A bill information collaborative management system and method based on a whole life cycle

Abstract

This invention discloses a collaborative governance system and method for invoice information based on the entire lifecycle, relating to the field of data processing technology. The invention first retrieves the review data within the current invoice's lifecycle; then extracts the usage attributes of the review data and matches the categories of the review data based on these attributes; finally, it calculates the correlation between the review data based on the categories, and determines whether the current invoice triggers a review based on historical correlation data, optimizing the correlation acquisition process based on the review results. This invention comprehensively captures compliance characteristics and shortens the processing cycle by covering the entire invoice lifecycle, integrating multi-dimensional review data, and calculating a review index based on historical data; by dividing usage attribute intervals and determining thresholds, it accurately identifies high-risk invoices, avoiding unnecessary reviews and improving governance efficiency; and by dynamically adjusting weights through optimization cycles, it achieves iterative optimization of the governance mechanism, ensuring long-term accuracy and adaptability.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a collaborative governance system and method for invoice information based on the entire lifecycle. Background Technology

[0002] With the deepening development of the digital economy, the form and circulation mode of fiscal receipts continue to evolve, and electronic receipts are gradually replacing paper receipts.

[0003] When companies submit invoices for reimbursement, the sources of invoices are quite diverse. External confirmation of invoices not only consumes a lot of communication costs and prolongs the reimbursement cycle, but may also lead to confirmation failure due to untimely response or lack of transparency from the issuing entity. Therefore, it is urgent to establish a review trigger judgment mechanism within the company for invoices from different sources. Summary of the Invention

[0004] The purpose of this invention is to provide a collaborative governance system and method for invoice information based on the entire lifecycle, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A collaborative governance method for invoice information based on the entire lifecycle, comprising the following steps:

[0007] Step S1: Retrieve the audit data within the lifecycle of the current invoice. The lifecycle includes the issuance generation cycle, the circulation authorization cycle, and the invoice verification cycle. Each lifecycle includes several types of audit data.

[0008] Furthermore, step S1 also includes:

[0009] The invoice is a fiscal electronic invoice, which contains several preset verification information items. The verification information is a structured field used to represent the content of the invoice. A preset standard library is used to match the verification information with data in the standard library to verify whether the invoice information is complete. The verification information is different in different lifecycles.

[0010] The core principle of matching is to use a pre-defined standardized rule system, i.e., a standard library, as a benchmark to perform multi-dimensional consistency and compliance verification of the target verification information in the invoice:

[0011] First, the standard library is broken down into structured benchmark rules that include data format, field connotation, logical constraints, and other dimensions. Then, the invoice verification information is standardized and transformed into a structured form that can be compared with the benchmark rules. Subsequently, through field-level semantic adaptation verification, format compliance verification, and logical constraint verification, the matching result is quantitatively output by the proportion of the number of information items that pass the verification to the total number of information items verified in the scenario. Essentially, it is an information validity verification process based on a unified benchmark.

[0012] The invoice generation cycle is the process from when the invoice issuer receives the invoice issuance request to when the invoice generation is completed.

[0013] The circulation authorization cycle is the process from the completion of the invoice generation to the time when the invoice user obtains the right to use the invoice through authorization. The use of the invoice includes the reimbursement of the invoice and the verification of the invoice.

[0014] The invoice verification period is the process from when the invoice user initiates the invoice usage request to when the invoice usage is completed;

[0015] The audit data includes the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i ;

[0016] The information matching rate X i This represents the percentage of complete verification information items in the i-th lifecycle out of all verification information items.

[0017] Each lifecycle contains several preset processes. Each time a ticket passes through a process, the ticket is marked; the process completion rate Y i This represents the percentage of the number of tags on the ticket within the i-th lifecycle in the total number of lifecycle processes.

[0018] The timeliness compliance rate Z i Z is the quantified value of the proportion of the time spent on a ticket in its i-th lifetime to the preset time limit for that lifetime. i =[1-(Z a / Z t )]×100%, where Z a Z represents the time elapsed during the lifecycle of the note. t This represents the preset time limit for this lifecycle, when Z a ≥Z t At that time, Z i Set to 0.

[0019] Furthermore, the methods for retrieving the audit data within the lifecycle of the current invoice in step S1 include:

[0020] Step S1-1: Preset the statistical time, obtain the audit data of all invoices within the lifecycle of the statistical time, and integrate the audit data into a historical audit set H. i H i H represents the historical review set for the i-th lifecycle; the historical review set H i It includes the verification information for each invoice, including the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i ;

[0021] Step S1-2: Statistical analysis of historical audit set H i For each type of audit data, calculate the frequency of each value in the corresponding audit data to form a frequency set S. i ={P Eij}, where S i P represents the set of frequencies for the i-th lifetime. Eij This represents the frequency of the j-th value of the E-th type of audit data during the i-th lifecycle.

[0022] Steps S1-3: Integrate the audit data within the current invoice's lifecycle into an audit set R={X} i Y i Z i}; Retrieve the historical review set H i The intersection of the audit set R and the frequency set S is used to obtain the numerical values ​​of the audit data within the intersection. i The frequency of occurrence in each element is used as the intersection frequency. The intersection frequencies of elements with the same lifecycle are summed to obtain the current audit index L of the invoice. i0 L i L represents the review index for the i-th lifecycle. i0 The audit index represents the current audit index at the i-th stage of the bill's lifecycle.

[0023] Step S2: Extract the usage attributes of the audit data, and match the category of the audit data according to the usage attributes;

[0024] Furthermore, step S2 also includes:

[0025] Step S2-1: Extract the review index of all invoices within the statistical period and integrate them into a historical usage attribute set U={U1, U2, ... U... n}, where U n U represents the usage attribute of the nth ticket within the statistical time period. nThe system includes the review index of the nth invoice within the statistical period; all review indices in the historical usage attribute set U are evenly divided into M usage attribute intervals according to their lifecycles, the number of invoices in each usage attribute interval is counted, and the coverage C of each usage attribute interval is calculated. m =m0 / N, where C m This represents the coverage of the m-th attribute interval, m0 represents the number of tickets in the m-th attribute interval, and N represents the total number of tickets within the statistical time period.

[0026] Step S2-2: Obtain the matching threshold for the corresponding lifecycle based on the median of the usage attribute interval corresponding to the largest coverage value in the same lifecycle.

[0027] Step S2-3: Obtain the current ticket's usage attribute set U0={L i0}; Obtain the corresponding matching threshold T i T i The matching threshold represents the i-th lifecycle. Based on the number of documents with an approval index greater than the corresponding matching threshold, the documents are divided into several categories, and then based on L... i0 Greater than T i The number of items matches the category of the current invoice's audit data.

[0028] Step S3: Obtain the correlation between the audit data according to the category, as the correlation data; extract historical correlation data, and determine whether the current invoice needs to be reviewed based on the historical correlation data;

[0029] Furthermore, step S3 also includes:

[0030] Step S3-1: Calculate the correlation degree based on the category of the audit data of the invoice;

[0031] ;

[0032] Among them G r L represents the correlation between the audit data of the r-th category. i w represents the review index for the i-th lifecycle. ir This represents the preset weight coefficient of the corresponding review index in the r-th category. The preset weight coefficients of different categories are independent of each other.

[0033] Step S3-2: Extract the correlation degree of all tickets within the statistical time period and integrate them into a historical correlation degree set G; divide the correlation degree in the historical correlation degree set G evenly into k correlation degree intervals, where M is the number of usage attribute intervals divided by the historical usage attribute set U, and k is rounded down; count the number of tickets in each correlation interval, take the correlation interval with the most tickets as the correlation threshold interval, and take the upper limit of the correlation threshold interval as the review threshold; if the correlation interval with the most tickets is not unique, take the upper limit of the correlation threshold interval with the largest upper limit as the review threshold.

[0034] Calculate the relevance of the current invoice. If the relevance of the current invoice is greater than or equal to the review threshold, it is determined that no review will be triggered; if the relevance of the current invoice is less than the review threshold, it is determined that a review will be triggered.

[0035] Step S4: Preset optimization period, and optimize the process of obtaining the correlation between audit data based on the review results within the optimization period;

[0036] Furthermore, step S4 also includes:

[0037] Step S4-1: The review results include review compliance and review non-compliance;

[0038] When the review result is compliant, update the current relevance of the invoice to obtain the updated relevance, and record the updated relevance into the historical relevance set G:

[0039] ;

[0040] in Represents the updated relevance of the current invoice. This represents the pre-update relevance of the current invoice for review, where F represents the review threshold. Represents the preset update coefficient;

[0041] When the review result is non-compliant, the current invoice will be marked as an abnormal invoice and included in the historical abnormal invoice database;

[0042] Step S4-2: When an optimization cycle ends, retrieve the historical abnormal invoice database for that optimization cycle, obtain the review index of all abnormal invoices in the database, and use the review indexes that are less than the corresponding matching threshold as the abnormal invoices for the corresponding categories; according to the category, adjust the preset weight coefficient of the review index for the corresponding category based on the proportion of the abnormal index of each lifecycle in all abnormal invoices:

[0043] ;

[0044] Where w ir D represents the preset weighting coefficient of the review index for the i-th lifecycle in the r-th category. i This represents the number of anomalous indices in the i-th lifecycle. The number of anomalous indices representing all lifecycles.

[0045] A collaborative governance system for invoice information based on the entire lifecycle, comprising an audit data retrieval module, an audit category matching module, a review trigger judgment module, and an association weight optimization module;

[0046] The audit data retrieval module is used to retrieve audit data within the lifecycle of the current invoice, integrate the historical audit set with the audit set of the current invoice, and calculate the audit index;

[0047] The audit category matching module is used to extract the usage attributes of the audit data, divide the usage attribute range, obtain the matching threshold, and match the category of the audit data of the current invoice.

[0048] The review trigger judgment module is used to obtain the correlation degree according to the category of the review data, extract historical correlation degree data, obtain the review threshold, and determine whether the current invoice is to be reviewed.

[0049] The association weight optimization module is used to preset the optimization period, update the association degree, mark abnormal tickets and adjust the preset weight coefficient according to the review results within the optimization period;

[0050] The output of the audit data retrieval module is connected to the input of the audit category matching module; the output of the audit category matching module is connected to the input of the review trigger judgment module; and the output of the review trigger judgment module is connected to the input of the association weight optimization module.

[0051] The audit data retrieval module also includes a historical data integration unit and an audit index calculation unit;

[0052] The historical set integration unit is used to preset the statistical time, obtain the audit data of all bills within the life cycle within the statistical time and integrate them into a historical audit set, and calculate the frequency of occurrence of each audit data value to form a frequency set.

[0053] The audit index calculation unit is used to integrate the audit data within the life cycle of the current invoice into an audit set, obtain the intersection with the historical audit set, and sum the frequencies of audit data with the same life cycle within the intersection to obtain the audit index.

[0054] The audit category matching module further includes a matching threshold determination unit and a category matching execution unit;

[0055] The matching threshold determination unit is used to extract the review index of all tickets within the statistical time period, integrate it into a historical usage attribute set, divide the usage attribute intervals and calculate the coverage, and filter the corresponding intervals to determine the matching threshold.

[0056] The category matching execution unit is used to obtain the set of usage attributes of the current invoice and the corresponding matching threshold, and to match the category of the current invoice review data according to the number of times the review index is greater than the matching threshold.

[0057] The review trigger judgment module also includes a review threshold determination unit and a review status judgment unit;

[0058] The review threshold determination unit is used to calculate the correlation degree based on the category of the invoice review data, extract the correlation degree of all invoices within the statistical period and integrate them into a historical correlation degree set, and divide the interval to determine the review threshold.

[0059] The review status determination unit is used to calculate the relevance of the current invoice, compare it with the review threshold, and determine whether the review is triggered if the relevance of the current invoice is greater than or equal to the review threshold, and trigger the review if it is less than the review threshold.

[0060] The correlation weight optimization module also includes a review result processing unit and a weight coefficient adjustment unit;

[0061] The review result processing unit is used to distinguish between review compliance and review non-compliance. When the review is compliant, the correlation degree is updated and the historical correlation degree set is entered. When the review is non-compliant, the invoice is marked as an abnormal invoice and included in the historical abnormal invoice database.

[0062] The weight coefficient adjustment unit is used to retrieve the abnormal index from the historical abnormal invoice database at the end of an optimization cycle, and adjust the preset weight coefficient of the corresponding category audit index according to the proportion of abnormal index in each life cycle.

[0063] Compared with the prior art, the beneficial effects of the present invention are:

[0064] 1. This invention covers the entire lifecycle of invoice issuance, generation, circulation authorization, and invoice verification, integrates multi-dimensional audit data such as information matching rate and process completeness rate, and calculates an audit index based on historical data. It comprehensively captures the compliance characteristics of invoices at each stage, avoids the limitations of judgment in a single link, reduces external confirmation and communication costs, and effectively shortens the invoice processing cycle.

[0065] 2. This invention divides the usage attribute range, determines the matching threshold and review threshold, calculates the correlation of review data by category, and accurately determines the need for invoice review. This not only avoids meaningless repeated review and wastes resources, but also accurately identifies high-risk invoices with low correlation, thus greatly improving the pertinence and efficiency of invoice governance.

[0066] 3. This invention, by setting an optimization cycle, updates the correlation degree, marks abnormal tickets, and dynamically adjusts the weight coefficients based on the review results, enabling the governance mechanism to continuously adapt to changes in business scenarios, strengthen the ability to identify abnormal tickets, realize the self-iterative optimization of the governance logic, and ensure the accuracy and adaptability of long-term governance. Attached Figure Description

[0067] Figure 1 This is a flowchart illustrating a collaborative governance method for invoice information based on the entire lifecycle, as proposed in this invention.

[0068] Figure 2 This is a schematic diagram of the structure of a collaborative governance system for invoice information based on the entire lifecycle of the present invention. Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] Example 1: As Figure 1 As shown, this invention provides a technical solution: a collaborative governance method for invoice information based on the entire lifecycle. This method includes the following steps:

[0071] Step S1: Retrieve the audit data within the lifecycle of the current invoice. The lifecycle includes the issuance generation cycle, the circulation authorization cycle, and the invoice verification cycle. Each lifecycle includes several types of audit data.

[0072] Step S1 further includes:

[0073] The invoice is a fiscal electronic invoice, which contains several preset verification information items. The verification information is a structured field used to represent the content of the invoice. A preset standard library is used to match the verification information with data in the standard library to verify whether the invoice information is complete. The verification information is different in different lifecycles.

[0074] The core principle of matching is to use a pre-defined standardized rule system, i.e., a standard library, as a benchmark to perform multi-dimensional consistency and compliance verification of the target verification information in the invoice:

[0075] First, the standard library is broken down into structured benchmark rules that include data format, field connotation, logical constraints, and other dimensions. Then, the invoice verification information is standardized and transformed into a structured form that can be compared with the benchmark rules. Subsequently, through field-level semantic adaptation verification, format compliance verification, and logical constraint verification, the matching result is quantitatively output by the proportion of the number of information items that pass the verification to the total number of information items verified in the scenario. Essentially, it is an information validity verification process based on a unified benchmark.

[0076] The invoice generation cycle is the process from when the invoice issuer receives the invoice issuance request to when the invoice generation is completed.

[0077] The circulation authorization cycle is the process from the completion of the invoice generation to the time when the invoice user obtains the right to use the invoice through authorization. The use of the invoice includes the reimbursement of the invoice and the verification of the invoice.

[0078] The invoice verification period is the process from when the invoice user initiates the invoice usage request to when the invoice usage is completed;

[0079] The audit data includes the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i ;

[0080] The information matching rate X i This represents the percentage of complete verification information items in the i-th lifecycle out of all verification information items.

[0081] Each lifecycle contains several preset processes. Each time a ticket passes through a process, the ticket is marked; the process completion rate Y i This represents the percentage of the number of tags on the ticket within the i-th lifecycle in the total number of lifecycle processes.

[0082] The timeliness compliance rate Z i Z is the quantified value of the proportion of the time spent on a ticket in its i-th lifetime to the preset time limit for that lifetime. i =[1-(Z a / Z t )]×100%, where Z a Z represents the time elapsed during the lifecycle of the note. t This represents the preset time limit for this lifecycle, when Z a ≥Z t At that time, Z i Set to 0.

[0083] The methods for retrieving the audit data within the lifecycle of the current invoice in step S1 include:

[0084] Step S1-1: Preset the statistical time, obtain the audit data of all invoices within the lifecycle of the statistical time, and integrate the audit data into a historical audit set H. i H i H represents the historical review set for the i-th lifecycle; the historical review set H i It includes the verification information for each invoice, including the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i;

[0085] Step S1-2: Statistical analysis of historical audit set H i For each type of audit data, calculate the frequency of each value in the corresponding audit data to form a frequency set S. i ={P Eij}, where S i P represents the set of frequencies for the i-th lifetime. Eij This represents the frequency of the j-th value of the E-th type of audit data during the i-th lifecycle.

[0086] Steps S1-3: Integrate the audit data within the current invoice's lifecycle into an audit set R={X} i Y i Z i}; Retrieve the historical review set H i The intersection of the audit set R and the frequency set S is used to obtain the numerical values ​​of the audit data within the intersection. i The frequency of occurrence in each element is used as the intersection frequency. The intersection frequencies of elements with the same lifecycle are summed to obtain the current audit index L of the invoice. i0 L i L represents the review index for the i-th lifecycle. i0 The audit index represents the current audit index at the i-th stage of the bill's lifecycle.

[0087] Step S2: Extract the usage attributes of the audit data, and match the category of the audit data according to the usage attributes;

[0088] Step S2 further includes:

[0089] Step S2-1: Extract the review index of all invoices within the statistical period and integrate them into a historical usage attribute set U={U1, U2, ... U... n}, where U n U represents the usage attribute of the nth ticket within the statistical time period. n The system includes the review index of the nth invoice within the statistical period; all review indices in the historical usage attribute set U are evenly divided into M usage attribute intervals according to their lifecycles, the number of invoices in each usage attribute interval is counted, and the coverage C of each usage attribute interval is calculated. m =m0 / N, where C m This represents the coverage of the m-th attribute interval, m0 represents the number of tickets in the m-th attribute interval, and N represents the total number of tickets within the statistical time period.

[0090] Step S2-2: Obtain the matching threshold for the corresponding lifecycle based on the median of the usage attribute interval corresponding to the largest coverage value in the same lifecycle.

[0091] Step S2-3: Obtain the current ticket's usage attribute set U0={L i0}; Obtain the corresponding matching threshold T i T i The matching threshold represents the i-th lifecycle. Based on the number of documents with an approval index greater than the corresponding matching threshold, the documents are divided into several categories, and then based on L... i0 Greater than T i The number of items matches the category of the current invoice's audit data.

[0092] Step S3: Obtain the correlation between the audit data according to the category, as the correlation data; extract historical correlation data, and determine whether the current invoice needs to be reviewed based on the historical correlation data;

[0093] Step S3 further includes:

[0094] Step S3-1: Calculate the correlation degree based on the category of the audit data of the invoice;

[0095] ;

[0096] Among them G r L represents the correlation between the audit data of the r-th category. i w represents the review index for the i-th lifecycle. ir This represents the preset weight coefficient of the corresponding review index in the r-th category. The preset weight coefficients of different categories are independent of each other.

[0097] Step S3-2: Extract the correlation degree of all tickets within the statistical time period and integrate them into a historical correlation degree set G; divide the correlation degree in the historical correlation degree set G evenly into k correlation degree intervals, where M is the number of usage attribute intervals divided by the historical usage attribute set U, and k is rounded down; count the number of tickets in each correlation interval, take the correlation interval with the most tickets as the correlation threshold interval, and take the upper limit of the correlation threshold interval as the review threshold; if the correlation interval with the most tickets is not unique, take the upper limit of the correlation threshold interval with the largest upper limit as the review threshold.

[0098] Calculate the relevance of the current invoice. If the relevance of the current invoice is greater than or equal to the review threshold, it is determined that no review will be triggered; if the relevance of the current invoice is less than the review threshold, it is determined that a review will be triggered.

[0099] Step S4: Preset optimization period, and optimize the process of obtaining the correlation between audit data based on the review results within the optimization period;

[0100] Step S4 further includes:

[0101] Step S4-1: The review results include review compliance and review non-compliance;

[0102] When the review result is compliant, update the current relevance of the invoice to obtain the updated relevance, and record the updated relevance into the historical relevance set G:

[0103] ;

[0104] in Represents the updated relevance of the current invoice. This represents the pre-update relevance of the current invoice for review, where F represents the review threshold. Represents the preset update coefficient;

[0105] When the review result is non-compliant, the current invoice will be marked as an abnormal invoice and included in the historical abnormal invoice database;

[0106] Step S4-2: When an optimization cycle ends, retrieve the historical abnormal invoice database for that optimization cycle, obtain the review index of all abnormal invoices in the database, and use the review indexes that are less than the corresponding matching threshold as the abnormal invoices for the corresponding categories; according to the category, adjust the preset weight coefficient of the review index for the corresponding category based on the proportion of the abnormal index of each lifecycle in all abnormal invoices:

[0107] ;

[0108] Where w ir D represents the preset weighting coefficient of the review index for the i-th lifecycle in the r-th category. i This represents the number of anomalous indices in the i-th lifecycle. The number of anomalous indices representing all lifecycles.

[0109] For example:

[0110] Lifecycle segmentation:

[0111] Invoice generation cycle i=1, circulation authorization cycle i=2, invoice verification cycle i=3;

[0112] Taking the information matching rate of obtaining the invoice generation cycle as an example:

[0113] First, break down the standard library into structured baseline rules for data format, field meaning, and logical constraints:

[0114] Data format: For example, amounts should be rounded to two decimal places, and dates should be in YYYY-MM-DD format;

[0115] Field requirements: For example, the name of the invoice issuer must include the complete administrative region, organization name, industry, and organizational form;

[0116] Logical constraints: such as total price including tax = amount excluding tax + tax amount;

[0117] A pre-trained large language model is used to standardize the invoice verification information, followed by field-by-field verification.

[0118] Semantic compatibility verification: Confirm that the name of the invoice issuer after normalization is consistent with the name in the standard library;

[0119] Format compliance check: Amount and date formats meet the requirements;

[0120] Logical constraint verification: The total price including tax is logically matched with the itemized amounts;

[0121] If the total number of verification information items for this period is 20, and 19 items pass the verification, then the information matching rate X1 = 19 / 20 × 100% = 75%;

[0122] The historical audit sets H1, H2, and H3 each contain the information matching rate, process completeness rate, and timeliness compliance rate of several invoices.

[0123] Current invoice review set R:

[0124] R1 (i=1): X1=95%, Y1=90%, Z1=88%;

[0125] R2 (i=2): X2=92%, Y2=85%, Z2=91%;

[0126] R3 (i=3): X3=89%, Y3=93%, Z3=86%;

[0127] Forming a frequency set S i :

[0128] S1:P X11 =3.2%, P Y12 =2.8%, P Z13 =2.5%;

[0129] S2:P X21 =3.5%, P Y22 =2.6%, P Z23 =3.0%;

[0130] S3:P X31 =2.9%, P Y32 =3.1%, P Z33 =2.7%;

[0131] Calculate the current invoice review index L i0 :

[0132] L10 =3.2% + 2.8% + 2.5% = 8.5%;

[0133] L 20 =3.5% + 2.6% + 3.0% = 9.1%;

[0134] L 30 =2.9% + 3.1% + 2.7% = 8.7%;

[0135] Divide the attribute range and calculate the coverage:

[0136] The historical use of the attribute set U={U1,U2,...}, where each element of each set contains L1, L2, and L3 of the corresponding ticket;

[0137] The coverage calculation results are as follows: The lifecycle is evenly divided into 10 usage attribute intervals.

[0138] i=1: The coverage of the interval [8.0%, 8.5%) is C8=1200 / 10000=12%;

[0139] i=2: The coverage of the interval [8.5%, 9.0%) is C9=1300 / 10000=13%;

[0140] i=3: The coverage of the interval [8.2%, 8.7%) is C7=1150 / 10000=11.5%;

[0141] Determine the matching threshold Ti:

[0142] T1: The median of the interval [8.0%, 8.5%) is 8.25%;

[0143] T2: The median of the interval [8.5%, 9.0%) is 8.75%;

[0144] T3: The median of the interval [8.2%, 8.7%) is 8.45%;

[0145] Match the current ticket type:

[0146] The current ticket uses the attribute set U0={L 10 =8.5%, L 20 =9.1%, L 30 =8.7%};

[0147] Judgment result: L 10 >T1、L 20 >T2、L 30 T3, a total of 3 conditions are met;

[0148] Matching category: Category 3, Division rule: Meeting 0 is Category 1, meeting 1 - 2 is Category 2, meeting 3 is Category 3;

[0149] Calculate the current bill correlation degree G r :

[0150] For Category 3, that is, the preset weight coefficient when r = 3: w 13 = 0.35, w 23 = 0.35, w 33 = 0.30

[0151] The correlation degree G3 = w 13 ×L 10 +w 23 ×L 20 +w 33 ×L 30 = 8.74%;

[0152] Determine the review threshold and judge the review status:

[0153] The number of correlation degree intervals k = √M = √10 ≈ 3 (rounded down), √ represents the arithmetic square root operation;

[0154] The historical correlation degree set G is divided into 3 intervals: [7.0%, 7.8%), [7.8%, 8.6%), [8.6%, 9.4%);

[0155] The number of bills in each interval: [7.0%, 7.8%) → 2800 sheets, [7.8%, 8.6%) → 3200 sheets, [8.6%, 9.4%) → 4000 sheets;

[0156] The review threshold F = 9.4%, which is the upper limit value of the interval [8.6%, 9.4%) with the largest number of bills;

[0157] Judgment result: G3 = 8.74% < F = 9.4%, triggering a review;

[0158] Process the review result:

[0159] Review result: Review compliance;

[0160] The updated correlation degree G3` = 8.77%;

[0161] Enter G3` = 8.77% into the historical correlation degree set G;

[0162] Adjust the preset weight coefficient:

[0163] There are a total of 500 abnormal bills in the historical abnormal bill library during the optimization period;

[0164] Statistics of the abnormal index for Category 3: D1 = 120, D2 = 130, D3 = 110;

[0165] The total number of abnormal indices D = 120 + 130 + 110 = 360;

[0166] Adjusted weighting coefficient: w 13 `=0.33、w 23 `=0.36、w 33 =0.31.

[0167] Example 2: Figure 2 As shown, the present invention provides a system, a collaborative governance system for invoice information based on the entire lifecycle, which includes an audit data retrieval module, an audit category matching module, a review trigger judgment module, and an association weight optimization module;

[0168] The audit data retrieval module is used to retrieve audit data within the lifecycle of the current invoice, integrate the historical audit set with the audit set of the current invoice, and calculate the audit index;

[0169] The audit category matching module is used to extract the usage attributes of the audit data, divide the usage attribute range, obtain the matching threshold, and match the category of the audit data of the current invoice.

[0170] The review trigger judgment module is used to obtain the correlation degree according to the category of the review data, extract historical correlation degree data, obtain the review threshold, and determine whether the current invoice is to be reviewed.

[0171] The association weight optimization module is used to preset the optimization period, update the association degree, mark abnormal tickets and adjust the preset weight coefficient according to the review results within the optimization period;

[0172] The output of the audit data retrieval module is connected to the input of the audit category matching module; the output of the audit category matching module is connected to the input of the review trigger judgment module; and the output of the review trigger judgment module is connected to the input of the association weight optimization module.

[0173] The audit data retrieval module also includes a historical data integration unit and an audit index calculation unit;

[0174] The historical set integration unit is used to preset the statistical time, obtain the audit data of all bills within the life cycle within the statistical time and integrate them into a historical audit set, and calculate the frequency of occurrence of each audit data value to form a frequency set.

[0175] The audit index calculation unit is used to integrate the audit data within the life cycle of the current invoice into an audit set, obtain the intersection with the historical audit set, and sum the frequencies of audit data with the same life cycle within the intersection to obtain the audit index.

[0176] The audit category matching module further includes a matching threshold determination unit and a category matching execution unit;

[0177] The matching threshold determination unit is used to extract the review index of all tickets within the statistical time period, integrate it into a historical usage attribute set, divide the usage attribute intervals and calculate the coverage, and filter the corresponding intervals to determine the matching threshold.

[0178] The category matching execution unit is used to obtain the set of usage attributes of the current invoice and the corresponding matching threshold, and to match the category of the current invoice review data according to the number of times the review index is greater than the matching threshold.

[0179] The review trigger judgment module also includes a review threshold determination unit and a review status judgment unit;

[0180] The review threshold determination unit is used to calculate the correlation degree based on the category of the invoice review data, extract the correlation degree of all invoices within the statistical period and integrate them into a historical correlation degree set, and divide the interval to determine the review threshold.

[0181] The review status determination unit is used to calculate the relevance of the current invoice, compare it with the review threshold, and determine whether the review is triggered if the relevance of the current invoice is greater than or equal to the review threshold, and trigger the review if it is less than the review threshold.

[0182] The correlation weight optimization module also includes a review result processing unit and a weight coefficient adjustment unit;

[0183] The review result processing unit is used to distinguish between review compliance and review non-compliance. When the review is compliant, the correlation degree is updated and the historical correlation degree set is entered. When the review is non-compliant, the invoice is marked as an abnormal invoice and included in the historical abnormal invoice database.

[0184] The weight coefficient adjustment unit is used to retrieve the abnormal index from the historical abnormal invoice database at the end of an optimization cycle, and adjust the preset weight coefficient of the corresponding category audit index according to the proportion of abnormal index in each life cycle.

[0185] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A collaborative governance method for invoice information based on the entire lifecycle, characterized in that: The method includes the following steps: Step S1: Retrieve the audit data within the lifecycle of the current invoice. The lifecycle includes the issuance generation cycle, the circulation authorization cycle, and the invoice verification cycle. Each lifecycle includes several types of audit data. The methods for retrieving the audit data within the lifecycle of the current invoice in step S1 include: Step S1-1: Preset the statistical time, obtain the audit data of all invoices within the lifecycle of the statistical time, and integrate the audit data into a historical audit set H. i H i H represents the historical review set for the i-th lifecycle; the historical review set H i It includes the verification information for each invoice, including the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i ; Step S1-2: Statistical analysis of historical audit set H i For each type of audit data, calculate the frequency of each value in the corresponding audit data to form a frequency set S. i ={P Eij }, where S i P represents the set of frequencies for the i-th lifetime. Eij This represents the frequency of the j-th value of the E-th type of audit data during the i-th lifecycle. Steps S1-3: Integrate the audit data within the current invoice's lifecycle into an audit set R={X} i Y i Z i }; Retrieve the historical review set H i The intersection of the audit set R and the frequency set S is used to obtain the numerical values ​​of the audit data within the intersection. i The frequency of occurrence in each element is used as the intersection frequency. The intersection frequencies of elements with the same lifecycle are summed to obtain the current audit index L of the invoice. i0 L i L represents the review index for the i-th lifecycle. i0 The review index represents the current audit index at the i-th stage of the bill's lifecycle. Step S2: Extract the usage attributes of the audit data, and match the category of the audit data according to the usage attributes; Step S2 further includes: Step S2-1: Extract the review index of all invoices within the statistical period and integrate them into a historical usage attribute set U={U1, U2, ... U... n }, where U n U represents the usage attribute of the nth ticket within the statistical time period. n The system includes the review index of the nth invoice within the statistical period; all review indices in the historical usage attribute set U are evenly divided into M usage attribute intervals according to their lifecycles, the number of invoices in each usage attribute interval is counted, and the coverage C of each usage attribute interval is calculated. m =m0 / N, where C m This represents the coverage of the m-th attribute interval, m0 represents the number of tickets in the m-th attribute interval, and N represents the total number of tickets within the statistical time period. Step S2-2: Obtain the matching threshold for the corresponding lifecycle based on the median of the usage attribute interval corresponding to the largest coverage value in the same lifecycle. Step S2-3: Obtain the current ticket's usage attribute set U0={L i0 }; Obtain the corresponding matching threshold T i T i The matching threshold represents the i-th lifecycle. Based on the number of documents with an approval index greater than the corresponding matching threshold, the documents are divided into several categories, and then based on L... i0 Greater than T i The number of items matches the category of the current invoice's audit data; Step S3: Obtain the correlation between the audit data according to the category, as the correlation data; extract historical correlation data, and determine whether the current invoice needs to be reviewed based on the historical correlation data; Step S3 further includes: Step S3-1: Calculate the correlation degree based on the category of the audit data of the invoice; ； Among them G r L represents the correlation between the audit data of the r-th category. i w represents the review index for the i-th lifecycle. ir This represents the preset weight coefficient of the corresponding review index in the r-th category. The preset weight coefficients of different categories are independent of each other. Step S3-2: Extract the correlation degree of all tickets within the statistical time period and integrate them into a historical correlation degree set G; divide the correlation degree in the historical correlation degree set G evenly into k correlation degree intervals, where M is the number of usage attribute intervals divided by the historical usage attribute set U, and k is rounded down; count the number of tickets in each correlation interval, take the correlation interval with the most tickets as the correlation threshold interval, and take the upper limit of the correlation threshold interval as the review threshold; if the correlation interval with the most tickets is not unique, take the upper limit of the correlation threshold interval with the largest upper limit as the review threshold. Calculate the relevance of the current invoice. If the relevance of the current invoice is greater than or equal to the review threshold, it is determined that no review will be triggered; if the relevance of the current invoice is less than the review threshold, it is determined that a review will be triggered. Step S4: Preset optimization period, and optimize the process of obtaining the correlation between audit data based on the review results within the optimization period.

2. The method for collaborative governance of invoice information based on the entire lifecycle as described in claim 1, characterized in that: Step S1 further includes: The invoice is a fiscal electronic invoice, which contains several preset verification information items. The verification information is a structured field used to represent the content of the invoice. A preset standard library is used to match the verification information with data in the standard library to verify whether the invoice information is complete. The verification information is different in different lifecycles. The invoice generation cycle is the process from when the invoice issuer receives the invoice issuance request to when the invoice generation is completed. The circulation authorization cycle is the process from the completion of the invoice generation to the time when the invoice user obtains the right to use the invoice through authorization. The use of the invoice includes the reimbursement of the invoice and the verification of the invoice. The invoice verification period is the process from when the invoice user initiates the invoice usage request to when the invoice usage is completed; The audit data includes the information matching rate X. i Process completeness rate Y i With timeliness compliance rate Z i ; The information matching rate X i This represents the percentage of complete verification information items in the i-th lifecycle out of all verification information items. Each lifecycle contains several preset processes. Each time a ticket passes through a process, the ticket is marked; the process completion rate Y i This represents the percentage of the number of tags on the ticket within the i-th lifecycle in the total number of lifecycle processes. The timeliness compliance rate Z i Z is the quantified value of the proportion of the time spent on a ticket in its i-th lifetime to the preset time limit for that lifetime. i =[1-(Z a / Z t )]×100%, where Z a Z represents the time elapsed during the lifecycle of the note. t This represents the preset time limit for this lifecycle, when Z a ≥Z t At that time, Z i Set to 0.

3. The method for collaborative governance of invoice information based on the entire lifecycle as described in claim 2, characterized in that: Step S4 further includes: Step S4-1: The review results include review compliance and review non-compliance; When the review result is compliant, update the current relevance of the invoice to obtain the updated relevance, and record the updated relevance into the historical relevance set G: ； in Represents the updated relevance of the current invoice. This represents the pre-update relevance of the current invoice for review, where F represents the review threshold. Represents the preset update coefficient; When the review result is non-compliant, the current invoice will be marked as an abnormal invoice and included in the historical abnormal invoice database; Step S4-2: When an optimization cycle ends, retrieve the historical abnormal invoice database for that optimization cycle, obtain the review index of all abnormal invoices in the database, and use the review indexes that are less than the corresponding matching threshold as the abnormal invoices for the corresponding categories; according to the category, adjust the preset weight coefficient of the review index for the corresponding category based on the proportion of the abnormal index of each lifecycle in all abnormal invoices: ； Where w ir D represents the preset weighting coefficient of the review index for the i-th lifecycle in the r-th category. i This represents the number of anomalous indices in the i-th lifecycle. The number of anomalous indices representing all lifecycles.

4. A collaborative governance system for invoice information based on the entire lifecycle, applied to the collaborative governance method for invoice information based on the entire lifecycle as described in any one of claims 1-3, characterized in that: The system includes a review data retrieval module, a review category matching module, a review trigger judgment module, and an association weight optimization module; The audit data retrieval module is used to retrieve audit data within the lifecycle of the current invoice, integrate the historical audit set with the audit set of the current invoice, and calculate the audit index; The audit category matching module is used to extract the usage attributes of the audit data, divide the usage attribute range, obtain the matching threshold, and match the category of the audit data of the current invoice. The review trigger judgment module is used to obtain the correlation degree according to the category of the review data, extract historical correlation degree data, obtain the review threshold, and determine whether the current invoice is to be reviewed. The association weight optimization module is used to preset the optimization period, update the association degree, mark abnormal tickets and adjust the preset weight coefficient according to the review results within the optimization period; The output of the audit data retrieval module is connected to the input of the audit category matching module; the output of the audit category matching module is connected to the input of the review trigger judgment module; and the output of the review trigger judgment module is connected to the input of the association weight optimization module.

5. A collaborative governance system for invoice information based on the entire lifecycle, as described in claim 4, characterized in that: The audit data retrieval module also includes a historical data integration unit and an audit index calculation unit; The historical set integration unit is used to preset the statistical time, obtain the audit data of all bills within the life cycle within the statistical time and integrate them into a historical audit set, and calculate the frequency of occurrence of each audit data value to form a frequency set. The audit index calculation unit is used to integrate the audit data within the life cycle of the current invoice into an audit set, obtain the intersection with the historical audit set, and sum the frequencies of audit data with the same life cycle within the intersection to obtain the audit index.

6. The collaborative governance system for invoice information based on the entire lifecycle as described in claim 4, characterized in that: The audit category matching module further includes a matching threshold determination unit and a category matching execution unit; The matching threshold determination unit is used to extract the review index of all tickets within the statistical time period, integrate it into a historical usage attribute set, divide the usage attribute intervals and calculate the coverage, and filter the corresponding intervals to determine the matching threshold. The category matching execution unit is used to obtain the set of usage attributes of the current invoice and the corresponding matching threshold, and to match the category of the current invoice review data according to the number of times the review index is greater than the matching threshold.

7. A collaborative governance system for invoice information based on the entire lifecycle, as described in claim 4, is characterized in that: The review trigger judgment module also includes a review threshold determination unit and a review status judgment unit; The review threshold determination unit is used to calculate the correlation degree based on the category of the invoice review data, extract the correlation degree of all invoices within the statistical period and integrate them into a historical correlation degree set, and divide the interval to determine the review threshold. The review status determination unit is used to calculate the relevance of the current invoice, compare it with the review threshold, and determine whether the review is triggered if the relevance of the current invoice is greater than or equal to the review threshold, and trigger the review if it is less than the review threshold. The correlation weight optimization module also includes a review result processing unit and a weight coefficient adjustment unit; The review result processing unit is used to distinguish between review compliance and review non-compliance. When the review is compliant, the correlation degree is updated and the historical correlation degree set is entered. When the review is non-compliant, the invoice is marked as an abnormal invoice and included in the historical abnormal invoice database. The weight coefficient adjustment unit is used to retrieve the abnormal index from the historical abnormal invoice database at the end of an optimization cycle, and adjust the preset weight coefficient of the corresponding category audit index according to the proportion of abnormal index in each life cycle.

Images