CN113793208A - Small-amount financial debt dispute smart litigation system based on block chain - Google Patents
Small-amount financial debt dispute smart litigation system based on block chain Download PDFInfo
- Publication number
- CN113793208A CN113793208A CN202111022109.1A CN202111022109A CN113793208A CN 113793208 A CN113793208 A CN 113793208A CN 202111022109 A CN202111022109 A CN 202111022109A CN 113793208 A CN113793208 A CN 113793208A
- Authority
- CN
- China
- Prior art keywords
- litigation
- loan
- data
- node
- unit
- 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.)
- Granted
Links
- 239000000463 material Substances 0.000 claims abstract description 62
- 238000003860 storage Methods 0.000 claims abstract description 39
- 238000013210 evaluation model Methods 0.000 claims abstract description 30
- 238000012795 verification Methods 0.000 claims abstract description 28
- 210000002569 neuron Anatomy 0.000 claims description 27
- 238000013145 classification model Methods 0.000 claims description 12
- 238000003062 neural network model Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000007726 management method Methods 0.000 claims description 10
- 230000014759 maintenance of location Effects 0.000 claims description 9
- 238000006467 substitution reaction Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 6
- 238000010191 image analysis Methods 0.000 claims description 6
- 238000000034 method Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000012544 monitoring process Methods 0.000 claims description 5
- 230000001360 synchronised effect Effects 0.000 claims 2
- 230000000694 effects Effects 0.000 abstract description 5
- 238000011084 recovery Methods 0.000 abstract description 4
- 230000006870 function Effects 0.000 description 6
- 238000011161 development Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
Images
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Computational Linguistics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Finance (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Technology Law (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Accounting & Taxation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Development Economics (AREA)
- Human Resources & Organizations (AREA)
- Primary Health Care (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention relates to the technical field of block chains, in particular to a small-amount financial debt dispute intelligent litigation system based on a block chain, which comprises a business subsystem and a supervision node, wherein the supervision node is arranged in a financial supervision institution, the business subsystem comprises a synchronization unit, a verification unit, a certificate-storing notarization unit, a storage unit and an litigation management unit, loan business data comprises loan application materials, approval data, loan data and repayment data, the certificate-storing notarization unit and the notarization certificate are obtained, a litigation form is generated according to a template, the loan business data, related certificate-storing notarization certificates and notarization certificates are packaged into an evidence package, the evidence package and the litigation form are submitted to the supervision node, the output of a difficulty evaluation model is a litigation cost score, the supervision node feeds the litigation cost score back to a bank node, and the litigation management unit progresses to a court system periodically and synchronously. The substantial effects of the invention are as follows: the preparation efficiency of litigation materials is improved, and the recovery rate of overdue funds is improved.
Description
Technical Field
The invention relates to the technical field of block chains, in particular to a small-amount financial debt dispute intelligent litigation system based on the block chains.
Background
The small and micro enterprises are the main channel for providing new employment posts, the main platform for entrepreneurial growth of the enterprises and the important force of technological innovation. Through diversified short-term loan service, the method is helpful for small and micro enterprises to enhance the risk resistance, grasp market opportunities in time and grasp development opportunities. But the risk assessment difficulty of small and micro enterprises is large, the bank loan risk is large, and the management cost is large. With the further development of economy, the situation that the business volume of the small loan rises exists, and meanwhile, the overdue rate and the bad account rate also tend to rise. Not only brings loss to the banking industry, but also integrally reduces the credit degree of the small and micro enterprises, and is finally not beneficial to the development of the small and micro enterprises. The deterrence force of judicial collection is not high due to low involved amount of overdue and bad account small loan, high collection urging cost and difficult judicial treatment.
Chinese patent CN109767193A, published 2019, 5, month 17, a method, equipment and readable storage medium for applying liability insurance to litigation property insurance, the method comprises the following steps: after the litigation property insurance applicant successfully logs in the insurance system, detecting whether an electronic application material submitted by the applicant is received; if the electronic application material is received and the electronic application material is detected to meet the corresponding preset condition, generating an electronic contract corresponding to the litigation property security liability risk, and outputting signing information to prompt the applicant to sign in the electronic contract; when a signature instruction for signing the electronic contract is detected, and the applicant is detected to pay corresponding insurance of the litigation property security liability insurance, outputting first prompt information to prompt the applicant that the litigation property security liability insurance is successfully applied. The technical scheme improves the insurance application efficiency of the litigation property insurance and reduces the insurance application duration of the litigation property insurance. The efficiency of litigation is improved to a certain extent, but the problem of the efficiency of intelligent litigation of a large number of small-amount financial debt rights disputes is still not solved.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: at present, the technical problem of lack of a preparation system for batch completion of low-volume financial debt right dispute litigation materials is solved. The small-amount financial-debt dispute litigation intelligent litigation system based on the block chain is provided, and the small-amount financial-debt dispute litigation material can be efficiently and quickly prepared.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: the utility model provides a small amount finance debt right dispute wisdom lawsuit system based on block chain, includes business subsystem and supervisory node, supervisory node sets up at financial supervisory institution, the business subsystem includes synchronization unit, verification unit, deposit the justice unit, memory cell and litigation administrative unit, synchronization unit synchronization loan service data, loan service data includes loan application material, examination and approval data, the data of paying for, when synchronization unit obtains loan service data, submits loan service data for verification unit and carries out integrality and correctness verification, if verification fails, then the verification unit sends out the warning, verifies by the business handling personnel and handles to submit loan service data by synchronization unit again, if verify then submit loan service data for deposit the justice unit and carry out deposit the justice and the justice, obtaining a certificate of storage and a certificate of notarization, submitting the loan service data, the certificate of storage and the certificate of notarization to a storage unit for storage, when a user does not pay according to an agreed repayment plan and causes overdue, a synchronization unit informs a litigation management unit, the litigation management unit calls the loan service data, generates a litigation form according to a template, packages the loan service data, related certificate of storage and notarization certificate into an evidence package, submits the evidence package and the litigation form to a supervision node, the supervision node verifies the authenticity and integrity of the evidence package and then verifies the compliance of the loan service, if the verification fails, returns the evidence package and the litigation form, if the verification passes, a difficulty evaluation model is established, the input of the difficulty evaluation model is financial service data and financial asset data of the loan user, and the output of the difficulty evaluation model is litigation cost score, the litigation cost score is high, which means that litigation cost of finally recovering overdue funds is high, the supervision node feeds the litigation cost score back to the bank node, the supervision node submits the evidence package, the prosecution state and the litigation cost score to the court system, and the litigation management unit periodically synchronizes the case progress to the court system.
As preferred, the unit of checking establishes the standard catalogue of loan application material, the unit of checking is with loan application material association standard catalogue, and when the entry in the standard catalogue all had associated loan application material, it is complete to judge loan application material, the unit of checking collects and stores behind sufficient quantity's loan application material, establishes classification model to every material category, classification model's input is the data with the loan application material of material category, and classification model establishes the back, with newly-increased loan application material input classification model, if newly-increased loan application material all is greater than preset threshold with the categorised distance of existing, then the suggestion corresponding loan application material is unusual, by the manual verification.
Preferably, the verification unit establishes an image analysis model for image data in the loan application material, the image analysis model extracts image data in the historical loan application material, extracts image parts with the same historical image data to be used as template regions, compares the image data in the newly added loan application material with the template regions, and if the similarity is lower than a preset threshold, prompts that the corresponding image data is abnormal and is checked manually.
Preferably, the storage unit comprises a plurality of storages, the bank node allocates a unique identifier for the certificate storage packet and a user identifier for the loan application user, the bank node divides the certificate storage packet into a plurality of sub-data packets, the sub-data packets have preset sizes, the sub-data packets are associated with the identifier and the user identifier, and the sub-data packets are randomly sent to the storages for storage.
Preferably, the supervisory node receives the service data reported by the financial institution, the supervisory node develops a corresponding storage table for the service data according to the field structure of the service data, the supervision node establishes a non-numerical value field substitution number table, the financial institution substitutes the non-numerical value field with a substitution number according to the substitution number table, the supervision node discloses an external main key field, the financial institution establishes two copies for a service data line, the field except the external main key field is divided into two addends which are respectively distributed to two copies for storage, the two copies both store the real value of the external main key field, one copy is used as a cooperation copy to be sent to the supervision node for storage, the other copy is used as a reserved copy to be stored in the financial institution, and when the supervision node runs the difficulty evaluation model, establishing safe multi-party calculation with the financial institution to obtain the output of the difficulty evaluation model.
Preferably, the difficulty evaluation model is a neural network model, the supervisory node splits the neural network model into a plurality of submodels and a master model, the number of the submodels is the same as that of neurons in the layer 1, the input of the submodels is input layer neurons connected with the neurons in the layer 1, the output of the submodels is the input number of the neurons in the layer 1, the input layer of the neural network model is deleted, the input of the neurons in the layer 1 is modified into the output of corresponding submodels to serve as the master model, the supervisory node establishes safe multi-party computation for each submodel and a financial institution to obtain the output of the submodels, and the supervisory node substitutes the output of the submodels into the master model to obtain the output of the master model, namely the lition cost score.
Preferably, when the supervisory node establishes the secure multi-party computation with the financial institution for each sub-model, the supervisory node informs the financial institution to check whether the service data is the latest data, if not, the variable quantity of the service data is computed, the variable quantity is superposed on the reserved copy, and the updated reserved copy participates in the secure multi-party computation.
Preferably, the supervisory node splits each connection of the submodel into two connections, the two connections are respectively marked as a cooperative connection and a reserved connection, weights of the cooperative connection and the reserved connection are respectively marked as a cooperative weight and a reserved weight, the financial institution randomly generates the reserved weight, sends the updated product of the value of the field corresponding to the reserved copy and the reserved weight to the supervisory node, calculates a ratio k of the value in the reserved copy in the corresponding field to the value in the cooperative copy, and sends the ratio k to the supervisory node, the supervision node calculates and obtains the cooperation weight according to the weight of the original connection, the ratio k and the retention weight, so that the ratio k, the cooperation weight and the retention weight form an equivalent weight equal to the weight of the original connection, and the supervision node obtains the output of the sub-model according to the product of the value of the corresponding field of the cooperative copy, the cooperative weight and the report of the financial institution.
The substantial effects of the invention are as follows: the loan service data is stored in a chain of blocks in time through the service subsystem, notarization is carried out through a notarization place to finish the collection of evidences, and the preparation of litigation materials is finished quickly by generating a litigation state through a template, so that the preparation efficiency of the litigation materials is improved; the litigation materials are verified through the supervision nodes, the integrity of the litigation materials is improved, the loan application materials are audited through the verification unit, the small-amount financial debt disputes are simplified and shunted through litigation cost grading, and finally the recovery rate of overdue funds is improved.
Drawings
FIG. 1 is a schematic diagram of a system for litigation.
Fig. 2 is a schematic view of a service data storage according to an embodiment.
Fig. 3 is a schematic diagram of a difficulty evaluation model according to an embodiment.
FIG. 4 is a diagram illustrating a sub-model according to an embodiment.
FIG. 5 is a diagram illustrating the execution of a second sub-model according to an embodiment.
Wherein: 10. the system comprises a business subsystem, 11, a synchronization unit, 12, a verification unit, 13, a certificate storage notarization unit, 14, a storage unit, 15, a litigation management unit, 20, a supervision node, 141, a business data line, 142, a addend, 143, a copy, 21, a main model, 22 and a sub model.
Detailed Description
The following provides a more detailed description of the present invention, with reference to the accompanying drawings.
The first embodiment is as follows:
a small amount of financial debt right dispute wisdom lawsuit system based on a block chain is disclosed, please refer to an attached figure 1, and comprises a service subsystem 10 and a supervision node 20, wherein the supervision node 20 is arranged in a financial supervision institution, the service subsystem 10 comprises a synchronization unit 11, a verification unit 12, a credit evidence storage unit 13, a storage unit 14 and a lawsuit management unit 15, the synchronization unit 11 synchronizes loan service data, the loan service data comprises loan application materials, approval data, loan payment data and repayment data, when the synchronization unit 11 obtains the loan service data, the loan service data is submitted to the verification unit 12 for integrity and correctness verification, if the verification fails, the verification unit 12 sends an alarm, service handling personnel verify the processing, the synchronization unit 11 submits the loan service data again, if the verification passes, the loan service data is submitted to the credit evidence storage public verification unit 13 for verification and public verification, obtaining a certificate of storage and a certificate of notarization, submitting the loan service data, the certificate of storage and the certificate of notarization to a storage unit 14 for storage, when the user does not pay according to the agreed repayment plan and causes overdue, a synchronization unit 11 informs a litigation management unit 15, the litigation management unit 15 calls the loan service data, generates a litigation state according to a template, packages the loan service data, the related certificate of storage and the certificate of notarization into an evidence package, submits the evidence package and the litigation state to a supervision node 20, the supervision node 20 verifies the authenticity and integrity of the evidence package and then verifies the compliance of the loan service, if the verification fails, the evidence package and the litigation state are returned, if the verification passes, a difficulty evaluation model is established, the input of the difficulty evaluation model is the financial service data and the financial data of the loan user, the output of the difficulty evaluation model is a litigation cost score, the high price score indicates that the overdue capital is finally recovered, the supervision node 20 feeds back the litigation cost scores to the bank nodes, the supervision node 20 submits the evidence packages, the complaints and the litigation cost scores to the court system, and the litigation management unit 15 periodically synchronizes the case progress to the court system. The specific difficulty evaluation model is that a bank formulates rules according to business requirements, the rules are converted into functions, and the functions are fitted by using a neural network model. Theoretically neural network models can fit arbitrary functions. Thus, only the monitoring node 20 can contact the financial service data of the loan user, and the financial institution originally needs to report the related financial service data to the monitoring institution, so that the disclosure of data privacy is avoided. The supervisory node 20 derives litigation cost scores based solely on the financial services data. And guiding the bank to carry out litigation work by the litigation cost score. The litigation cost of the bank can be reduced, and the recovery rate of overdue money can be improved. And further, effective deterrence is formed for the loan users, and the loan users are promoted to make full and judicious decisions when applying for loans and expanding the operation scale. The loan behavior of the loan applicant is reduced for those products that do not have advantages. The bank loan is more directed to those loan application persons with product advantages and development prospects. Eventually promoting the development of economy.
The checking unit 12 establishes a standard catalog of loan application materials, the checking unit 12 associates the loan application materials with the standard catalog, when the entries in the standard catalog all have associated loan application materials, the loan application materials are judged to be complete, after the checking unit 12 collects and stores enough loan application materials, a classification model is established for each material category, the input of the classification model is the data of the loan application materials of the same material category, after the classification model is established, the newly added loan application materials are input into the classification model, if the distances between the newly added loan application materials and the existing classifications are larger than a preset threshold, the corresponding loan application materials are prompted to be abnormal and are manually checked.
The checking unit 12 establishes an image analysis model for the image data in the loan application material, the image analysis model extracts the image data in the historical loan application material, extracts the image part with the same historical image data as the template region, compares the image data in the newly added loan application material with the template region, and if the similarity is lower than a preset threshold, prompts that the corresponding image data is abnormal and is checked manually.
The credit card saving system comprises a storage unit 14, a bank node, a plurality of sub-data packets and a plurality of memory units, wherein the bank node allocates a unique identifier for a credit card saving packet and a user identifier for a user applying for a loan, the bank node divides the credit card saving packet into the plurality of sub-data packets, the sub-data packets have preset sizes, associates the identifier with the user identifier, and randomly sends the plurality of sub-data packets to the plurality of memory units for storage.
The supervisory node 20 receives the service data reported by the financial institution, please refer to fig. 2, the supervisory node 20 creates a corresponding storage table for the service data according to the field structure of the service data, the supervisory node 20 creates a non-numerical field substitution number table, the financial institution replaces the non-numerical field with a substitution number according to the substitution number table, the supervisory node 20 discloses an external main key field, the financial institution creates two copies 143 for the service data line 141, the field except the external main key field is split into two addends 142, which are respectively allocated to the two copies 143 for storage, the two copies 143 both store the true value of the external main key field, one copy 143 is sent to the supervisory node 20 as a cooperative copy 143 for storage, the other copy 143 is stored in the financial institution as a reserved copy 143, when the supervisory node 20 runs the difficulty evaluation model, the supervisory node 20 establishes a secure multi-party calculation with the financial institution, and obtaining the output of the difficulty evaluation model.
The difficulty evaluation model is a neural network model, the supervision node 20 divides the neural network model into a plurality of sub-models 22 and a main model 21, please refer to fig. 3 and fig. 4, the number of the sub-models 22 is the same as that of neurons in layer 1, the input of the sub-models 22 is input layer neurons connected with the neurons in layer 1, the output of the sub-models 22 is the input number of the neurons in layer 1, the input layer of the neural network model is deleted, the input of the neurons in layer 1 is modified into the output of the corresponding sub-models 22 to serve as the main model 21, the supervision node 20 establishes safe multi-party calculation for each sub-model 22 and a financial institution to obtain the output of the sub-model 22, and the supervision node 20 substitutes the output of the sub-model 22 into the main model 21 to obtain the output of the main model 21, namely the litigation cost score.
In the difficulty evaluation model, the input layer has three neurons, which respectively correspond to the bank deposit balance, the monthly average consumption amount and the monthly average consumption frequency, the first layer of neurons has two neurons, one of the neurons is connected with the three neurons of the input layer, the excitation function is a sigmod function, the weights are represented by a11, a12 and a13, the offset is represented by b1, and the output is equal to sigmod (x), wherein the first layer of neurons is fully connected, and x = a11 bank deposit balance + a12 monthly average consumption amount + a13 monthly average consumption frequency + b 1. The financial data generated by the financial institution is specifically: the balance of the bank deposit is 33 ten thousand, the average monthly consumption amount is 1 ten thousand, and the average monthly consumption frequency is 16 times.
Wherein generating 2 addends 142 for the bank deposit balance respectively is: 33=12+21, the 2 copies 143 are assigned the following values: 12 and 21. The average monthly consumption amount of 1 ten thousand generates 2 addends 142: 10,000.00=6,000.00+4,000.00, the 2 copies 143 being assigned the following values: 6,000.00, and 4,000.00. The generated addend 142 for the average monthly consumption frequency 16 is: 16=13+3, and the 2 copies 143 are assigned the following values: 13 and 3.
Assume that the collaborative copy 143 stores data that is: 12,6,000.00,13, the supervisory node 20 calculates the sum as: a11 + a12 6,000.00+ a13 13, summed with the financial institution, and summed to give: a11 (12 + 21) + a12 (6,000.00 +4,000.00) + a13 (13 + 3). Namely: a11 x 33+ a12 x 10,000.00+ a13 x 16, which is exactly equal to the result of directly substituting the original real value into the weighted sum formula. And adding the offset value b1 to obtain the value of x, and substituting the value into a sigmod (x) function to obtain the output of the neuron. In the calculation process, the original real value is mixed in the multiple confusion values and the addend 142, so that the original real value is hidden and is difficult to be accurately found, and the privacy and the safety of data are improved.
When the supervisory node 20 establishes secure multiparty computation with the financial institution for each sub-model 22, the financial institution is notified to check whether the service data is the latest data, if not, the variation of the service data is computed, the variation is superimposed on the reserved copy 143, and the updated reserved copy 143 participates in the secure multiparty computation. If the difficulty evaluation model is executed, the deposit of the user is changed to 30, that is, the deposit is reduced by 3 ten thousand, the balance of the bank deposit in the reserved copy 14323 is subtracted from 21 by 3, and the result 19 is substituted into the difficulty evaluation model for execution.
In practice, a user usually sets up savings and deposit accounts in multiple banks, so that data reported to a supervising authority by a financial institution needs to be substituted into a difficulty evaluation model by balances of multiple bank accounts of the same user. If the user has enough deposits and the consumption amount and the consumption times are more, the repayment capability of the user is stronger, and the overdue loan is only caused by weak repayment willingness, so the difficulty of recovering the overdue amount through a judicial approach is lower. If the user has less deposits and less consumption, the overdue loan of the user may be caused by lack of funds, and even though the user passes through the judicial approach, the user has insufficient property, so that the overdue amount cannot be effectively recovered. And the user may intentionally delay or not cooperate with judicial activities due to insufficient property, so that the judicial difficulty is higher, and the personnel and cost for inputting and tracking cases by banks are higher. Only in the case that the bank is involved with a few cases and the personnel is sufficient, the cases are advanced.
The beneficial technical effects of this embodiment are: the loan service data is stored in a chain of blocks in time through the service subsystem 10, notarization is carried out through a notarization place to finish the collection of evidences, a litigation state is generated through a template, the preparation of litigation materials is finished quickly, and the preparation efficiency of the litigation materials is improved; the litigation materials are verified through the supervision node 20, the integrity of the litigation materials is improved, the loan application materials are verified through the verification unit 12, small financial debt disputes are simplified and shunted through litigation cost grading, and finally the recovery rate of overdue funds is improved.
Example two:
a small amount financial debt weight dispute intelligent litigation system based on a block chain is disclosed, please refer to the attached figure 5, a supervision node 20 divides each connection of a sub-model 22 into two connections, the two connections are respectively marked as a cooperative connection and a reserved connection, weights of the cooperative connection and the reserved connection are respectively marked as a cooperative weight and a reserved weight, financial institutions randomly generate reserved weights, products of values of fields corresponding to updated reserved copies 143 and the reserved weights are sent to the supervision node 20, a ratio k of values of the reserved copies 143 in the corresponding fields and values of the cooperative copies 143 is calculated, the ratio k is sent to the supervision node 20, the supervision node 20 calculates and obtains the cooperative weights according to the weights, the ratio k and the reserved weights of original connections, equivalent weights formed by the ratio k, the cooperative weights and the reserved weights are equal to weights of the original connections, and the supervision node 20 calculates and obtains the cooperative weights according to the values of the fields corresponding to the cooperative copies 143, and, The product of the collaborative weight and the financial institution's reported results in the output of the submodel 22.
Taking the 1 st neuron at layer 1 as an example, it involves 3 connections. The 3 neurons of the input layer are respectively connected, the corresponding weights are w11, w12 and w13, and the 3 connections involved in the connection are respectively split into 2 connections. As shown in fig. 5, 3 connections involved in the 1 st neuron of layer 1 are split into a reserved connection and a cooperative connection. So that the connections of the 1 st neuron of level 1 are changed from 3 to 6, the corresponding output y1= Sigmod (Σ w1_ ri × xi _ r + ∑ w1_ ci × xi _ c + b 1).
For convenience of expression, a scaling factor q is set, q represents the ratio of the value in the cooperative copy 143 to the original value, then the ratio k = (1-q)/q, when splitting into the reserved connection and the cooperative connection, the number of 1 st neurons participating in level 1 of the cooperation weight w1_ c1, the scaling factor q and the reserved weight w1_ r1 is equal to x1_ r 68692 × w _ 7378 _ 6 + x1_ c1 × 1_ c 1= (1-q) × 1 × 1_ r1 + q × 1 × 1_ c1, the equivalent weight is (1-q) × 1_ r1 + q × w1_ c1, and the equivalent weight is equal to the weight w11 of the original connection, that is: w11= (1-q) × w1_ r1 + q × w1_ c 1. Wherein the proportionality coefficient q is obtained by calculation by the financial institution, and the cooperation weight w1_ c1 is generated by the supervisory node 20. The retention weight w1_ r1 is calculated from the equation w11= (1-q) × w1_ r1 + q × w1_ c 1. Since the financial institution does not know the collaboration weight w1_ c1, the equivalent weight, namely the weight w11 of the original connection, cannot be solved, so that the connection weight of the difficulty evaluation model remains secret. For the calculation of the reserved weight, only addition and multiplication are involved, and homomorphic encryption techniques of addition and multiplication belong to the prior art in the field, and are not described herein again.
In this embodiment, when the financial institution assists the supervisory node 20 to execute the sub-model 22, the specific connection weight of the sub-model 22 cannot be known, and the sub-model 22 is protected. And the leakage of the difficulty evaluation model is avoided. If the difficulty evaluation model is leaked, the problem that part of the applicants control data maliciously aiming at the difficulty evaluation model can be brought, and the effect of the difficulty evaluation model is reduced. The above-described embodiments are only preferred embodiments of the present invention, and are not intended to limit the present invention in any way, and other variations and modifications may be made without departing from the spirit of the invention as set forth in the claims.
Claims (8)
1. A small amount financial debt right dispute intelligent litigation system based on a block chain is characterized in that,
the loan transaction data is submitted to the checking unit for integrity and correctness checking when the synchronous unit obtains the loan transaction data, if the checking fails, the checking unit sends an alarm, the service handling personnel checks and processes the loan transaction data and submits the loan transaction data again by the synchronous unit, if the checking passes, the loan transaction data is submitted to the verification and notarization unit for verification and notarization, a certificate of deposit and a notarization is obtained, and the loan transaction data, the certificate of deposit and the notarization are submitted to the storage unit for storage, when the user does not repay according to the agreed repayment plan and overdue is caused, the synchronizing unit informs the litigation managing unit, the litigation managing unit calls the loan service data, generates a litigation form according to the template, packages the loan service data, the related certificate of deposit and the notarization certificate into an evidence package, submits the evidence package and the litigation form to the supervising node,
the supervision node verifies authenticity and integrity of the evidence package, then verifies compliance of loan service, returns the evidence package and litigation state if verification fails, establishes a difficulty evaluation model if verification passes, inputs of the difficulty evaluation model are financial service data and financial asset data of loan users, outputs of the difficulty evaluation model are litigation cost scores, the high litigation cost scores represent high litigation cost for finally retrieving overdue funds, the supervision node feeds the litigation cost scores back to the bank node, the supervision node submits the evidence package, the litigation state and the litigation cost scores to a court system, and the litigation management unit progresses to the court system periodically and synchronously.
2. The system of claim 1, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
the checking unit establishes a standard catalogue of loan application materials, the checking unit associates the loan application materials with the standard catalogue, and when the items in the standard catalogue have associated loan application materials, the integrity of the loan application materials is judged,
the checking unit collects and stores sufficient loan application materials, establishes classification models for each material category, inputs the classification models for the loan application materials of the same material category, inputs the newly added loan application materials into the classification models after the classification models are established, and prompts the corresponding loan application materials to be abnormal if the distance between the newly added loan application materials and the existing classification is greater than a preset threshold value, and the loan application materials are manually checked.
3. The system of claim 2, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
the checking unit establishes an image analysis model for image data in the loan application material, the image analysis model extracts the image data in the historical loan application material, extracts the image part with the same historical image data, and is used as a template region to compare the image data in the newly added loan application material with the template region, if the similarity is lower than a preset threshold, the corresponding image data is prompted to be abnormal, and the image data is manually checked.
4. The system of claim 1, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
the credit card saving system comprises a plurality of storages of the storage unit, bank nodes distribute unique identification for credit card saving packages, user identification for users applying for loans, the bank nodes divide the credit card saving packages into a plurality of sub data packages, the sub data packages are preset in size, the sub data packages are associated with the identification and the user identification, and the sub data packages are randomly sent to the storages to be stored.
5. The system according to any one of claims 1 to 4, wherein the system comprises a plurality of intelligent litigation systems,
the supervisory node receives the business data reported by the financial institution, the supervisory node develops a corresponding storage table for the business data according to the field structure of the business data, the supervision node establishes a non-numerical value field substitution number table, the financial institution substitutes the non-numerical value field with a substitution number according to the substitution number table, the supervision node discloses an external main key field, the financial institution establishes two copies for a service data line, the field except the external main key field is divided into two addends which are respectively distributed to two copies for storage, the two copies both store the real value of the external main key field, one copy is used as a cooperation copy to be sent to the supervision node for storage, the other copy is used as a reserved copy to be stored in the financial institution, and when the supervision node runs the difficulty evaluation model, establishing safe multi-party calculation with the financial institution to obtain the output of the difficulty evaluation model.
6. The system of claim 5, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
the difficulty evaluation model is a neural network model, the monitoring node divides the neural network model into a plurality of submodels and a main model, the number of the submodels is the same as that of neurons in a layer 1, the input of the submodels is input layer neurons connected with the neurons in the layer 1, the output of the submodels is the input number of the neurons in the layer 1, the input layer of the neural network model is deleted, the input of the neurons in the layer 1 is modified into the output of corresponding submodels to serve as the main model, the monitoring node establishes safe multi-party calculation for each submodel and a financial institution to obtain the output of the submodels, and the monitoring node substitutes the output of the submodels into the main model to obtain the output of the main model, namely the lition score.
7. The system of claim 6, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
when the supervision node establishes safe multi-party calculation for each sub-model and the financial institution, the financial institution is informed to check whether the service data is the latest data, if not, the variable quantity of the service data is calculated, the variable quantity is superposed on the reserved copy, and the updated reserved copy participates in the safe multi-party calculation.
8. The system of claim 7, wherein the small amount financial debt dispute litigation smart action system based on the blockchain,
the supervisory node divides each connection of the submodel into two connections, the two connections are respectively recorded as a cooperative connection and a reserved connection, the weights of the cooperative connection and the reserved connection are respectively recorded as a cooperative weight and a reserved weight, the financial institution randomly generates the reserved weight, the product of the value of the field corresponding to the updated reserved copy and the reserved weight is sent to the supervisory node, the ratio k of the value in the reserved copy in the corresponding field to the value in the cooperative copy is calculated, the ratio k is sent to the supervisory node, the supervision node calculates and obtains the cooperation weight according to the weight of the original connection, the ratio k and the retention weight, so that the ratio k, the cooperation weight and the retention weight form an equivalent weight equal to the weight of the original connection, and the supervision node obtains the output of the sub-model according to the product of the value of the corresponding field of the cooperative copy, the cooperative weight and the report of the financial institution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111022109.1A CN113793208B (en) | 2021-09-01 | 2021-09-01 | Block chain-based small financial creditor dispute intelligent litigation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111022109.1A CN113793208B (en) | 2021-09-01 | 2021-09-01 | Block chain-based small financial creditor dispute intelligent litigation system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113793208A true CN113793208A (en) | 2021-12-14 |
| CN113793208B CN113793208B (en) | 2024-04-09 |
Family
ID=79182588
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111022109.1A Active CN113793208B (en) | 2021-09-01 | 2021-09-01 | Block chain-based small financial creditor dispute intelligent litigation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113793208B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881772A (en) * | 2022-07-05 | 2022-08-09 | 浙江数秦科技有限公司 | Loan processing method based on block chain |
| CN117350850A (en) * | 2023-12-06 | 2024-01-05 | 四川享宇科技有限公司 | Method for credit evaluation based on position and gap filling behavior data |
| CN117829987A (en) * | 2024-03-05 | 2024-04-05 | 山东兰图地理信息科技有限公司 | Real estate information management method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160232631A1 (en) * | 2015-01-13 | 2016-08-11 | Timothy J. Young | Repayment through non recourse contracts related to litigation |
| US20180285971A1 (en) * | 2017-03-31 | 2018-10-04 | International Business Machines Corporation | Management of consumer debt collection using a blockchain and machine learning |
| CN111459958A (en) * | 2020-03-30 | 2020-07-28 | 潘小胜 | Realization method of strength-endowing notarization technology based on block chain |
| CN111897770A (en) * | 2020-07-15 | 2020-11-06 | 中国建设银行股份有限公司 | Method and system for initiating litigation request |
| CN112837075A (en) * | 2021-02-08 | 2021-05-25 | 中船（浙江）海洋科技有限公司 | Remote notarization system and method based on block chain |
-
2021
- 2021-09-01 CN CN202111022109.1A patent/CN113793208B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160232631A1 (en) * | 2015-01-13 | 2016-08-11 | Timothy J. Young | Repayment through non recourse contracts related to litigation |
| US20180285971A1 (en) * | 2017-03-31 | 2018-10-04 | International Business Machines Corporation | Management of consumer debt collection using a blockchain and machine learning |
| CN111459958A (en) * | 2020-03-30 | 2020-07-28 | 潘小胜 | Realization method of strength-endowing notarization technology based on block chain |
| CN111897770A (en) * | 2020-07-15 | 2020-11-06 | 中国建设银行股份有限公司 | Method and system for initiating litigation request |
| CN112837075A (en) * | 2021-02-08 | 2021-05-25 | 中船（浙江）海洋科技有限公司 | Remote notarization system and method based on block chain |
Non-Patent Citations (2)
| Title |
|---|
| 张玮智;: "民事诉讼中区块链技术应用问题研究――基于区块链存证的视角", 江苏理工学院学报, no. 01, pages 72 - 78 * |
| 陈新: "民事诉讼中基于区块链技术的电子证据应用研究", 法制博览, no. 3, pages 215 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881772A (en) * | 2022-07-05 | 2022-08-09 | 浙江数秦科技有限公司 | Loan processing method based on block chain |
| CN114881772B (en) * | 2022-07-05 | 2022-09-23 | 浙江数秦科技有限公司 | Loan processing method based on block chain |
| CN117350850A (en) * | 2023-12-06 | 2024-01-05 | 四川享宇科技有限公司 | Method for credit evaluation based on position and gap filling behavior data |
| CN117350850B (en) * | 2023-12-06 | 2024-02-23 | 四川享宇科技有限公司 | Method for credit evaluation based on position and gap filling behavior data |
| CN117829987A (en) * | 2024-03-05 | 2024-04-05 | 山东兰图地理信息科技有限公司 | Real estate information management method and system |
| CN117829987B (en) * | 2024-03-05 | 2024-05-10 | 山东兰图地理信息科技有限公司 | Real estate information management method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113793208B (en) | 2024-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113793208A (en) | Small-amount financial debt dispute smart litigation system based on block chain | |
| Brühl | Virtual currencies, distributed ledgers and the future of financial services | |
| Huda et al. | Fuzzy MADM approach for rating of process-based fraud | |
| CN113268760B (en) | Distributed data fusion platform based on block chain | |
| CN106296390A (en) | A kind of bill that improves processes method and the bill processing system of safety | |
| WO2019021312A1 (en) | An automated system for default probability prediction of loans and method thereof | |
| CN108182627A (en) | A kind of system that user credit assessment is realized according to user behavior | |
| CN114492605A (en) | Federal learning feature selection method, device and system and electronic equipment | |
| CN111061718A (en) | Data checking method and device | |
| CN113793209B (en) | Intelligent litigation system based on blockchain | |
| CN113849567B (en) | Credited dispute early warning system based on data fusion | |
| CN113159796A (en) | Trade contract verification method and device | |
| CN112801780A (en) | Method, device and system for identifying international and international risk customers based on federal learning | |
| CN111539724A (en) | Electronic commercial acceptance bill financing method and device based on block chain architecture | |
| CN106779723A (en) | A kind of mobile terminal methods of risk assessment and device | |
| CN115564591A (en) | Financing product determination method and related equipment | |
| CN115170295A (en) | Enterprise credit risk assessment processing method and device | |
| Choi et al. | Classification model for detecting and managing credit loan fraud based on individual-level utility concept | |
| CN114240588A (en) | Annuity plan reconciliation method, device, equipment and medium based on block chain | |
| CN113240420A (en) | Financial encryption payment method and system based on big data | |
| Xu | The influence of artificial intelligence on the financial industry | |
| CN112288582A (en) | Information processing method and device for improving policy approval efficiency | |
| Jain | From Crypto Wild West to Regulated Frontier: Unleashing the Potential of Blockchain Technology | |
| CN117094722B (en) | Security supervision method and system for online payment | |
| Mendes | Forecasting bitcoin prices: ARIMA vs LSTM |
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 |