CN111489254A - Credit risk assessment intelligent engine system based on historical credit big data - Google Patents
Credit risk assessment intelligent engine system based on historical credit big data Download PDFInfo
- Publication number
- CN111489254A CN111489254A CN202010289207.0A CN202010289207A CN111489254A CN 111489254 A CN111489254 A CN 111489254A CN 202010289207 A CN202010289207 A CN 202010289207A CN 111489254 A CN111489254 A CN 111489254A
- Authority
- CN
- China
- Prior art keywords
- credit
- risk
- data
- running water
- intelligent engine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012502 risk assessment Methods 0.000 title claims abstract description 22
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 100
- 238000004458 analytical method Methods 0.000 claims abstract description 29
- 238000005206 flow analysis Methods 0.000 claims abstract description 8
- 238000004457 water analysis Methods 0.000 claims abstract description 7
- 230000003993 interaction Effects 0.000 claims abstract 2
- 238000012546 transfer Methods 0.000 claims description 20
- 238000000034 method Methods 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 8
- 239000002131 composite material Substances 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 3
- 238000012797 qualification Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011217 control strategy Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 239000000463 material Substances 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
- 238000011160 research Methods 0.000 description 1
- 229910052709 silver Inorganic materials 0.000 description 1
- 239000004332 silver Substances 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Databases & Information Systems (AREA)
- Accounting & Taxation (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Finance (AREA)
- Technology Law (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a credit risk assessment intelligent engine system based on historical credit big data, which comprises a pipeline analysis intelligent engine, a credit decision intelligent engine and a risk suggestion intelligent engine, wherein the pipeline analysis intelligent engine, the credit decision intelligent engine and the risk suggestion intelligent engine are respectively in data interaction with a data center; the running water analysis intelligent engine is used for acquiring running water data of a user and analyzing the running water data so as to judge and mark the type of the running water; the credit decision intelligent engine acquires user information and an output result of the flow analysis intelligent engine, and inputs the user information and the output result into a preset algorithm to obtain a decision whether the user passes or not; and the risk suggestion letter intelligent engine acquires a received historical credit data set, receives the output result of the pipelining analysis intelligent engine and the output result of the risk suggestion letter engine, and inputs the output result as input data into a preset rule so as to output risk suggestions.
Description
Technical Field
The invention relates to a bank wind control technology, in particular to a credit risk assessment intelligent engine system based on historical credit big data.
Background
Wind control is the core of financial business, and the wind control core is embodied in credit risk assessment. In the big data era, the infrastructure of the social credit system is continuously perfected, all levels of industrial and commercial management departments and administrative and judicial functional departments regularly disclose relevant credit information which takes industrial and commercial enterprises or individuals as main bodies on their credit platforms, and each large financial institution vigorously builds a financial credit information sharing platform. How to efficiently and accurately utilize the data to support credit risk assessment becomes one of the research cores of various large financial institutions.
In past credit risk approval work, traditional financial institutions often relied on the experience of creditors. For example, for the original running bill of a bank, a lot of people are relied on to extract relevant data from the complicated running water; aiming at the historical credit information of the client, such as judicial complaints, historical industrial and commercial changes, historical overdue conditions and the like, whether the client meets loan conditions or not is manually checked according to a wind control strategy. Such an approach has the following problems:
1) lack of effective technical assistance, time and labor consumption, long response time and easy occurrence of human errors.
2) There are risk points that are not black or white and that the individual treatment criteria will not be consistent. Some people consider not to be a risk point, and another person may be a risk point.
Therefore, for analysis processing of historical credit data of banks, individuals and enterprises, a more intelligent system based on rules is still needed to reduce manual intervention, realize automatic operation and help financial institutions improve credit risk assessment efficiency.
Disclosure of Invention
The invention aims to provide a credit risk assessment intelligent engine system based on historical credit big data, which can carry out intelligent risk assessment based on the historical credit data of individuals and enterprises.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
the credit risk assessment intelligent engine system based on the historical credit big data comprises three parts: the system comprises a flow analysis intelligent engine, a credit decision intelligent engine and a risk suggestion intelligent engine.
1) The running water analysis intelligent engine is used for acquiring running water data of a user and analyzing the running water data so as to judge and mark the type of the running water;
2) the credit decision intelligent engine acquires user information and an output result of the flow analysis intelligent engine, and inputs the user information and the output result into a preset algorithm to obtain a decision whether the user passes or not;
3) and the risk suggestion letter intelligent engine acquires a received historical credit data set, receives the output result of the pipelining analysis intelligent engine and the output result of the risk suggestion letter engine, and inputs the output result as input data into a preset rule so as to output risk suggestions.
As a preferred mode, the workflow of the pipeline analysis intelligent engine is as follows:
(1) acquiring original bank flow data uploaded by a user at a terminal;
(2) the flow analysis intelligent engine receives an analysis command and acquires credit subject association data from the data center;
(3) after the credit principal associated data is received, the intelligent engine executes a calculation command and begins to analyze the flow;
(4) and outputting an analysis result, wherein the analysis result comprises a monthly running water index statistical table and a bank running water bill with a category label obtained by a classification marking method.
As a preferred solution, the category label includes but is not limited to: whether the running water is valid, loan income, suspected cash register income, the running water of the same-name transfer accounts, the running water of posting accounts, asset income, running water, suspected loan payment expenditure, to public transfer, to private transfer and cash withdrawal expenditure.
As a preferred technical solution, the classification labeling method is as follows:
and presetting a keyword corresponding to the category label in the system, if the input bank flow data contains the keyword, marking the keyword as the category label corresponding to the keyword, and outputting the keyword.
As a preferred technical solution, the credit intelligence decision engine works as follows:
(1) receiving from the data center customer historical credit data relating to customer credit qualification metrics;
(2) receiving an output result of the intelligent engine for the running water analysis;
(3) after the intelligent decision engine receives the data, outputting a result through an intelligent decision method;
the step (2) is an unnecessary step and may be omitted.
As a preferred technical solution, the intelligent decision method includes the following processes:
t1, presetting a risk judgment formula by the system based on the risk factors, wherein the risk judgment formula takes one or more as a collection, each collection corresponds to a decision conclusion, and the risk factors are credit risk association dimensions;
t2: acquiring an actual risk factor from the data obtained in the step (1) and/or the step (2);
t3: and inputting the actual risk factor as input data into a risk judgment formula, and outputting a decision conclusion corresponding to the collection to which the risk judgment formula belongs.
As a preferred solution, the credit decision intelligence engine is further configured to output a composite credit score.
As a preferred technical solution, the method for outputting the composite credit score includes:
the system presets weight for the risk factor, creates a judgment formula based on the risk factor, and gives a score to a judgment result corresponding to the judgment formula;
receiving the actual risk factor, inputting a judgment formula to obtain the score of the credit subject, and outputting a comprehensive credit score by the system by using a comprehensive credit score calculation formula, wherein the comprehensive credit score calculation formula is as follows:
s ═ a × X + N; in the formula, S represents credit score, A represents judgment result score, X represents weight, and N represents N A X X.
As a preferred technical scheme, the operation of the risk suggestion letter intelligent engine comprises the following processes:
(1) the system presets risk factors and the value ranges of the risk factors, and a preset conclusion is made if the risk factors are located in the value ranges;
(2) receiving a historical credit data set, receiving a streamline analysis intelligent engine output result, receiving risk suggestion letter engine analysis data, and obtaining an actual risk factor of a credit principal from the analysis data;
(3) and inputting the actual risk factor into a value range, obtaining a conclusion and outputting the conclusion.
Compared with the prior art, the invention has the following beneficial effects:
the invention provides a set of pipelining analysis intelligent engine based on original bank pipelining statement, which accurately analyzes original pipelining data of a bank, efficiently identifies and outputs key indexes such as effective bank pipelining, borrowing income, suspected cash register income, suspected repayment expenditure, public transfer, private transfer, cash withdrawal expenditure and the like, accurately measures the actual capital pipelining condition of a client, and automatically identifies risk indexes in the pipelining for credit risk assessment;
the invention provides a credit decision intelligent engine based on personal and enterprise historical credit big data, which automatically judges whether a customer accords with a wind control policy according to the historical credit data (mainly online big data) of the customer, the operation data (mainly bank flow data) and the like by combining a product policy, gives out a comprehensive credit score and completes automatic approval;
the invention provides a credit risk suggestion letter intelligent engine integrating personal and enterprise historical credit big data and original bank running statement data, which automatically gives decision suggestions and points to be manually verified for each risk factor according to client historical credit data (mainly online data), operating data (mainly bank running data) and the like in combination with product policies, so that offline full-scale (including telephone full-scale) is given a target, and risks contained in online data are furthest checked.
Drawings
FIG. 1 is a block diagram of the system of the present invention;
fig. 2 is a flow chart of the system.
Detailed Description
The invention aims to overcome the defects of the prior art and provides a credit risk assessment intelligent engine system based on historical credit big data, and the invention is further detailed in the following by combining with the embodiment.
Examples
The credit risk assessment intelligent engine system based on the historical credit big data comprises a stream analysis intelligent engine, a credit decision intelligent engine and a risk suggestion intelligent engine;
the running water analysis intelligent engine is used for acquiring running water data of a user and analyzing the running water data so as to judge the attribute of the running water;
the credit decision intelligent engine acquires the user information and the related flow attribute thereof, inputs the user information and the related flow attribute into a preset rule, and acquires the decision whether the user passes and the comprehensive credit score;
and the risk suggestion letter intelligent engine acquires a received historical credit data set, receives the output result of the pipelining analysis intelligent engine and the output result of the risk suggestion letter engine, and accordingly outputs the risk suggestion.
First, the workflow of the flow analysis intelligent engine is explained, and the process is as follows:
(a1) acquiring an original bank flow form uploaded by a user at a terminal, wherein data contained in the form comprises but is not limited to: transaction date, transaction amount, balance, transaction location, transaction mode, transaction channel, transaction summary, transaction remark, name of the opposite party, account number of the opposite party, bank of the bank, account of the bank and account type;
(a2) the stream analysis intelligent engine receives an analysis command and acquires credit subject association data from the data center, wherein the association data comprises but is not limited to personal names, associated natural person names and associated enterprise names;
(a3) after the credit principal associated data is received, the intelligent engine executes a calculation command and begins to analyze the flow;
(a4) and outputting an analysis result, wherein the analysis result comprises a monthly running water index statistical table and a bank running water bill with a category label obtained by a classification marking method. Category labels include, but are not limited to: whether the running water is valid, loan income, suspected cash register income, the running water of the same-name transfer accounts, the running water of posting accounts, asset income, running water, suspected loan payment expenditure, to public transfer, to private transfer and cash withdrawal expenditure.
The effective pipelining judgment means that not all income banks in the original bank pipelining are identified, and only the effective pipelining which can truly reflect the stable and continuous income sources of customers is a key index for identifying the income level of the customers;
the debit and credit income judgment refers to the debit and credit income used for judging the original running water of the bank, and the debit and credit income is an important index for measuring the debt condition and the debit and credit habit of the client;
the posting flow judgment is used for judging the posting flow in the original flow of the bank. The posting running water is fast-forward running water with fast-out speed and simultaneous-in running water with simultaneous-out speed, and is generally brushed running water, so that the running water is high in virtual condition and is a main part in invalid running water;
the homonymous transfer assembly line is used for judging the homonymous transfer assembly line in the original assembly line of the bank. The homonymous transfer pipelining refers to transfer pipelining of a person associated with a pipelining account main body, and is a main part in invalid pipelining;
the positive flowing water judgment is used for judging positive flowing water in the original flowing water of the bank. The positive running water refers to running water which returns the expenditure running water after the transaction fails and belongs to invalid running water;
the suspected cash register judgment is used for judging the suspected cash register flowing water in the original flowing water of the bank. The suspected cash flow refers to suspected water flow which is obtained by improperly utilizing a POS machine to cash the amount in the credit card to a savings card and belongs to invalid water flow;
the asset income determination is used for determining the asset income in the original flowing water of the bank. The property income refers to income obtained by household owned mobile property (such as bank deposit, securities) and real property (such as house, vehicle, collection, etc.). Including interest, rent, patent income gained by giving property rights of use; the dividend income, the property value-added income and the like obtained by property operation have extremely important value for judging the asset allocation condition of the client;
the other invalid income judgment is used for judging other unclassified invalid running water in the original running water of the bank;
the suspected repayment expense judgment is used for judging the suspected repayment expense in the original flowing water of the bank. The suspected loan payment is an important index for measuring the client liability, and is particularly beneficial to digging a large amount of hidden liabilities;
the revolution-out judgment is used for judging the revolution-out flow in the original flow of the bank, so that the fund flow condition of a client to a company account can be measured, and the judgment of the type of a main transaction opponent of the client is facilitated;
the private-pair export judgment is used for judging the private-pair export stream in the original bank stream, can measure the fund outflow condition of a customer to a personal account, and is beneficial to judging the type of a main transaction opponent of the customer;
the cash withdrawal expenditure judgment is used for judging the cash withdrawal running water in the original running water of the bank, and can measure the cash use condition in the daily operation of a client;
the average daily deposit balance calculation is used for calculating the average daily deposit balance of the running water of the client bank. The average daily deposit balance, namely the average daily deposit amount of the account, is an index for measuring the daily service level of the account by a financial department, and is an important index for measuring the repayment capacity of a client.
In the step (a4), the classification labeling method is to preset a keyword corresponding to the category label in the system, and if the input bank flow data contains the keyword, the input bank flow data is labeled with the category label corresponding to the keyword and output the category label.
Examples are as follows:
r1: flushing positive flowing water: if the bank running water contains any one of the key words of 'rushing to the right', 'rushing to the bill', 'account number error', the 'counter party name', 'transaction remark', 'transaction abstract', 'transaction channel', 'transaction mode' or 'transaction place' column, the 'rushing to the right' running water is marked;
r2: income borrowing and lending: if the bank running water contains any one of the keywords "borrow" and "loan", the columns of "opposite party name", "trade remark", "trade summary", "trade channel", "trade mode" or "trade place" are marked as "borrow income";
r3: posting running water: in the bank flow, if the 'opposite side name' of the income flow and the 'opposite side name', 'trade remark', 'trade abstract', 'trade channel', 'trade mode' or 'trade place' of the expenditure flow in the previous hour and the next hour can be accurately matched, the income is marked as 'posting flow'; in the next hour, if the two running water sums are 0, the previous running water is the entry running water, and the next running water account does not contain the name of the credit subject associated object, the previous entry running water is marked as 'posting running water'; if a plurality of items of the outgoing running water are the same as the amount of the current item of the incoming running water, and one item of the incoming running water meets the conditions, the items of the outgoing running water are marked as 'posting running water'.
R4: and (3) transferring the account by the same name: (1) in the running water, the running water containing the name of the credit subject associated object in the columns of the name of the opposite party, the remark of the transaction, the summary of the transaction, the channel of the transaction, the mode of the transaction or the place of the transaction is marked as the running water of the same-name transfer; (2) in the running water, the running water of the 'opposite side name', 'transaction remark', 'transaction summary', 'transaction channel', 'transaction mode' or 'transaction place' column contains the key word 'borrow', and simultaneously contains the key word 'customer name/spouse name/customer company name/customer external investment or external job company name/customer company shareholder name/assigned customer relatives name/legal representative of the bank card owner/legal representative' is marked as 'same name transfer running water';
r5: suspected set of running water: in the running water, the running water of the name of the opposite party, the remarks of the transaction, the summary of the transaction, the channel of the transaction, the mode of the transaction or the place of the transaction, which contains the key words and the merge, is marked as 'suspected cash register';
r6: and (4) income of assets: in the running water, the running water of the keyword 'silver certificate transfer' listed in the 'name of the opposite party', 'trade remark', 'trade abstract', 'trade channel', 'trade mode' or 'trade place' is marked as 'asset income';
r7: other invalid revenues: in the running water, the running water of the name of the opposite party, the remarks of the transaction, the summary of the transaction, the channel of the transaction, the mode of the transaction or the place of the transaction, which contains the key word "overdraft and recollection" or "claim money", is marked as "other invalid income";
r8: the third party pays for the valid flow: in the running water, the running water of the key word 'transfer cash', which is listed as 'opposite side name', 'trade remark', 'trade abstract', 'trade channel', 'trade mode' or 'trade place', is marked as 'effective running water';
r9: effective flowing water: for all income item streams, streams not labeled with the above-described markers R1-R8 are labeled as "valid streams";
for the running water with the 'transaction amount' less than or equal to 0, each original running water is marked in sequence according to the steps, except for the rule R4(2) and the rule R8, the rules arranged in the back can not cover the marks of the rules in the front, and the rule R4(2) and the rule R8 can cover the running water marks of the rules in the front.
R10: and (3) revolution and play: in the running water, the column character of the opposite side name is more than or equal to 5, and the running water containing the keywords company and group and not containing the keyword financial payment is marked as 'going out of the pair of the revolution'; in the running water, the "opposite side name" column contains keywords "company" and "group", and also contains keywords "property payment through-" "property payment through: "the running water mark is" go out to the revolution ";
r11: and (4) private roll-out: in the running water, the column character of the name of the opposite party is more than 0 and less than or equal to 6, or the running water containing the keywords 'Paibao' and 'WeChat' is marked as 'Paiping out';
r12: and (3) paying out: in the running water, the columns of 'opposite side name', 'trade remark', 'trade abstract', 'trade channel', 'trade mode' or 'trade place' contain the key word 'current payment', and the running water without 'current payment transfer' is marked as 'cash-out expenditure';
and for the pipeline with the 'transaction amount' larger than 0, sequentially marking each original pipeline according to the steps, wherein the 'pair private roll-out' mark cannot cover the 'pair revolution roll-out' mark, and the 'withdrawal' mark can cover the 'pair private roll-out' and 'pair revolution roll-out' marks.
The credit intelligence decision engine works as follows:
(b1) receiving from the data center customer historical credit data related to customer credit qualification metrics including, but not limited to: personal basic information, consumption label information, external investment information, external job information, judicial complaint information, enterprise industry and commerce information, tax information, judicial information and intellectual property information;
(b2) receiving an output result of the intelligent engine for the running water analysis;
(b3) the intelligent decision engine outputs a result by analyzing after receiving the data; the step (b2) is not essential and can be omitted.
Credit intelligence decision engine parsing includes the following processes:
and constructing a risk factor. Defining a credit risk association dimension as a risk factor, selecting the risk factor according to the actual wind control policy of a credit product, wherein the risk factor is taken from the steps (b1) and (b2), constructing a risk factor set, classifying the risk factor set, and dividing the risk factor set into four categories by taking enterprises and individuals as objects, wherein the four categories are detailed in the following table;
the rule attribute comprises two types of condition relation and condition range, the condition relation is divided into three types, namely, all the conditions are met, one type of condition is met, and two types of condition range are met, the condition range is divided into two types, namely interval type and single condition, the interval type comprises interval type which is greater than F1 ═ and less than F1 ≦ and the like, the single condition is that the condition is equal to, not equal to, <, ∈,
", wherein F1 is a risk factor;
and (5) constructing a rule set. The rule set is a set of rules, each rule is composed of a risk factor, a judgment condition and a corresponding output conclusion, and is simply expressed as "F1 > -" X, if the rule passes, F1 < X, then rejection is performed, F1 is the risk factor, "> -" is the judgment condition, "X" is the value range of the risk factor satisfying the rule, and "pass" is the conclusion. Such as: age ═ 25, customer passed, age < 25, customer rejected; if rule 1 is R1, rule 2 is R2, and rule 3 is R3, then the customer passes if three rules are satisfied simultaneously, or the customer passes if one of the three rules is satisfied, then the rule set is considered; according to the actual wind control strategy of the product, the rule set can be flexibly adjusted;
connecting a plurality of rules or rule sets, constructing a complete credit decision tree,
and constructing a scoring card. The scoring card is realized by endowing different weight scores to different values of the risk factors, and can be customized and adjusted according to the actual wind control rule of the credit product, wherein the scoring card is taken as an example of the following customers:
the scoring card rules are as follows:
1. age, weight coefficient 0.05, 55 > age > 25, score 100; age < 25, score 0; age greater than 55, score 0;
2. whether the mobile phone number is real-name or not is judged, the weight coefficient is 0.15, whether the mobile phone number is real-name or not is judged, and the score is 100; whether the mobile phone number is real-name or not is judged, the score is 0, whether the mobile phone number is real-name or not is judged, and the score is 50;
3. the number of executed records, the weight coefficient is 0.2, the number of executed records is 0, and the score is 100; 0 < number of executed records < 3, and score 50; 3 < ═ number of executed records < 5 with a score of 20, number of executed records > -5 with a score of 0;
4. the number of executed records of lost mail is 0.3, the weight coefficient is 0, the number of executed records of lost mail is 0, and the score is 100; the number of the executed records of losing credit is more than 0 and less than 3, and the score is 30; the number of the executed records of losing the information is 3, and the score is 0;
5. the registration age is 0.15, the weight coefficient is less than 1, and the score is 0; 1 & ltregistration year & lt 3, and the score is 50; 3 & ltregistration life & lt 5, and the score is 70; 5 & ltregistration life & lt 15, and the score is 100; registration age > -15with a score of 70;
6. the daily average deposit balance is 0.15, the daily average deposit balance is less than 10000, and the score is 20; 10000 < ═ daily average deposit balance < 50000 and the score is 40; 10000 ≦ daily average deposit balance < 50000; the daily average deposit balance is less than 100000 and the score is 60; the daily average deposit balance is less than 500000, and the score is 60; the average daily deposit balance is 500000 with the score of 100;
then it is determined that,
client SY 1-0.05 x 100+0.15 x 100+0.2 x 100+0.15 x 50+0.15 x 40-73.5;
client SY2 ═ 0.05 × 100+0.15 × 0+0.2 × 20+0.2 × 30+0.15 × 100+0.15 × 60 ═ 39;
respectively obtaining the comprehensive credit scores of the two clients and evaluating the credit risks of the two clients;
the output conclusion comprises decision suggestions whether the client passes or not and the comprehensive credit score.
The working process of the risk suggestion letter intelligent engine is as follows:
(c1) the system presets risk factors and the value ranges of the risk factors, and a preset conclusion is made if the risk factors are within the value ranges;
(c2) receiving a historical credit data set, receiving a streamline analysis intelligent engine output result, receiving risk suggestion letter engine analysis data, and obtaining an actual risk factor of a credit principal from the analysis data;
(c3) and inputting the actual risk factor into a value range, obtaining a conclusion and outputting the conclusion.
The specific implementation case is as follows:
constructing risk factors in the same way as a credit decision intelligent engine;
the rule attribute is constructed in the same way as the credit decision intelligent engine;
and constructing a single decision suggestion rule. The single decision suggestion rule consists of a risk factor, a judgment condition and a corresponding output conclusion, and is simply expressed as that when F1 is equal to X, Y is output, wherein F1 is the risk factor, and when equal to X is the judgment condition under the rule attribute, X is the value range of the risk factor meeting the rule, and Y is the output conclusion and is generally a segment of text content. For example, the rule is that when the registration age is < 1, then "the enterprise registration age is less than 1 year, the customer needs to provide other certification material that can certify that the customer has actually operated for one year, otherwise the customer is rejected". All the suggestion rules are based on the wind control policy of the product, and corresponding risk response suggestions are output according to actual return data of the client;
all the conclusions are suggested to be listed item by item, and are gathered in a Word or PDF document to be output.
Specifically, the method for operating and using the program code of the present invention includes:
1) running entirely on the user's local computing device
2) Partly performed on user equipment
3) Executed as a stand-alone software package
4) Execution as a set of WEB service calls
When implemented as a suite of Web services, may be connected to the user computing devices via any form of network, such as a local area network (L AN) or a Wide Area Network (WAN).
The invention is well implemented in accordance with the above-described embodiments. It should be noted that, based on the above structural design, in order to solve the same technical problems, even if some insubstantial modifications or colorings are made on the present invention, the adopted technical solution is still the same as the present invention, and therefore, the technical solution should be within the protection scope of the present invention.
Claims (9)
1. The credit risk assessment intelligent engine system based on the historical credit big data is characterized by comprising a flow analysis intelligent engine, a credit decision intelligent engine and a risk suggestion intelligent engine which are respectively in data interaction with a data center;
the running water analysis intelligent engine is used for acquiring running water data of a user and analyzing the running water data so as to judge and mark the type of the running water;
the credit decision intelligent engine acquires user information and an output result of the flow analysis intelligent engine, and inputs the user information and the output result into a preset algorithm to obtain a decision whether the user passes or not;
and the risk suggestion letter intelligent engine acquires a received historical credit data set, receives the output result of the pipelining analysis intelligent engine and the output result of the risk suggestion letter engine, and inputs the output result as input data into a preset rule so as to output risk suggestions.
2. The credit risk assessment intelligence engine system based on historical credit big data, according to claim 1, wherein the workflow of the pipeline parsing intelligence engine is as follows:
(1) acquiring original bank flow data uploaded by a user at a terminal;
(2) the flow analysis intelligent engine receives an analysis command and acquires credit subject association data from the data center;
(3) after the credit principal associated data is received, the intelligent engine executes a calculation command and begins to analyze the flow;
(4) and outputting an analysis result, wherein the analysis result comprises a monthly running water index statistical table and a bank running water bill with a category label obtained by a classification marking method.
3. The intelligent engine system for credit risk assessment based on historical credit big data according to claim 2, wherein category labels include but are not limited to: whether the running water is valid, loan income, suspected cash register income, the running water of the same-name transfer accounts, the running water of posting accounts, asset income, running water, suspected loan payment expenditure, to public transfer, to private transfer and cash withdrawal expenditure.
4. The intelligent engine system for credit risk assessment based on historical credit big data according to claim 3, wherein the classification labeling method is as follows:
and presetting a keyword corresponding to the category label in the system, if the input bank flow data contains the keyword, marking the keyword as the category label corresponding to the keyword, and outputting the keyword.
5. The intelligent engine system for credit risk assessment based on historical credit big data according to claim 1, wherein [ s1] works as follows:
(1) receiving from the data center customer historical credit data relating to customer credit qualification metrics;
(2) receiving an output result of the intelligent engine for the running water analysis;
(3) after the intelligent decision engine receives the data, outputting a result through an intelligent decision method;
the step (2) is an unnecessary step and may be omitted.
6. The credit risk assessment intelligence engine system based on historical credit big data, according to claim 5, characterized by, the intelligence decision method comprises the following processes:
t1, presetting a risk judgment formula by the system based on the risk factors, wherein the risk judgment formula takes one or more as a collection, each collection corresponds to a decision conclusion, and the risk factors are credit risk association dimensions;
t2: acquiring an actual risk factor from the data obtained in the step (1) and/or the step (2);
t3: and inputting the actual risk factor as input data into a risk judgment formula, and outputting a decision conclusion corresponding to the collection to which the risk judgment formula belongs.
7. The credit risk assessment intelligence engine system based on historical credit big data, as claimed in claim 1, wherein the credit decision intelligence engine is further configured to output a composite credit score.
8. The intelligent engine system for credit risk assessment based on historical credit big data according to claim 7, wherein said method for outputting composite credit score is:
the system presets weight for the risk factor, creates a judgment formula based on the risk factor, and gives a score to a judgment result corresponding to the judgment formula;
receiving the actual risk factor, inputting a judgment formula to obtain the score of the credit subject, and outputting a comprehensive credit score by the system by using a comprehensive credit score calculation formula, wherein the comprehensive credit score calculation formula is as follows:
s = a X + N; in the formula, S represents credit score, A represents judgment result score, X represents weight, and N represents N A X X.
9. The credit risk assessment intelligence engine system based on historical credit big data, according to claim 1, characterized by a risk advice intelligence engine whose work comprises the following processes:
(1) the system presets risk factors and the value ranges of the risk factors, and a preset conclusion is made if the risk factors are within the value ranges;
(2) receiving a historical credit data set, receiving a streamline analysis intelligent engine output result, receiving risk suggestion letter engine analysis data, and obtaining an actual risk factor of a credit principal from the analysis data;
(3) and inputting the actual risk factor into a value range, obtaining a conclusion and outputting the conclusion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010289207.0A CN111489254A (en) | 2020-04-14 | 2020-04-14 | Credit risk assessment intelligent engine system based on historical credit big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010289207.0A CN111489254A (en) | 2020-04-14 | 2020-04-14 | Credit risk assessment intelligent engine system based on historical credit big data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111489254A true CN111489254A (en) | 2020-08-04 |
Family
ID=71811824
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010289207.0A Pending CN111489254A (en) | 2020-04-14 | 2020-04-14 | Credit risk assessment intelligent engine system based on historical credit big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111489254A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112287237A (en) * | 2020-12-24 | 2021-01-29 | 银联智惠信息服务（上海）有限公司 | Transaction data analysis method and device for third-party transaction platform and terminal |
| CN112783963A (en) * | 2021-03-17 | 2021-05-11 | 上海数喆数据科技有限公司 | Enterprise offline and online multi-source data integration method and device based on business circle division |
| CN112991034A (en) * | 2020-11-30 | 2021-06-18 | 重庆誉存大数据科技有限公司 | Model-based mini-enterprise credit assessment method, equipment and storage medium |
| CN113129021A (en) * | 2021-05-18 | 2021-07-16 | 中国银行股份有限公司 | Block chain-based method and device for preventing malicious overdraft of credit card |
| CN113988671A (en) * | 2021-11-02 | 2022-01-28 | 城云科技（中国）有限公司 | Enterprise credit risk assessment method, device and application |
| CN116225417A (en) * | 2023-05-08 | 2023-06-06 | 无锡锡商银行股份有限公司 | Financial platform decision engine management system and method based on big data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109064312A (en) * | 2018-07-17 | 2018-12-21 | 深圳汇加优运互联数据服务有限公司 | A kind of loan for vehicle risk control method, electronic device and storage medium |
| CN109146663A (en) * | 2018-07-20 | 2019-01-04 | 中国邮政储蓄银行股份有限公司 | The processing method and system of flowing water information |
| CN109767323A (en) * | 2018-12-28 | 2019-05-17 | 交通银行股份有限公司 | Capital chain generation method and device based on transaction journal |
| CN110135700A (en) * | 2019-04-23 | 2019-08-16 | 北京淇瑀信息科技有限公司 | Credit Risk Assessment method and device based on expandtabs data |
-
2020
- 2020-04-14 CN CN202010289207.0A patent/CN111489254A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109064312A (en) * | 2018-07-17 | 2018-12-21 | 深圳汇加优运互联数据服务有限公司 | A kind of loan for vehicle risk control method, electronic device and storage medium |
| CN109146663A (en) * | 2018-07-20 | 2019-01-04 | 中国邮政储蓄银行股份有限公司 | The processing method and system of flowing water information |
| CN109767323A (en) * | 2018-12-28 | 2019-05-17 | 交通银行股份有限公司 | Capital chain generation method and device based on transaction journal |
| CN110135700A (en) * | 2019-04-23 | 2019-08-16 | 北京淇瑀信息科技有限公司 | Credit Risk Assessment method and device based on expandtabs data |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112991034A (en) * | 2020-11-30 | 2021-06-18 | 重庆誉存大数据科技有限公司 | Model-based mini-enterprise credit assessment method, equipment and storage medium |
| CN112287237A (en) * | 2020-12-24 | 2021-01-29 | 银联智惠信息服务（上海）有限公司 | Transaction data analysis method and device for third-party transaction platform and terminal |
| CN112287237B (en) * | 2020-12-24 | 2021-04-13 | 银联智惠信息服务（上海）有限公司 | Transaction data analysis method and device for third-party transaction platform and terminal |
| CN112783963A (en) * | 2021-03-17 | 2021-05-11 | 上海数喆数据科技有限公司 | Enterprise offline and online multi-source data integration method and device based on business circle division |
| CN113129021A (en) * | 2021-05-18 | 2021-07-16 | 中国银行股份有限公司 | Block chain-based method and device for preventing malicious overdraft of credit card |
| CN113988671A (en) * | 2021-11-02 | 2022-01-28 | 城云科技（中国）有限公司 | Enterprise credit risk assessment method, device and application |
| CN116225417A (en) * | 2023-05-08 | 2023-06-06 | 无锡锡商银行股份有限公司 | Financial platform decision engine management system and method based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111489254A (en) | Credit risk assessment intelligent engine system based on historical credit big data | |
| Unegbu | Theories of accounting: evolution & developments, income-determination and diversities in use | |
| JP5096642B2 (en) | Economic activity index presentation system | |
| Yin et al. | Evaluating the credit risk of SMEs using legal judgments | |
| Frame et al. | Credit scoring and the availability of small business credit in low‐and moderate‐income areas | |
| Franceschetti et al. | Do bankrupt companies manipulate earnings more than the non-bankrupt ones? | |
| Lueg et al. | Does transition to IFRS substantially affect key financial ratios in shareholder-oriented common law regimes? Evidence from the UK | |
| CN109670945B (en) | Comprehensive risk early warning decision platform based on big data | |
| CN111476660A (en) | Intelligent wind control system and method based on data analysis | |
| Flögel et al. | The Banking Systems of Germany, the UK and Spain from a Spatial Perspective: The German Case | |
| CN109242664A (en) | It is a kind of towards the tax risk prediction technique for newly setting up enterprise | |
| CN104103009A (en) | Construction method of database based on credit report | |
| CN109345372A (en) | Credit-graded approach, system and computer readable storage medium | |
| CN112613977A (en) | Personal credit loan admission credit granting method and system based on government affair data | |
| CN113989019A (en) | Method, device, equipment and storage medium for identifying risks | |
| KR20110051463A (en) | Automatic entry generation appartus and method thereof | |
| CN113988726A (en) | Enterprise industry credit evaluation management system based on block chain | |
| Ageeva et al. | The application of digital technologies in financial reporting and auditing | |
| CN113902546A (en) | Credit risk early warning method and system based on knowledge graph | |
| CN114840579B (en) | Hospital internal auditing system | |
| Zhitlukhina et al. | Issues of Falsifying Financial Statements in Terms of Economic Security. | |
| CN101877112A (en) | Economic intelligence forecast | |
| Narayan et al. | State-level politics: Do they influence corporate investment decisions? | |
| CN110532296A (en) | A kind of citizen's credit score assessment system based on the open data of government | |
| CN117011065A (en) | Financial product dynamic matching recommendation method and equipment based on enterprise tag |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200804 |