CN111340611A - Risk early warning method and device - Google Patents
Risk early warning method and device Download PDFInfo
- Publication number
- CN111340611A CN111340611A CN202010105184.3A CN202010105184A CN111340611A CN 111340611 A CN111340611 A CN 111340611A CN 202010105184 A CN202010105184 A CN 202010105184A CN 111340611 A CN111340611 A CN 111340611A
- Authority
- CN
- China
- Prior art keywords
- community
- node
- family
- nodes
- information
- 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
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000012546 transfer Methods 0.000 claims abstract description 65
- 238000004422 calculation algorithm Methods 0.000 claims description 23
- 238000012544 monitoring process Methods 0.000 claims description 17
- 238000007477 logistic regression Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 8
- 238000000638 solvent extraction Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 230000001902 propagating effect Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 19
- 238000012545 processing Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 11
- 238000004891 communication Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000002265 prevention Effects 0.000 description 6
- 238000011835 investigation Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000000644 propagated effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000012550 audit Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000006467 substitution reaction 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
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a risk early warning method and device, and relates to the technical field of computers. One embodiment of the method comprises: acquiring the relativity of the borrower with overdue loan, and constructing a family member network according to the relativity; calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period; updating the weight to a family member network to obtain a rights undirected graph, and carrying out community division on the rights undirected graph to obtain a family community of the borrower; and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score. According to the embodiment, the community division is carried out on the family member network of the loan persons who are overdue, the family community of the loan persons who are overdue is reasonably excavated, the default risk of the family members in the family community is predicted, early warning prompt is carried out, the risk precaution degree is improved, and the loan risk is reduced.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a risk early warning method and device.
Background
Loan transactions have various uncertain risk factors that may result in the loan funds not being withdrawn as expected, putting the property of the commercial bank that issued the loan at risk, or even causing loss. Thus, the commercial bank needs to make a loan risk assessment to the customer before or after the loan is issued. In the prior art, a commercial bank audits data such as age, education degree, marital status, monthly income, credit investigation condition and the like of a client to estimate a risk level for business personnel to take precautions in advance.
In the process of implementing the invention, the inventor finds that at least the following problems exist in the prior art:
if a client's loan at a commercial bank is overdue, its family fund chain may have been broken, or there is a case that a malicious debt is not yet made, then the client's family member will have a greater risk of default on the loan stock at the commercial bank. How to acquire family members of the client and predict default conditions of the family members is a problem to be solved.
Disclosure of Invention
In view of this, embodiments of the present invention provide a risk early warning method and apparatus, which perform community division on a network of family members of loaners who have overdue, reasonably dig out a family community of loaners who have overdue, and further predict default risks of family members in the family community and perform early warning prompting, thereby improving risk precaution degree and reducing loan risks.
To achieve the above object, according to an aspect of an embodiment of the present invention, a risk early warning method is provided.
The risk early warning method of the embodiment of the invention comprises the following steps: acquiring the relationship of the borrowers with overdue loan, and constructing a family member network according to the relationship; calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period; updating the weight to the family member network to obtain a right undirected graph, and carrying out community division on the right undirected graph to obtain a family community of the borrower; and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score.
Optionally, the community partitioning of the weighted undirected graph includes: and (3) node selection: selecting a monitoring node from the plurality of nodes of the authorized undirected graph, and determining an audience node according to the connection relation between the monitoring node and other nodes; and (3) propagating and updating: the listener node transmits the label with the largest occurrence frequency in the self label list to the monitoring node, and the monitoring node selects the label with the largest weight and updates the label to the self label list; wherein, the initial label of the label list is a node name; repeatedly executing the node selection step and the propagation updating step until the set first iteration times are reached, and counting the frequency of the tags in the tag lists of the nodes; and selecting the label with the maximum frequency as the community label name of the node, and taking the node with the same community label name as a community.
Optionally, the community partitioning of the weighted undirected graph includes: initializing each node of the weighted undirected graph as a community; distributing: traversing the nodes of the authorized undirected graph, trying to distribute the nodes to communities where neighbor nodes are located, calculating the modularity change before and after distribution, and distributing the nodes to the communities where the neighbor nodes with the largest modularity change are located; repeating: repeatedly executing the distribution step until the community in which the node is located remains unchanged; and (3) reconstruction: reconstructing nodes belonging to the same community into a new community to obtain a new weighted undirected graph, and updating the weight of a ring of the new community and the weight of an edge between the new communities; and repeatedly executing the distribution step, the repetition step and the reconstruction step until the modularity of the new authorized undirected graph is kept unchanged or a set second iteration number is reached.
Optionally, calculating the weight between the nodes comprises: calculating the ratio of the transfer amount between the node i and the node j to the total transfer amount of the node i and the total transfer amount of the node j in the set time period to obtain a first ratio and a second ratio; calculating the ratio of the transfer times between the node i and the node j to the total transfer times of the node i and the node j in the set time period to obtain a third ratio and a fourth ratio; summing the first ratio and the second ratio to obtain a first sum, and summing the third ratio and the fourth ratio to obtain a second sum; and weighting and summing the first sum value and the second sum value, and determining the weight between the node i and the node j according to the weighted and summed result.
Optionally, updating the weights of the rings of the new community and the weights of the edges between the new communities comprises: calculating the sum of the weights among the nodes contained in the new community to obtain a third sum, and taking twice of the third sum as the weight of the ring of the new community; and calculating the weight sum of edges connected across communities between the new communities to obtain a fourth sum value, and taking the fourth sum value as the weight of the edges between the new communities.
Optionally, determining a risk score for the family member comprises: acquiring sample data of a set time window, wherein the sample data comprises basic information, loan information and family fund chain associated information of a plurality of applicants; determining feature variables and classification variables affecting a risk score based on the sample data; wherein the categorical variable is used to indicate whether the applicant repayment is normal or abnormal; constructing a risk early warning model by using a logistic regression algorithm and the characteristic variables, and converting a logistic regression result into a benchmark score; and inputting the basic information, the loan information and the family fund chain association information of the family members into the risk early warning model to obtain the risk score of the family members.
Optionally, determining feature variables affecting a risk score based on the sample data comprises: analyzing data of different dimensions in the sample data to obtain initial variables influencing risk scores; and calculating an evidence weight value and an information value of the initial variable, and screening out a characteristic variable from the initial variable according to the evidence weight value and the information value.
Optionally, determining a risk score for the family member comprises: acquiring basic information and loan information of family members contained in the family community, and transfer information and guarantee information in the family community; respectively converting the basic information, the loan information, the transfer information and the guarantee information into numerical values according to set rules; and carrying out weighted summation on the numerical values to obtain the risk score of the family community, and taking the risk score of the family community as the risk score of family members contained in the family community.
In order to achieve the above object, according to another aspect of the embodiments of the present invention, a risk early warning apparatus is provided.
The risk early warning device of the embodiment of the invention comprises: the system comprises a network construction module, a family member network and a client side, wherein the network construction module is used for acquiring the relatives of a borrower with overdue loan and constructing the family member network according to the relatives; the weight calculation module is used for calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period; the community division module is used for updating the weight to the family member network to obtain a entitled undirected graph and carrying out community division on the entitled undirected graph to obtain a family community of the borrower; and the score early warning module is used for carrying out early warning prompt according to the risk score after determining the risk score of the family member according to the basic information and loan information of the family member in the family community.
Optionally, the community dividing module further includes a node selection module, a propagation update module, a repeat execution module, and a community determination module; the node selection module is used for selecting a monitoring node from the plurality of nodes of the authorized undirected graph and determining an audience node according to the connection relation between the monitoring node and other nodes; the propagation updating module is used for the audience node to propagate the label with the largest occurrence frequency in the self label list to the monitoring node, and the monitoring node selects the label with the largest weight and updates the label to the self label list; wherein, the initial label of the label list is a node name; the repeated execution module is used for repeatedly executing the processing procedures of the node selection module and the propagation updating module until the set first iteration times is reached, and counting the frequency of the tags in the tag lists of the nodes; and the community determining module is used for selecting the label with the maximum frequency as the community label name of the node and taking the node with the same community label name as a community.
Optionally, the community dividing module further includes: the system comprises an initialization module, a distribution module, a first repetition module, a reconstruction module and a second repetition module; the initialization module is configured to initialize each node of the weighted undirected graph as a community; the distribution module is used for traversing the nodes of the authorized undirected graph, trying to distribute the nodes to communities where the neighbor nodes are located, calculating the modularity change before and after distribution, and distributing the nodes to the communities where the neighbor nodes with the largest modularity change are located; the first repeating module is used for repeatedly executing the processing process of the distribution module until the community where the node is located keeps unchanged; the reconstruction module is used for reconstructing nodes belonging to the same community into a new community to obtain a new weighted undirected graph and updating the weight of a ring of the new community and the weight of an edge between the new communities; and the second repeating module is used for repeatedly executing the processing procedures of the distribution module, the repeating module and the reconstructing module until the modularity of the new weighted undirected graph is kept unchanged or a set second iteration number is reached.
Optionally, the weight calculation module is further configured to calculate ratios of transfer amounts between the node i and the node j to a total transfer amount of the node i and a total transfer amount of the node j in the set time period, so as to obtain a first ratio and a second ratio; calculating the ratio of the transfer times between the node i and the node j to the total transfer times of the node i and the node j in the set time period to obtain a third ratio and a fourth ratio; summing the first ratio and the second ratio to obtain a first sum, and summing the third ratio and the fourth ratio to obtain a second sum; and weighting and summing the first sum value and the second sum value, and determining the weight between the node i and the node j according to the weighted and summed result.
Optionally, the reconstruction module is further configured to calculate a sum of weights between nodes included in the new community to obtain a third sum, and use twice of the third sum as a weight of a ring of the new community; and calculating the weight sum of edges connected across communities between the new communities to obtain a fourth sum value, and taking the fourth sum value as the weight of the edges between the new communities.
Optionally, the scoring early warning module is further configured to obtain sample data of a set time window, where the sample data includes basic information, loan information, and family fund chain association information of a plurality of applicants; determining feature variables and classification variables affecting a risk score based on the sample data; wherein the categorical variable is used to indicate whether the applicant repayment is normal or abnormal; constructing a risk early warning model by using a logistic regression algorithm and the characteristic variables, and converting a logistic regression result into a benchmark score; and inputting the basic information, the loan information and the family fund chain association information of the family members into the risk early warning model to obtain the risk score of the family members.
Optionally, the score early warning module is further configured to analyze data of different dimensions in the sample data to obtain initial variables affecting the risk score; and calculating an evidence weight value and an information value of the initial variable, and screening out a characteristic variable from the initial variable according to the evidence weight value and the information value.
Optionally, the scoring early warning module is further configured to obtain basic information and loan information of family members included in the family community, and transfer information and guarantee information in the family community; respectively converting the basic information, the loan information, the transfer information and the guarantee information into numerical values according to set rules; and carrying out weighted summation on the numerical values to obtain the risk score of the family community, and taking the risk score of the family community as the risk score of family members contained in the family community.
To achieve the above object, according to still another aspect of embodiments of the present invention, there is provided an electronic apparatus.
An electronic device of an embodiment of the present invention includes: one or more processors; a storage device, configured to store one or more programs, which when executed by the one or more processors, cause the one or more processors to implement a risk pre-warning method according to an embodiment of the present invention.
To achieve the above object, according to still another aspect of embodiments of the present invention, there is provided a computer-readable medium.
A computer-readable medium of an embodiment of the present invention stores thereon a computer program, which when executed by a processor implements a risk early warning method of an embodiment of the present invention.
One embodiment of the above invention has the following advantages or benefits: the method comprises the steps that community division is carried out on a family member network of a loan person who is overdue, a family community of the loan person who is overdue is reasonably excavated, the default risk of the family member in the family community is predicted, early warning prompt is carried out, the risk precaution degree is improved, and the loan risk is reduced; based on an SLPA algorithm or a Louvain algorithm, community division is carried out on a weighted undirected graph (namely a family member network with weights), and the rationality of the divided communities is further ensured; the intimate degree of capital exchange between the nodes is measured through the transfer information between the nodes, so that the weight between the nodes is convenient to determine; predicting the risk score of the family members by using a logistic regression algorithm, converting the loan risk into a classification problem, and providing a basis for business personnel to make decisions; and risk early warning is carried out from the family community dimension, the early warning range is further expanded, and the loan risk is reduced.
Further effects of the above-mentioned non-conventional alternatives will be described below in connection with the embodiments.
Drawings
The drawings are included to provide a better understanding of the invention and are not to be construed as unduly limiting the invention. Wherein:
fig. 1 is a schematic diagram of main steps of a risk early warning method according to a first embodiment of the present invention;
fig. 2 is a schematic main flow chart of a risk early warning method according to a second embodiment of the present invention;
FIG. 3 is a diagram illustrating a right undirected graph according to a second embodiment of the present invention;
FIG. 4 is a schematic diagram of a tag list of a node according to a second embodiment of the present invention;
FIG. 5 is a schematic diagram of a risk scoring process for an individual dimension according to a second embodiment of the present invention;
fig. 6 is a schematic main flow chart of a risk early warning method according to a third embodiment of the present invention;
fig. 7 is a schematic diagram of main modules of a risk early warning apparatus according to an embodiment of the present invention;
FIG. 8 is an exemplary system architecture diagram in which embodiments of the present invention may be employed;
FIG. 9 is a schematic diagram of a computer apparatus suitable for use in an electronic device to implement an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention are described below with reference to the accompanying drawings, in which various details of embodiments of the invention are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
Fig. 1 is a schematic diagram of main steps of a risk early warning method according to a first embodiment of the present invention. As shown in fig. 1, a risk early warning method according to an embodiment of the present invention mainly includes the following steps:
step S101: and acquiring the relationship of the borrowers with overdue loan, and constructing a family member network according to the relationship. If the loan of the loan person in the bank is overdue, the relationship of the loan person can be obtained based on the application form filled by the loan person, the information source of the third party and other data, the loan person and the members of the family are both used as a node, and the two nodes with the relationship of the first degree of the relationship are connected to obtain the family member network. Taking the borrower as an example, the relatives members having a degree of relativity with the borrower are the parents, spouses, children and brothers of the borrower.
Step S102: and calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period. And acquiring the transfer information of each node in the family member network within a set time period, and calculating the weight between the nodes according to the transfer information. The calculation method of the weight between any two nodes (node i and node j) is as follows:
calculating the ratio of the transfer amount between the node i and the node j to the total transfer amount of the node i and the total transfer amount of the node j respectively to obtain a first ratio and a second ratio; calculating the ratio of the transfer times between the node i and the node j to the total transfer times of the node i and the node j respectively to obtain a third ratio and a fourth ratio; summing the first ratio and the second ratio to obtain a first sum, and summing the third ratio and the fourth ratio to obtain a second sum; and weighting and summing the first sum value and the second sum value to obtain the weight between the node i and the node j.
Step S103: and updating the weight to the family member network to obtain a rights undirected graph, and carrying out community division on the rights undirected graph to obtain the family community of the borrower. And correspondingly adding the weight between the nodes obtained in the step S102 to the connection edge of the corresponding node in the family member network to obtain the weighted undirected graph. A community structure of the weighted undirected graph is mined by using a community discovery algorithm, so that the relationship between nodes in the same community is tight, and the relationship between communities is sparse.
In an embodiment, the community discovery Algorithm may adopt a SLPA (Speaker-inside label discovery Algorithm) Algorithm or a Louvain Algorithm. The weighted undirected graph is divided into at least one community by processing of a community discovery algorithm. The community in which the borrower is located is the lender's family community.
Step S104: and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score. The method comprises the steps of obtaining basic information and loan information of each family member contained in a family community, wherein the basic information comprises age, occupation, credit investigation information and the like, and the loan information comprises loan amount, overdue times and the like in a set time period. And risk scoring of personal dimension and/or family dimension can be carried out based on the information, and the risk scoring can be listed in a close observation list when the risk scoring exceeds a threshold value, so that subsequent risk prevention is facilitated.
Fig. 2 is a schematic main flow chart of a risk early warning method according to a second embodiment of the present invention. As shown in fig. 2, the risk early warning method according to the second embodiment of the present invention is implemented based on the SLPA algorithm, and mainly includes the following steps:
step S201: and acquiring the relatives of the borrowers with overdue loan, and constructing a family member network according to the relatives. The borrower needs to register his or her first degree relationship when transacting loan-related business at the commercial bank. And acquiring the one-degree relationship of the members contained in the one-degree relationship of the borrower, then continuously acquiring the one-degree relationship of the members contained in the one-degree relationship, and acquiring the one-degree relationship for multiple times according to the method to obtain the relatives related to the borrower so as to construct a family member network.
Wherein the first degree relatives are parents, spouses, children, brothers and sisters. The nodes in the family member network are the borrower and the relatives of the borrower. The number of nodes between two nodes determines the relationship between the two nodes, and specifically, if the number of nodes is N, the two nodes are in (N +1) degree relationship. For example, there is a second degree relationship between a borrower's parent and a borrower's daughter.
Step S202: and calculating the weight among the nodes according to the transfer information of each node in the family member network within a set time period, and then updating the weight to the family member network to obtain a weighted undirected graph. The weight between the nodes is processed quantitatively according to the close degree of the money in the coming and going, and the processing characteristics comprise a plurality of factors such as the transfer amount between the two nodes, the transfer times between the two nodes, the total transfer amount of the nodes and the like. In an embodiment, the calculation formula of the weight between the node i and the node j may be as follows:
in the formula, WijIs the weight between node i and node j; mijThe transfer amount between the node i and the node j is obtained; n is a radical ofijThe number of account transfers between the node i and the node j is obtained; mi、MjThe total transfer amount of the node i and the node j is respectively; n is a radical ofi、NjThe total number of account transfers of the node i and the node j respectively; w1、W2All are weighted values, and the sum of the weighted values and the weighted value is 1.
In practical application, factors influencing the weights between the nodes may be increased according to requirements, for example, the calculation formula of the weights between the node i and the node j may be as follows:
in the formula, W1、W2、W3All are weight values, and the sum of the weight values and the weight values is 1. W1、W2、W3The specific value of (a) can be determined empirically.
And then, updating the calculated weight to the family member network to obtain the authorized undirected graph. Fig. 3 is a schematic diagram of a right undirected graph according to a second embodiment of the invention. As shown in fig. 3, the borrower is a customer a, the name of each node in the drawing is the name of the relationship of the customer a, the numerical value on the connecting line between the nodes is the weight between two nodes, and the ellipses indicate that the drawing is a part of a directed undirected graph.
Step S203: an initial label is set in the label list of each node of the entitled undirected graph. And setting a label list for storing the occurrence times of label names for each node, wherein the initial label can be the name of the node, and the occurrence times is 1. Taking the client A as an example, the initial label of the corresponding node is the client A, and the initial label of the parent node is the parent name.
Step S204: randomly selecting a listening node from all nodes of the weighted undirected graph. And randomly selecting one node from all the nodes of the weighted undirected graph as a monitoring node, namely Listener, at each iteration.
Step S205: and each listener node of the listening nodes transmits the label with the largest occurrence frequency in the self label list to the listening nodes. The node directly connected to Listener is the Listener node, named Speaker. Assuming that the customer A is Listener, the father, mother, spouse and son directly connected to the customer A are Speaker. Each Speaker propagates the most frequently appearing tags in its tag list to the Listener.
Step S206: and the monitoring node selects the label with the maximum weight and updates the label to the label list of the monitoring node. Listener updates the label with the largest weight between itself and Speaker to the own label list. With reference to fig. 3, if the weight between the customer a and the spouse is the greatest, the tag propagated by the spouse (the tag at the first iteration is the initial tag, the name of the spouse) is updated to the tag list of the customer a, and the tag list of the customer a includes the customer a and the name of the spouse.
Fig. 4 is a schematic diagram of a tag list of a node according to a second embodiment of the present invention. As shown in FIG. 4, each node maintains its own tag list, LnTRepresenting the label of listener node n after the T-th iteration.
Step S207: judging whether the current iteration times reach a set first iteration time, if not, executing the step S204; if the first number of iterations has been reached, step S208 is performed. The first number of iterations T is preset, for example to 8.
Step S208: and counting the occurrence frequency of the labels in the label list of each node, and selecting the label with the maximum frequency as the community label name of the node. And after traversing for T times, counting the frequency of the tags in the tag list of each node, and selecting the tag with the maximum frequency as the community tag name of the node.
Step S209: and taking the nodes with the same community tag name as a community after passing through each node of the digraph, wherein the community where the borrower is located is the family community of the borrower. And traversing all the nodes of the authorized undirected graph again, searching out the nodes with the same community tag name, and adding the nodes into a set. The nodes in a set form a community. Assuming that the community tag name of the client A is the spouse name, and the community named by the spouse name of the client A comprises 50 members, the 50 members form the family community of the client A.
It should be noted that after determining good communities, if the community division is not reasonable, iteration may continue. The rationality of community division may be whether the number of members in a community is within a set threshold interval.
Step S210: and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score. In the step, risk early warning can be carried out based on personal dimensionality, and risk early warning can also be carried out based on family dimensionality. The risk early warning based on the individual dimension can be realized by logistic regression, and is described in detail later with reference to fig. 5. The realization process of the risk early warning based on family dimensionality is as follows:
acquiring basic information and loan information of all family members in a family community, and transfer information and guarantee information in the family community; and then, respectively converting the information into numerical values according to a set rule, and carrying out weighted summation on the numerical values to obtain the risk score of each family community. The basic information comprises personal information and credit investigation information, and the transfer information comprises transfer amount and transfer times.
Based on the risk scores, in a pre-loan link, the association relationship of family members in a family community with lower risk scores (namely high default risk) can be mainly found out; in the middle-loan link, the family community credit limit with lower risk score needs to be strictly checked; in the post-loan link, capital exchange monitoring of the family community with lower risk score needs to be enhanced.
Through the processing from step S201 to step S209, the second-degree, third-degree and other multi-degree relationships and the tightly connected communities can be obtained from the first-degree relationships of the borrowers. Therefore, loan data of family members including second-degree and third-degree relatives in the family community of the overdue client can be inquired, the family fund chain risk is monitored, the monitored early warning range is expanded, and the risk prevention degree is improved.
Fig. 5 is a schematic diagram of a risk scoring process of an individual dimension according to a second embodiment of the present invention. As shown in fig. 5, the risk scoring process (i.e., step S210) of the individual dimension according to the second embodiment of the present invention includes the following steps:
step S501: and acquiring sample data of a set time window. Wherein the sample data comprises basic information, loan information and family fund chain related information of a plurality of applicants. The applicant is a customer who has applied for loan service at a bank. The basic information of the applicant includes age, occupation, credit information such as the number of strokes of a long loan, and the like. The applicant's loan information includes the number of loan overdues over a period of time, such as the number of loan overdues of approximately 1 year. The family fund chain related information comprises the loan overdue times of family members in the family community of the applicant, such as the loan overdue times of nearly 3 months.
Step S502: feature variables and classification variables that affect the risk score are determined based on the sample data. Analyzing data of different dimensions in the sample data to obtain initial variables influencing risk scores; then, an Evidence Weight Value (WOE) and an Information Value (IV) Of the initial variable are calculated, and a characteristic variable is screened out from the initial variable according to the Evidence Weight Value and the information Value.
Wherein the initial variables comprise the following three types of data: basic information, loan information and family fund chain related information. The classification variable is used for indicating whether the payment of the applicant is normal or abnormal, and the standards of normal payment and abnormal payment can be defined by self, for example, the definition of normal payment can be as follows: the payment exception can be defined as: the half-year appearance period is over 2 or more.
And preprocessing the initial variables and the classification variables, and then screening out characteristic variables from the preprocessed initial variables by combining the significance and the stability of the variables. Wherein the preprocessing comprises binning discretization and WOE encoding. Variable Stability is measured as the Population Stability Index (PSI).
After preprocessing, the distinguishing strength of a single variable on the model is inspected through IV, PSI of the calculated variable on a development sample and a time-span sample is calculated, the stability of the variable is evaluated, and effective variables are screened out for model parameter fitting. In the example, the number of initial variables is 200, and 18 feature variables are selected by the above processing.
Step S503: and constructing a risk early warning model by using a logistic regression algorithm and the characteristic variables, and converting the logistic regression result into a benchmark score. And establishing a model by using logistic regression, and converting a result obtained by the logistic regression into a benchmark scoring card format by using a rounding.
Step S504: the risk pre-warning model was validated using KS and kini coefficients. KS is Kolmogorov-Smirnov, also known as Kolmogorov-Smirnov. The Gini coefficient is the Gini coefficient. The calculated risk early warning model has the KS of 56.71 percent and Gini of 68.79 percent, and the model has better discrimination. Compared with the method without adding the family fund chain related information, KS is improved by 5 points, and Gini is improved by 7 points.
Step S505: and inputting the basic information, the loan information and the family fund chain association information of the family members into a risk early warning model to obtain the risk score of the family members. And inputting the information of the loan clients into a risk early warning model to obtain the risk score of the loan clients. The higher the risk score, the lower the risk of an abnormal repayment (i.e., more than 2 years overdue within a half year).
When the benchmark score is set to be 600 minutes, the ratio of the number of the clients with normal repayment and abnormal repayment is as follows: 20:1, the ratio is doubled for each 20 minutes increase. When the risk score of the client is 600 minutes, the ratio of the number of the clients with normal repayment and abnormal repayment is 20: 1; when the risk score of the client is 620, the ratio of the number of the clients with normal repayment and abnormal repayment is 40: 1; the ratio of the number of customers with a normal and abnormal repayment is 10:1 when the customer risk score is 580.
When the risk score of a family member is below a set threshold, the family member is placed on a close watch list. For the customers in the close observation list, the capital exchange monitoring level needs to be improved; if the client has a loan at present, the collection hastening force is improved; if the client is in the loan application phase, the line of credit may be decreased.
Fig. 6 is a schematic main flow chart of a risk early warning method according to a third embodiment of the present invention. As shown in fig. 6, the risk early warning method according to the third embodiment of the present invention is implemented based on a Louvain algorithm, and mainly includes the following steps:
step S601: and acquiring the relatives of the borrowers with overdue loan, and constructing a family member network according to the relatives. The implementation process of this step is the same as step S201, and is not described herein again.
Step S602: and calculating the weight among the nodes according to the transfer information of each node in the family member network within a set time period, and then updating the weight to the family member network to obtain a weighted undirected graph. In this embodiment, the calculation formula of the weight between the node i and the node j may be as follows:
in the formula, WijIs the weight between node i and node j; mijThe transfer amount between the node i and the node j is obtained; n is a radical ofijThe number of account transfers between the node i and the node j is obtained; mi、MjThe total transfer amount of the node i and the node j is respectively; n is a radical ofi、NjThe total number of account transfers of the node i and the node j respectively; w1、W2All are weighted values, and the sum of the weighted values and the weighted value is 1.
Step S603: each node of the weighted undirected graph is initialized as a community. And each node in the authorized undirected graph is regarded as an independent community, and the number of the communities is the same as that of the nodes.
Step S604: and (4) trying to distribute the nodes to communities where neighbor nodes are located after nodes of the authorized undirected graph are traversed, and calculating the modularity change before and after distribution. The modularity Q is also called a modularization metric value, the size of the modularity value mainly depends on the community division condition of the network, and the modularity value can be used for quantitatively measuring the network community division quality, the closer the value is to 1, the stronger the strength of the community structure divided by the network is, namely the better the division quality is. Optimal network community partitioning can be achieved by maximizing the modularity Q.
And (3) for each node i, sequentially trying to distribute the node i to the community where each neighbor node is located, calculating the modularity change delta Q before and after distribution, and recording the neighbor node with the maximum delta Q (namely max delta Q) for each node i.
Step S605: and distributing the nodes to the communities where the neighbor nodes with the most variation of the modularity exist. In a preferred embodiment, whether the maximum modularity change max Δ Q is greater than 0 is judged, and if max Δ Q is greater than 0, the node i is allocated to a community where a neighbor node with the maximum Δ Q is located; if max Δ Q is less than or equal to 0, the community in which the node i is located is kept unchanged.
Step S606: and judging whether the community in which the node is located keeps unchanged, if so, executing the step S604, and if not, executing the step S607. The above steps S604 to S606 belong to the first stage iteration, and are used for node transfer between communities and for node transfer evaluation.
After several iterations, the embodiment judges whether the community where the node is located remains unchanged. And if the communities to which all the nodes belong do not change any more, the node transfer of the social interval is considered to be finished, and the iteration of the round is maximized.
Step S607: and reconstructing the nodes belonging to the same community into new communities to obtain new weighted undirected graphs, and updating the weight of the ring of each new community and the weight of the edges between the new communities. In the first stage of iteration, some community nodes are changed (nodes are added and nodes are reduced), so that the original authorized undirected graph needs to be reconstructed. This step is a second stage iteration for reconstructing the weighted undirected graph. In the reconstruction, all nodes in the same community are reconstructed into a new community.
And calculating the sum of the weights among the nodes contained in the new community, and updating twice of the sum of the weights as the weight of the ring of the new community. For example, the new community 1 contains 3 nodes, and there are 2 connecting edges, and the weights of the 2 connecting edges are summed, and the sum is 2 times the weight of the ring of the new community 1. And calculating the weight sum of edges connected across communities among the new communities, and updating the weight sum into the weight of the edges among the new communities. For example, there are 3 connecting edges between the new community 1 and the new community 2, and the weight of the edge between the two communities is the sum of the weights of the 3 connecting edges.
Step S608: judging whether the modularity of the new authorized undirected graph is kept unchanged or reaches a set second iteration number, if so, executing a step S609; otherwise, step S604 is executed. And continuing to start the first-stage iteration and the second-stage iteration of the next round until the modularity of the new weighted undirected graph is kept unchanged or the set second iteration number is reached. After several rounds of iteration, the method judges whether the modularity of the new authorized undirected graph is unchanged or reaches the set second iteration number.
Step S609: and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score. The implementation process of this step is the same as step S210, and is not described herein again.
According to the risk early warning method provided by the embodiment of the invention, the family community of the loan officer who is overdue is reasonably excavated by carrying out community division on the family member network of the loan officer who is overdue, so that the family fund chain risk is monitored, the risk prevention degree is improved, and the loan risk is reduced; based on an SLPA algorithm or a Louvain algorithm, community division is carried out on a weighted undirected graph (namely a family member network with weights), and the rationality of the divided communities is further ensured; the intimate degree of capital exchange between the nodes is measured through the transfer information between the nodes, so that the weight between the nodes is convenient to determine; predicting the risk score of the family members by using a logistic regression algorithm, converting the loan risk into a classification problem, and providing a basis for business personnel to make decisions; and risk early warning is carried out from the family community dimension, the early warning range is further expanded, and the loan risk is reduced.
Fig. 7 is a schematic diagram of main modules of a risk early warning device according to an embodiment of the present invention. As shown in fig. 7, a risk early warning apparatus 700 according to an embodiment of the present invention mainly includes:
the network construction module 701 is used for acquiring the relatives of the borrowers with overdue loan and constructing the family member network according to the relatives. If the loan of the loan person in the bank is overdue, the relationship of the loan person can be obtained based on the application form filled by the loan person, the information source of the third party and other data, the loan person and the members of the family are both used as a node, and the two nodes with the relationship of the first degree of the relationship are connected to obtain the family member network. Taking the borrower as an example, the relatives members having a degree of relativity with the borrower are the parents, spouses, children and brothers of the borrower.
A weight calculation module 702, configured to calculate weights among the nodes according to transfer information of multiple nodes in the family member network within a set time period. And acquiring the transfer information of each node in the family member network within a set time period, and calculating the weight between the nodes according to the transfer information. The calculation method of the weight between any two nodes (node i and node j) is as follows:
calculating the ratio of the transfer amount between the node i and the node j to the total transfer amount of the node i and the total transfer amount of the node j respectively to obtain a first ratio and a second ratio; calculating the ratio of the transfer times between the node i and the node j to the total transfer times of the node i and the node j respectively to obtain a third ratio and a fourth ratio; summing the first ratio and the second ratio to obtain a first sum, and summing the third ratio and the fourth ratio to obtain a second sum; and weighting and summing the first sum value and the second sum value to obtain the weight between the node i and the node j.
The community division module 703 is configured to update the weight to the family member network to obtain a rights undirected graph, and perform community division on the rights undirected graph to obtain a family community of the borrower. The weights between the nodes obtained by the weight calculation module 702 are correspondingly added to the connection edges of the corresponding nodes in the family member network to obtain a weighted undirected graph. A community structure of the weighted undirected graph is mined by using a community discovery algorithm, so that the relationship between nodes in the same community is tight, and the relationship between communities is sparse.
In an embodiment, the community discovery algorithm may adopt an SLPA algorithm or a Louvain algorithm. The weighted undirected graph is divided into at least one community by processing of a community discovery algorithm. The community in which the borrower is located is the lender's family community.
And the score early warning module 704 is used for performing early warning prompt according to the risk score after determining the risk score of the family member according to the basic information and loan information of the family member in the family community. The method comprises the steps of obtaining basic information and loan information of each family member contained in a family community, wherein the basic information comprises age, occupation, credit investigation information and the like, and the loan information comprises loan amount, overdue times and the like in a set time period. And risk scoring of personal dimension and/or family dimension can be carried out based on the information, and the risk scoring can be listed in a close observation list when the risk scoring exceeds a threshold value, so that subsequent risk prevention is facilitated.
From the above description, the community division is performed on the family member network of the loan officer who is overdue, and the family community of the loan officer who is overdue is reasonably excavated, so that the family fund chain risk is monitored, the risk prevention degree is improved, and the loan risk is reduced.
Fig. 8 shows an exemplary system architecture 800 to which the risk pre-warning method or risk pre-warning apparatus of the embodiments of the invention may be applied.
As shown in fig. 8, the system architecture 800 may include terminal devices 801, 802, 803, a network 804, and a server 805. The network 804 serves to provide a medium for communication links between the terminal devices 801, 802, 803 and the server 805. Network 804 may include various types of connections, such as wire, wireless communication links, or fiber optic cables, to name a few.
A user may use the terminal devices 801, 802, 803 to interact with a server 805 over a network 804 to receive or send messages or the like. The terminal devices 801, 802, 803 may have installed thereon various communication client applications, such as shopping applications, web browser applications, search applications, instant messaging tools, mailbox clients, social platform software, and the like.
The terminal devices 801, 802, 803 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 805 may be a server that provides various services, such as a background management server that an administrator processes using risk assessment instructions sent by the terminal devices 801, 802, 803. The background management server can construct a family member network, calculate node weights, divide communities, calculate risk scores, and feed back processing results (such as risk scores) to the terminal equipment.
It should be noted that the risk pre-warning method provided in the embodiment of the present application is generally executed by the server 805, and accordingly, the risk pre-warning apparatus is generally disposed in the server 805.
It should be understood that the number of terminal devices, networks, and servers in fig. 8 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
The invention also provides an electronic device and a computer readable medium according to the embodiment of the invention.
The electronic device of the present invention includes: one or more processors; a storage device, configured to store one or more programs, which when executed by the one or more processors, cause the one or more processors to implement a risk pre-warning method according to an embodiment of the present invention.
The computer readable medium of the present invention has stored thereon a computer program which, when executed by a processor, implements a risk pre-warning method of an embodiment of the present invention.
Referring now to FIG. 9, shown is a block diagram of a computer system 900 suitable for use in implementing an electronic device of an embodiment of the present invention. The electronic device shown in fig. 9 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 9, the computer system 900 includes a Central Processing Unit (CPU)901 that can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)902 or a program loaded from a storage section 908 into a Random Access Memory (RAM) 903. In the RAM 903, various programs and data necessary for the operation of the computer system 900 are also stored. The CPU 901, ROM 902, and RAM 903 are connected to each other via a bus 904. An input/output (I/O) interface 905 is also connected to bus 904.
The following components are connected to the I/O interface 905: an input portion 906 including a keyboard, a mouse, and the like; an output section 907 including components such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 908 including a hard disk and the like; and a communication section 909 including a network interface card such as a LAN card, a modem, or the like. The communication section 909 performs communication processing via a network such as the internet. The drive 910 is also connected to the I/O interface 905 as necessary. A removable medium 911 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 910 as necessary, so that a computer program read out therefrom is mounted into the storage section 908 as necessary.
In particular, the processes described above with respect to the main step diagrams may be implemented as computer software programs, according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program containing program code for performing the method illustrated in the main step diagram. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 909, and/or installed from the removable medium 911. The above-described functions defined in the system of the present invention are executed when the computer program is executed by a Central Processing Unit (CPU) 901.
It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules described in the embodiments of the present invention may be implemented by software or hardware. The described modules may also be provided in a processor, which may be described as: a processor comprises a network construction module, a weight calculation module, a community division module and a grading early warning module. The names of the modules do not limit the modules in a certain condition, for example, the network construction module can also be described as a module for acquiring the relatives of the borrowers with overdue loan and constructing the family member network according to the relatives.
As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to comprise: acquiring the relationship of the borrowers with overdue loan, and constructing a family member network according to the relationship; calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period; updating the weight to the family member network to obtain a right undirected graph, and carrying out community division on the right undirected graph to obtain a family community of the borrower; and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score.
According to the technical scheme of the embodiment of the invention, the family community of the loan officer who is overdue is reasonably excavated by carrying out community division on the family member network of the loan officer who is overdue, so that the family fund chain risk is monitored, the risk prevention degree is improved, and the loan risk is reduced.
The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
The above-described embodiments should not be construed as limiting the scope of the invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions can occur, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (11)
1. A risk early warning method is characterized by comprising the following steps:
acquiring the relationship of the borrowers with overdue loan, and constructing a family member network according to the relationship;
calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period;
updating the weight to the family member network to obtain a right undirected graph, and carrying out community division on the right undirected graph to obtain a family community of the borrower;
and after determining the risk score of the family member according to the basic information and loan information of the family member in the family community, carrying out early warning prompt according to the risk score.
2. The method of claim 1, wherein community partitioning the weighted undirected graph comprises:
and (3) node selection: selecting a monitoring node from the plurality of nodes of the authorized undirected graph, and determining an audience node according to the connection relation between the monitoring node and other nodes;
and (3) propagating and updating: the listener node transmits the label with the largest occurrence frequency in the self label list to the monitoring node, and the monitoring node selects the label with the largest weight and updates the label to the self label list; wherein, the initial label of the label list is a node name;
repeatedly executing the node selection step and the propagation updating step until the set first iteration times are reached, and counting the frequency of the tags in the tag lists of the nodes;
and selecting the label with the maximum frequency as the community label name of the node, and taking the node with the same community label name as a community.
3. The method of claim 1, wherein community partitioning the weighted undirected graph comprises:
initializing each node of the weighted undirected graph as a community;
distributing: traversing the nodes of the authorized undirected graph, trying to distribute the nodes to communities where neighbor nodes are located, calculating the modularity change before and after distribution, and distributing the nodes to the communities where the neighbor nodes with the largest modularity change are located;
repeating: repeatedly executing the distribution step until the community in which the node is located remains unchanged;
and (3) reconstruction: reconstructing nodes belonging to the same community into a new community to obtain a new weighted undirected graph, and updating the weight of a ring of the new community and the weight of an edge between the new communities;
and repeatedly executing the distribution step, the repetition step and the reconstruction step until the modularity of the new authorized undirected graph is kept unchanged or a set second iteration number is reached.
4. The method of claim 1, 2 or 3, wherein calculating the weights between the nodes comprises:
calculating the ratio of the transfer amount between the node i and the node j to the total transfer amount of the node i and the total transfer amount of the node j in the set time period to obtain a first ratio and a second ratio;
calculating the ratio of the transfer times between the node i and the node j to the total transfer times of the node i and the node j in the set time period to obtain a third ratio and a fourth ratio;
summing the first ratio and the second ratio to obtain a first sum, and summing the third ratio and the fourth ratio to obtain a second sum;
and weighting and summing the first sum value and the second sum value, and determining the weight between the node i and the node j according to the weighted and summed result.
5. The method of claim 3, wherein updating the weights of the rings of the new community and the weights of the edges between the new community comprises:
calculating the sum of the weights among the nodes contained in the new community to obtain a third sum, and taking twice of the third sum as the weight of the ring of the new community;
and calculating the weight sum of edges connected across communities between the new communities to obtain a fourth sum value, and taking the fourth sum value as the weight of the edges between the new communities.
6. The method of claim 1, wherein determining the risk score for the family member comprises:
acquiring sample data of a set time window, wherein the sample data comprises basic information, loan information and family fund chain associated information of a plurality of applicants;
determining feature variables and classification variables affecting a risk score based on the sample data; wherein the categorical variable is used to indicate whether the applicant repayment is normal or abnormal;
constructing a risk early warning model by using a logistic regression algorithm and the characteristic variables, and converting a logistic regression result into a benchmark score;
and inputting the basic information, the loan information and the family fund chain association information of the family members into the risk early warning model to obtain the risk score of the family members.
7. The method of claim 6, wherein determining feature variables that affect a risk score based on the sample data comprises:
analyzing data of different dimensions in the sample data to obtain initial variables influencing risk scores;
and calculating an evidence weight value and an information value of the initial variable, and screening out a characteristic variable from the initial variable according to the evidence weight value and the information value.
8. The method of claim 1, wherein determining the risk score for the family member comprises:
acquiring basic information and loan information of family members contained in the family community, and transfer information and guarantee information in the family community;
respectively converting the basic information, the loan information, the transfer information and the guarantee information into numerical values according to set rules;
and carrying out weighted summation on the numerical values to obtain the risk score of the family community, and taking the risk score of the family community as the risk score of family members contained in the family community.
9. A risk early warning device, comprising:
the system comprises a network construction module, a family member network and a client side, wherein the network construction module is used for acquiring the relatives of a borrower with overdue loan and constructing the family member network according to the relatives;
the weight calculation module is used for calculating the weight among the nodes according to the transfer information of a plurality of nodes in the family member network within a set time period;
the community division module is used for updating the weight to the family member network to obtain a entitled undirected graph and carrying out community division on the entitled undirected graph to obtain a family community of the borrower;
and the score early warning module is used for carrying out early warning prompt according to the risk score after determining the risk score of the family member according to the basic information and loan information of the family member in the family community.
10. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-8.
11. A computer-readable medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010105184.3A CN111340611B (en) | 2020-02-20 | 2020-02-20 | Risk early warning method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010105184.3A CN111340611B (en) | 2020-02-20 | 2020-02-20 | Risk early warning method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111340611A true CN111340611A (en) | 2020-06-26 |
| CN111340611B CN111340611B (en) | 2024-03-08 |
Family
ID=71187090
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010105184.3A Active CN111340611B (en) | 2020-02-20 | 2020-02-20 | Risk early warning method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111340611B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111915426A (en) * | 2020-08-04 | 2020-11-10 | 中投国信(北京)科技发展有限公司 | Method and system for distinguishing credit users based on graph calculation and community division algorithm |
| CN111951104A (en) * | 2020-08-24 | 2020-11-17 | 上海银行股份有限公司 | Risk conduction early warning method based on associated graph |
| CN112150270A (en) * | 2020-09-18 | 2020-12-29 | 中国建设银行股份有限公司 | Method, device, equipment and computer readable medium for monitoring vehicle information |
| CN113129127A (en) * | 2021-04-21 | 2021-07-16 | 建信金融科技有限责任公司 | Early warning method and device |
| CN113409139A (en) * | 2021-07-27 | 2021-09-17 | 深圳前海微众银行股份有限公司 | Credit risk identification method, apparatus, device, and program |
| CN114240215A (en) * | 2021-12-22 | 2022-03-25 | 中国建设银行股份有限公司 | User loss of contact grade acquisition method and device and storage medium |
| CN114386488A (en) * | 2021-12-22 | 2022-04-22 | 百果园技术(新加坡)有限公司 | User category identification method and device, electronic equipment and readable storage medium |
| CN114626701A (en) * | 2022-02-28 | 2022-06-14 | 中国建设银行股份有限公司 | Community risk early warning method and device, computer equipment and storage medium |
| CN115017155A (en) * | 2022-05-25 | 2022-09-06 | 厦门市美亚柏科信息股份有限公司 | Method, device and storage medium for determining user intimacy |
| CN115098508A (en) * | 2022-07-04 | 2022-09-23 | 成都秦川物联网科技股份有限公司 | Smart city check list generation method, system and device based on Internet of things |
| CN115797044A (en) * | 2022-11-15 | 2023-03-14 | 东方微银科技股份有限公司 | Credit wind control early warning method and system based on cluster analysis |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160267587A1 (en) * | 2015-03-12 | 2016-09-15 | Backed Inc. | Systems and methods for online guarantorship of loans |
| CN106530078A (en) * | 2016-11-29 | 2017-03-22 | 流量海科技成都有限公司 | Loan risk early warning method and system based on multi-industry data |
| CN107194623A (en) * | 2017-07-20 | 2017-09-22 | 深圳市分期乐网络科技有限公司 | A kind of discovery method and device of clique's fraud |
| CN107292509A (en) * | 2017-06-16 | 2017-10-24 | 兴业数字金融服务（上海）股份有限公司 | A kind of enterprise's credit risk early-warning monitoring method |
| CN109360084A (en) * | 2018-09-27 | 2019-02-19 | 平安科技（深圳）有限公司 | Appraisal procedure and device, storage medium, the computer equipment of reference default risk |
| CN109636565A (en) * | 2018-10-16 | 2019-04-16 | 深圳壹账通智能科技有限公司 | Processing method, device, equipment and the computer readable storage medium of risk data |
| CN109685355A (en) * | 2018-12-19 | 2019-04-26 | 重庆百行智能数据科技研究院有限公司 | Business risk recognition methods, device and computer readable storage medium |
| CN110009093A (en) * | 2018-12-07 | 2019-07-12 | 阿里巴巴集团控股有限公司 | For analyzing the nerve network system and method for relational network figure |
| CN110135853A (en) * | 2019-04-25 | 2019-08-16 | 阿里巴巴集团控股有限公司 | Clique's user identification method, device and equipment |
| CN110362639A (en) * | 2019-06-27 | 2019-10-22 | 上海淇馥信息技术有限公司 | A kind of method for prewarning risk, device, electronic equipment calculating analysis based on demographic associations |
| CN110647590A (en) * | 2019-09-23 | 2020-01-03 | 税友软件集团股份有限公司 | Target community data identification method and related device |
| CN110659799A (en) * | 2019-08-14 | 2020-01-07 | 深圳壹账通智能科技有限公司 | Attribute information processing method and device based on relational network, computer equipment and storage medium |
-
2020
- 2020-02-20 CN CN202010105184.3A patent/CN111340611B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160267587A1 (en) * | 2015-03-12 | 2016-09-15 | Backed Inc. | Systems and methods for online guarantorship of loans |
| CN106530078A (en) * | 2016-11-29 | 2017-03-22 | 流量海科技成都有限公司 | Loan risk early warning method and system based on multi-industry data |
| CN107292509A (en) * | 2017-06-16 | 2017-10-24 | 兴业数字金融服务（上海）股份有限公司 | A kind of enterprise's credit risk early-warning monitoring method |
| CN107194623A (en) * | 2017-07-20 | 2017-09-22 | 深圳市分期乐网络科技有限公司 | A kind of discovery method and device of clique's fraud |
| CN109360084A (en) * | 2018-09-27 | 2019-02-19 | 平安科技（深圳）有限公司 | Appraisal procedure and device, storage medium, the computer equipment of reference default risk |
| CN109636565A (en) * | 2018-10-16 | 2019-04-16 | 深圳壹账通智能科技有限公司 | Processing method, device, equipment and the computer readable storage medium of risk data |
| CN110009093A (en) * | 2018-12-07 | 2019-07-12 | 阿里巴巴集团控股有限公司 | For analyzing the nerve network system and method for relational network figure |
| CN109685355A (en) * | 2018-12-19 | 2019-04-26 | 重庆百行智能数据科技研究院有限公司 | Business risk recognition methods, device and computer readable storage medium |
| CN110135853A (en) * | 2019-04-25 | 2019-08-16 | 阿里巴巴集团控股有限公司 | Clique's user identification method, device and equipment |
| CN110362639A (en) * | 2019-06-27 | 2019-10-22 | 上海淇馥信息技术有限公司 | A kind of method for prewarning risk, device, electronic equipment calculating analysis based on demographic associations |
| CN110659799A (en) * | 2019-08-14 | 2020-01-07 | 深圳壹账通智能科技有限公司 | Attribute information processing method and device based on relational network, computer equipment and storage medium |
| CN110647590A (en) * | 2019-09-23 | 2020-01-03 | 税友软件集团股份有限公司 | Target community data identification method and related device |
Non-Patent Citations (1)
| Title |
|---|
| 王李冬 等: "\"大规模社交网络重叠社区发现技术综述\"", 《杭州师范大学学报(自然科学版)》, vol. 15, no. 3, pages 1 - 6 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111915426A (en) * | 2020-08-04 | 2020-11-10 | 中投国信(北京)科技发展有限公司 | Method and system for distinguishing credit users based on graph calculation and community division algorithm |
| CN111951104A (en) * | 2020-08-24 | 2020-11-17 | 上海银行股份有限公司 | Risk conduction early warning method based on associated graph |
| CN112150270A (en) * | 2020-09-18 | 2020-12-29 | 中国建设银行股份有限公司 | Method, device, equipment and computer readable medium for monitoring vehicle information |
| CN113129127A (en) * | 2021-04-21 | 2021-07-16 | 建信金融科技有限责任公司 | Early warning method and device |
| CN113409139A (en) * | 2021-07-27 | 2021-09-17 | 深圳前海微众银行股份有限公司 | Credit risk identification method, apparatus, device, and program |
| CN113409139B (en) * | 2021-07-27 | 2024-05-28 | 深圳前海微众银行股份有限公司 | Credit risk identification method, apparatus, device and program |
| CN114386488A (en) * | 2021-12-22 | 2022-04-22 | 百果园技术(新加坡)有限公司 | User category identification method and device, electronic equipment and readable storage medium |
| CN114240215A (en) * | 2021-12-22 | 2022-03-25 | 中国建设银行股份有限公司 | User loss of contact grade acquisition method and device and storage medium |
| CN114626701A (en) * | 2022-02-28 | 2022-06-14 | 中国建设银行股份有限公司 | Community risk early warning method and device, computer equipment and storage medium |
| CN115017155A (en) * | 2022-05-25 | 2022-09-06 | 厦门市美亚柏科信息股份有限公司 | Method, device and storage medium for determining user intimacy |
| CN115098508A (en) * | 2022-07-04 | 2022-09-23 | 成都秦川物联网科技股份有限公司 | Smart city check list generation method, system and device based on Internet of things |
| CN115797044A (en) * | 2022-11-15 | 2023-03-14 | 东方微银科技股份有限公司 | Credit wind control early warning method and system based on cluster analysis |
| CN115797044B (en) * | 2022-11-15 | 2024-03-29 | 东方微银科技股份有限公司 | Credit wind control early warning method and system based on cluster analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111340611B (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111340611B (en) | Risk early warning method and device | |
| US20190259095A1 (en) | Determining present and future virtual balances for a client computing device | |
| CN110119413A (en) | The method and apparatus of data fusion | |
| US11227217B1 (en) | Entity transaction attribute determination method and apparatus | |
| Li et al. | Credit scoring by incorporating dynamic networked information | |
| CN113570396A (en) | Time series data abnormity detection method, device, equipment and storage medium | |
| CN112950359B (en) | User identification method and device | |
| CN113487410A (en) | Credit granting management method and device, electronic equipment and computer readable medium | |
| CN112581270A (en) | Risk account identification method and device, electronic equipment and storage medium | |
| CN114880482A (en) | Graph embedding-based relation graph key personnel analysis method and system | |
| CN118096170A (en) | Risk prediction method and apparatus, device, storage medium, and program product | |
| CN112734352A (en) | Document auditing method and device based on data dimensionality | |
| CN112231299B (en) | Method and device for dynamically adjusting feature library | |
| CN111245815B (en) | Data processing method and device, storage medium and electronic equipment | |
| CN117113222A (en) | Data analysis model generation method and device and electronic equipment | |
| CN114239985A (en) | Exchange rate prediction method and device, electronic equipment and storage medium | |
| CN114897607A (en) | Data processing method and device for product resources, electronic equipment and storage medium | |
| CN114331540A (en) | Method for training model, method and device for determining asset valuation | |
| CN113159937A (en) | Method and device for identifying risks and electronic equipment | |
| CN113129127A (en) | Early warning method and device | |
| CN112116253A (en) | Method, device and system for selecting central mesh point | |
| CN113779335A (en) | Information generation method and device, electronic equipment and computer readable medium | |
| CN118245341B (en) | Service model switching method, device, electronic equipment and computer readable medium | |
| CN115496393A (en) | Abnormality detection method, abnormality detection device, electronic apparatus, and computer-readable medium | |
| CN112101397B (en) | Method and device for predicting book weight interval |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220926 Address after: 25 Financial Street, Xicheng District, Beijing 100033 Applicant after: CHINA CONSTRUCTION BANK Corp. Address before: 25 Financial Street, Xicheng District, Beijing 100033 Applicant before: CHINA CONSTRUCTION BANK Corp. Applicant before: Jianxin Financial Science and Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |