an RPA-based credit data automatic review method according to an embodiment of the present invention;
fig. 4 is a flowchart showing a step S5 of the automatic credit information review method based on RPA according to the embodiment of the present invention.
Detailed Description
In order that the objects and advantages of the invention will become more apparent, the invention will be further described with reference to the following examples; it should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are merely for explaining the technical principles of the present invention, and are not intended to limit the scope of the present invention.
Fig. 1, fig. 2, fig. 3, and fig. 4 show an overall flowchart of an RPA-based credit data automatic review method, a specific flowchart of step S3, a specific flowchart of step S4, and a specific flowchart of step S5, respectively, according to an embodiment of the present invention. The invention discloses an automatic credit information examination method based on RPA, which comprises the following steps:
Step S1, an RPA logs in an image system to acquire a credit information image of a user, carries out information identification on the credit information image to output to-be-inspected credit data of the user, and compares the to-be-inspected credit data of the user with user data input into the RPA to finish credit information inspection;
Step S2, counting the number of error information in a plurality of examination periods, and calculating the variance of the information error rate according to the number counting result;
Step S3, if the stability of the information acquisition is lower than the allowable range based on the variance of the information error rate, the number of the data collection sources is reconfigured, or the receptive field is redetermined based on the average change amount of the image noise area so as to output a first corresponding receptive field;
Step S4, training the credit data image after outputting the first corresponding receptive field, and if the image acquisition effectiveness is judged to be lower than the allowable range based on the increasing speed of the error quantity of the training set, re-determining the first corresponding receptive field to output a second corresponding receptive field;
and S5, if the information leakage risk is judged to be out of the allowable range based on the loss rate of the storage data after the number of the data collection sources is reconfigured, the frequency of the data backup is adjusted.
Specifically, the specific steps of the step S3 include:
step S31, if the stability of the information acquisition is lower than the allowable range, reconfiguring the number of data collection sources, or preliminarily judging that the comprehensiveness of the image feature extraction is lower than the allowable range;
in step S32, if the comprehensiveness of the image feature extraction is determined to be lower than the allowable range based on the average variation of the image noise area, the receptive field is redetermined to output the first corresponding receptive field.
Specifically, the specific steps of the step S4 include:
Step S41, training the credit data image after outputting the first corresponding receptive field, obtaining the error quantity of the training set and calculating the increasing speed of the error quantity of the training set;
step S42, judging the effectiveness of image acquisition according to the increasing speed of the error quantity based on the training set;
and S43, calculating the difference value between the increasing speed of the error quantity of the training set and the preset speed when the image acquisition effectiveness is lower than the allowable range, and redetermining the first corresponding receptive field to output a second corresponding receptive field.
Specifically, the RPA is collectively referred to as robotic flow automation (Robotic Process Automation).
Specifically, the calculation formula of the information error rate is:，
Where F is an information error rate, Q j is the number of error information in the jth inspection period, Q j is the total number of information in the jth inspection period, l is the number of inspection periods, and l is a natural number of 1 or more.
Specifically, the error information is information in which the same data feature does not exist in the user data.
Specifically, the variance of the information error rate is the variance of the proportion of the number of error messages in several inspection periods to the total number of messages, and it can be understood that the calculation method of the variance of the information error rate is a common technical means for those skilled in the art, and will not be described herein.
Specifically, the receptive field means the size of the region pixels divided at the time of single feature extraction.
Specifically, the calculation formula of the increasing speed of the number of training set errors is as follows:
V = (m2-m1)/(t2-t1)
Wherein V is the increasing speed of the error quantity of the training set, m1 is the error quantity of the training set at the time t1, and m2 is the error quantity of the training set at the time t 2.
Specifically, the calculation formula of the loss rate of the stored data is as follows:
P = p1/p0
where P is the loss rate of the storage data, P1 is the loss amount of the storage data, and P0 is the total amount of the storage data.
According to the method, the credit data is automatically inspected based on RPA (remote procedure analysis) in the steps S1-S5, so that the time and effort of manual operation are reduced, the stability of information acquisition reflected by the variance of the information error rate is judged and the number of data collection sources is reconfigured in the inspection process, the influence on the information credibility due to the collection of a single collection channel is reduced, or the loss of information is reduced on the premise that the image feature extraction effect is ensured due to the redetermination of a receptive field, the frequency of data backup is reduced, the influence on the inspection process due to the data loss is reduced, and the improvement of the efficiency and accuracy of data inspection is realized.
With continued reference to fig. 2, the determining process for the stability of the information acquisition includes determining that the stability of the information acquisition is lower than an allowable range if the variance of the information error rate satisfies a preset first variance condition or satisfies a preset second variance condition;
if the variance of the information error rate only meets a preset second variance condition, reconfiguring the number of data collection sources;
And if the variance of the information error rate only meets a preset first variance condition, primarily judging that the comprehensiveness of the image feature extraction is lower than the allowable range, and secondarily judging the comprehensiveness of the image feature extraction according to the average change quantity of the image noise area.
The preset first variance condition is that the variance of the information error rate is larger than the preset first variance and smaller than or equal to the preset second variance; the preset second variance condition is that the variance of the information error rate is larger than the preset second variance; the preset first variance is smaller than the preset second variance.
The calculation formula of the average change amount of the image noise area is as follows: Wherein B is the average variation of the image noise area, (B i-bi-1) is the difference between the i-th area noise area and the i-1-th area noise area, n is the number of image dividing areas, and n is a natural number greater than or equal to 2. With continued reference to fig. 2, the process of reconfiguring the number of data collection sources includes increasing the number of data collection sources according to a difference between a variance of the information error rate and a preset second variance.
Specifically, the variance of the information error rate is denoted as X, the preset first variance is denoted as X1, the preset second variance is denoted as X2, the difference between the variance of the information error rate and the preset first variance is denoted as DeltaX, the preset variance difference is denoted as DeltaX 0, the number of data collection sources is denoted as S,
If Δx is less than or equal to Δx0, the number of data collection sources is adjusted to S' =ceil (7S/5);
if Δx > Δx0, the number of data collection sources is adjusted to S' =ceil (7S/3).
With continued reference to fig. 2, the process of performing the secondary determination on the comprehensiveness of the image feature extraction includes, if the average variation of the image noise area is greater than the preset variation, re-determining the receptive field if the comprehensiveness of the secondary determination image feature extraction is lower than the allowable range.
The process of redetermining the receptive field comprises the step of reducing the receptive field according to the difference value of the average change quantity of the image noise area and the preset change quantity so as to output a first corresponding receptive field.
Specifically, the average change amount of the image noise area is denoted as B, the preset change amount is denoted as B0, the difference between the average change amount of the image noise area and the preset change amount is denoted as DeltaB, the difference between the preset change amount is denoted as DeltaB 0, the receptive field is denoted as Y, the first adjustment coefficient alpha 1 is preset, the second adjustment coefficient alpha 2 is preset, wherein 1 < alpha 2,
If Δb is less than or equal to Δb0, the first corresponding receptive field Y' =y×α1;
If Δb > - Δb0, then the first corresponding receptive field Y' =y×α2.
In the method, in the process of checking the credit information, because of the image quality problem of the acquired credit image or more image noise points caused by the interference of different signal sources in the transmission process, if the image features are still extracted according to the original receptive field at the moment, the error information is extracted, so that the checking accuracy is reduced, and the receptive field is reduced to improve the capturing effect of the local features by setting the preset variation difference value, the preset first regulating coefficient and the preset second regulating coefficient, so that the efficiency and the accuracy of checking the information are further improved.
With continued reference to fig. 3, in the step S4, the process of determining the validity of the image acquisition includes: and acquiring the error quantity of the training set and calculating the increasing speed of the error quantity of the training set when the first corresponding receptive field is output, and if the increasing speed of the error quantity of the training set is larger than the preset speed, judging that the image acquisition effectiveness is lower than the allowable range, and re-determining the first corresponding receptive field.
The process of redefining the first corresponding receptive field comprises the step of increasing the first corresponding receptive field according to the difference value between the increasing speed of the number of errors of the training set and the preset speed so as to output a second corresponding receptive field.
Specifically, the increasing speed of the number of training set errors is denoted as V, the preset speed is denoted as V0, the difference between the increasing speed of the number of training set errors and the preset speed is denoted as DeltaV, deltaV=V-V0 is set, the preset increasing speed difference is denoted as DeltaV 0, the third adjustment coefficient alpha 3 is preset, the fourth adjustment coefficient alpha 4 is preset, wherein 0 < alpha 3 < alpha 4 <1,
If DeltaV is less than or equal to DeltaV 0, the second corresponding receptive field Y '=Y' × (1- α3);
if Δv > - Δv0, then the second corresponding receptive field Y "=y' × (1- α4).
The method trains the credit data images after outputting the first corresponding receptive fields, but the extraction area is too small, the extraction of part of characteristic information cannot reach the effective range, so that the training set continuously generates errors, the first corresponding receptive fields are redetermined by setting a preset increase speed difference value, a preset third regulating coefficient and a preset fourth regulating coefficient, more original information is reserved on the premise of ensuring the extraction precision, the loss of information is reduced, and the efficiency and the accuracy of data examination are further improved.
With continued reference to fig. 4, the specific steps of step S5 include:
step S51, when the reconfiguration of the number of the data collection sources is completed, calculating the loss rate of the stored data, and judging whether the information leakage risk is within an allowable range according to the loss rate of the stored data;
And step S52, if the information leakage risk exceeds the allowable range, adjusting the frequency of the data backup.
The process of adjusting the data backup comprises the step of adjusting the frequency of the data backup according to the difference value of the loss rate of the stored data and the preset loss rate, wherein the frequency of the data backup is inversely proportional to the difference value of the loss rate of the stored data and the preset loss rate.
Specifically, the loss rate of the stored data is denoted as P, the preset loss rate is denoted as P0, the difference Δp between the loss rate of the stored data and the preset loss rate is set to Δp=p-P0, the preset loss rate difference is denoted as Δp0, the frequency of data backup is denoted as H,
If delta P is less than or equal to delta P0, the frequency H' =3H/4 of the data backup after adjustment;
If Δp > Δp0, the frequency of the post-data backup is adjusted H' =h/2.
The system of the invention reconfigures the data collection sources, and improves the risk of information exposure, which can lead to the increase of attack degree when effective information is extracted by different collection sources, so that the efficiency of examination is reduced due to the loss of stored information and the decrease of information integrity, and the efficiency and accuracy of data examination are further improved by reducing the influence of data loss on the examination process by reducing the frequency of data backup.
Example 1
In this embodiment, when the data collection sources are reconfigured to 7, the loss rate p=0.13 of the stored data is calculated, the preset loss rate is denoted as p0=0.08, the difference Δp0=0.04 between the preset loss rate and the difference Δp=0.13-0.08=0.05 between the loss rate of the stored data and the preset loss rate, the frequency h=1/H of data backup,
If Δp > - Δp0 is determined, the frequency of the data backup after adjustment is H' = (H/2) =0.5 times/H.
The method of the invention ensures the integrity of the data by reducing the frequency of data backup, reduces the possibility of data loss and further improves the efficiency and accuracy of data examination.
Thus far, the technical solution of the present invention has been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of protection of the present invention is not limited to these specific embodiments. Equivalent modifications and substitutions for related technical features may be made by those skilled in the art without departing from the principles of the present invention, and such modifications and substitutions will be within the scope of the present invention.
Claims (6)
1. An automatic review method for credit information based on RPA, comprising:
Step S1, an RPA logs in an image system to acquire a credit information image of a user, carries out information identification on the credit information image to output to-be-inspected credit data of the user, and compares the to-be-inspected credit data of the user with user data input into the RPA to finish credit information inspection;
Step S2, counting the number of error information in a plurality of examination periods, and calculating the variance of the information error rate according to the number counting result;
Step S3, if the stability of the information acquisition is lower than the allowable range based on the variance of the information error rate, the number of the data collection sources is reconfigured, or the receptive field is redetermined based on the average change amount of the image noise area so as to output a first corresponding receptive field;
Step S4, training the credit data image after outputting the first corresponding receptive field, and if the image acquisition effectiveness is judged to be lower than the allowable range based on the increasing speed of the error quantity of the training set, re-determining the first corresponding receptive field to output a second corresponding receptive field;
step S5, if the information leakage risk is judged to be out of the allowable range based on the loss rate of the storage data after the number of the data collection sources is reconfigured, the frequency of the data backup is adjusted;
the judging process of the stability of the information acquisition comprises the step of judging that the stability of the information acquisition is lower than an allowable range if the variance of the information error rate meets a preset first variance condition or a preset second variance condition;
if the variance of the information error rate only meets a preset second variance condition, reconfiguring the number of data collection sources;
if the variance of the information error rate only meets a preset first variance condition, primarily judging that the comprehensiveness of the image feature extraction is lower than an allowable range, and secondarily judging the comprehensiveness of the image feature extraction according to the average variation of the image noise area;
The preset first variance condition is that the variance of the information error rate is larger than the preset first variance and smaller than or equal to the preset second variance; the preset second variance condition is that the variance of the information error rate is larger than the preset second variance; the preset first variance is smaller than the preset second variance;
the process of reconfiguring the number of the data collection sources comprises the steps of adjusting the number of the data collection sources according to the difference value between the variance of the information error rate and the preset second variance;
The process of carrying out secondary judgment on the comprehensiveness of the image feature extraction comprises the steps of if the average change amount of the image noise area is larger than a preset change amount, carrying out secondary judgment on the comprehensiveness of the image feature extraction is lower than an allowable range, and carrying out re-determination on a receptive field;
the process of redetermining the receptive field comprises the step of reducing the receptive field according to the difference value of the average change quantity of the image noise area and the preset change quantity so as to output a first corresponding receptive field.
2. The automatic review method of credit information based on RPA according to claim 1, wherein the calculation formula of the average variation of the image noise area is:
，
Wherein B is the average variation of the image noise area, B i-bi-1 is the difference between the i-th area noise area and the i-1-th area noise area, n is the number of image dividing areas, and n is a natural number greater than or equal to 2.
3. The automatic review of RPA-based credit data method according to claim 1, wherein in the step S4, the determining process for the validity of the image acquisition includes: and acquiring the error quantity of the training set and calculating the increasing speed of the error quantity of the training set when the first corresponding receptive field is output, and if the increasing speed of the error quantity of the training set is larger than the preset speed, judging that the image acquisition effectiveness is lower than the allowable range, and re-determining the first corresponding receptive field.
4. The RPA-based credit data automatic review method of claim 3 wherein the step of redefining the first corresponding receptive field comprises increasing the first corresponding receptive field to output a second corresponding receptive field based on a difference between a rate of increase in the number of training set errors and a preset rate.
5. The method for automatically checking credit information based on RPA according to claim 4, wherein the specific steps of step S5 include:
step S51, when the reconfiguration of the number of the data collection sources is completed, calculating the loss rate of the stored data, and judging whether the information leakage risk is within an allowable range according to the loss rate of the stored data;
And step S52, if the information leakage risk exceeds the allowable range, adjusting the frequency of the data backup.
6. The RPA-based credit profile automatic review method of claim 5, wherein the adjusting the data backup comprises adjusting a frequency of the data backup based on a difference between a loss rate of the stored data and a preset loss rate, wherein the frequency of the data backup is inversely proportional to the difference between the loss rate of the stored data and the preset loss rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410281954.8A CN117934158B (en) | 2024-03-13 | 2024-03-13 | Credit data automatic examination method based on RPA |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410281954.8A CN117934158B (en) | 2024-03-13 | 2024-03-13 | Credit data automatic examination method based on RPA |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117934158A CN117934158A (en) | 2024-04-26 |
| CN117934158B true CN117934158B (en) | 2024-08-13 |
Family
ID=90750999
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410281954.8A Active CN117934158B (en) | 2024-03-13 | 2024-03-13 | Credit data automatic examination method based on RPA |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117934158B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5828789A (en) * | 1996-02-19 | 1998-10-27 | Fuji Xerox Co., Ltd. | Image coding system and image decoding system |
| CN111723669A (en) * | 2020-05-21 | 2020-09-29 | 湖南三湘银行股份有限公司 | Bank client identity processing method based on space positioning |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1999045713A1 (en) * | 1998-03-05 | 1999-09-10 | Matsushita Electric Industrial Co., Ltd. | Image coding method, image coding / decoding method, image coder, or image recording/reproducing apparatus |
| JP2006072550A (en) * | 2004-08-31 | 2006-03-16 | Sumitomo Mitsui Banking Corp | Credit management method and system |
| FR2925193B1 (en) * | 2007-12-14 | 2009-12-11 | Advanced Track & Trace | METHOD AND DEVICE FOR FIGHT AGAINST COUNTERFEITING |
| US10268862B2 (en) * | 2015-04-16 | 2019-04-23 | Huawei Technologies Co., Ltd. | Fingerprint collection method, fingerprint collector, and terminal |
| CN106096348B (en) * | 2016-06-07 | 2019-04-12 | 华中科技大学 | A kind of card verifying system and method based on multidimensional code |
| CN110443137B (en) * | 2019-07-03 | 2023-07-25 | 平安科技（深圳）有限公司 | Multi-dimensional identity information identification method and device, computer equipment and storage medium |
| JP7325755B2 (en) * | 2020-01-21 | 2023-08-15 | 本田技研工業株式会社 | Information processing device, information processing method and program |
| CN111860481B (en) * | 2020-06-30 | 2021-08-10 | 湖南三湘银行股份有限公司 | Online banking service system based on multiple identification parameters |
| CN113034263A (en) * | 2020-08-20 | 2021-06-25 | 深圳大学 | Credit approval method, credit approval device, credit approval server and credit approval medium |
| CN115019166B (en) * | 2022-05-24 | 2024-02-09 | 深圳大学 | Method, device, medium and terminal for extracting marsh wetland information |
| CN117058236A (en) * | 2023-08-17 | 2023-11-14 | 中国矿业大学 | Target identification positioning method based on multi-vision system self-switching |
| CN117437482A (en) * | 2023-11-15 | 2024-01-23 | 广东电网有限责任公司 | Cable main insulation part defect identification method and system |
-
2024
- 2024-03-13 CN CN202410281954.8A patent/CN117934158B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5828789A (en) * | 1996-02-19 | 1998-10-27 | Fuji Xerox Co., Ltd. | Image coding system and image decoding system |
| CN111723669A (en) * | 2020-05-21 | 2020-09-29 | 湖南三湘银行股份有限公司 | Bank client identity processing method based on space positioning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117934158A (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537215A (en) | A kind of flame detecting method based on image object detection | |
| CN106570631B (en) | P2P platform-oriented operation risk assessment method and system | |
| CN112819093A (en) | Man-machine asynchronous recognition method based on small data set and convolutional neural network | |
| CN117237126B (en) | Insurance platform and insurance data processing method | |
| CN117934158B (en) | Credit data automatic examination method based on RPA | |
| CN112037006A (en) | Credit risk identification method and device for small and micro enterprises | |
| CN113095579B (en) | Daily-scale rainfall forecast correction method coupled with Bernoulli-gamma-Gaussian distribution | |
| CN113435464B (en) | Abnormal data detection method and device, electronic equipment and computer storage medium | |
| CN116416078B (en) | Audit supervision method for maintaining fund accounting safety | |
| CN112329862A (en) | Decision tree-based anti-money laundering method and system | |
| CN115660834B (en) | Individual loan risk assessment method based on decision tree | |
| KR102525491B1 (en) | Method of providing structure damage detection report | |
| CN112418313B (en) | Big data online noise filtering system and method | |
| CN111598184B (en) | DenseNet-based image noise identification method and device | |
| CN111767815A (en) | Tunnel water leakage identification method | |
| KR20210105774A (en) | Method and apparatus for sampling of cloud outsorcing task for inspection by using artificial intelligent | |
| CN112396648A (en) | Target identification method and system capable of positioning mass center of target object | |
| CN117436770B (en) | BIM-based building engineering construction management method and system | |
| KR100686466B1 (en) | System and method for valuing loan portfolios using fuzzy clustering | |
| CN118333633B (en) | Credit and debt standing book management system and method based on big data | |
| CN118261700B (en) | Comprehensive merchant data management system based on cloud computing | |
| CN118505282A (en) | Base difference trend prediction method based on generation type countermeasure network | |
| CN118864854A (en) | Statistical information-introduced weighted Focal loss function and semantic segmentation network model training method | |
| CN118296531A (en) | Business abnormality detection method, device, equipment and medium | |
| CN118195561A (en) | Human resource management method and system with intelligent accounting salary |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |