CN115994818A - Intelligent collection system and method for special assets of bank - Google Patents
Intelligent collection system and method for special assets of bank Download PDFInfo
- Publication number
- CN115994818A CN115994818A CN202211345603.6A CN202211345603A CN115994818A CN 115994818 A CN115994818 A CN 115994818A CN 202211345603 A CN202211345603 A CN 202211345603A CN 115994818 A CN115994818 A CN 115994818A
- Authority
- CN
- China
- Prior art keywords
- data
- special
- tag
- intelligent
- module
- 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
- 238000000034 method Methods 0.000 title claims abstract description 18
- 238000011156 evaluation Methods 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 22
- 238000013139 quantization Methods 0.000 claims abstract description 20
- 238000012544 monitoring process Methods 0.000 claims abstract description 13
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 6
- 238000010276 construction Methods 0.000 claims abstract description 6
- 238000010801 machine learning Methods 0.000 claims description 17
- 238000012795 verification Methods 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 11
- 238000004891 communication Methods 0.000 claims description 10
- 238000002372 labelling Methods 0.000 claims description 10
- 238000012550 audit Methods 0.000 claims description 8
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000013524 data verification Methods 0.000 claims description 7
- 238000005065 mining Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000035515 penetration Effects 0.000 claims description 3
- 230000010354 integration Effects 0.000 abstract description 6
- 238000007726 management method Methods 0.000 description 16
- 239000010931 gold Substances 0.000 description 15
- 229910052737 gold Inorganic materials 0.000 description 15
- 238000011084 recovery Methods 0.000 description 13
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 12
- 230000006870 function Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 238000012360 testing method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001447 compensatory effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000008571 general function Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000007789 sealing Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an intelligent collection system and method for special assets of a bank, comprising the following modules: the customer portrait construction module: providing customer portraits presentation, portrayal tagging, and customer monitoring; and the intelligent decision module: carrying out characteristic quantization on the treatment scheme through data modeling, and giving quantization treatment and reference recommendation of case similarity; and (5) expanding property clue module: displaying and managing suspected property clue data of a user client, and exploring special property clues through artificial intelligence and machine intelligence; an estimation center module: performing intelligent evaluation on the special assets; and a system security module: different security measures are implemented for different security threats. The special asset intelligent collection system provided by the invention greatly improves the data integration speed and the service processing speed.
Description
Technical Field
The invention relates to the technical field of asset collection, in particular to an intelligent collection system and method for special assets of a bank.
Background
Along with the periodical entering of economy into a downlink channel, the complex change of domestic and foreign economic situation, the bad business of banks is continuously increased, and the scale of special assets of each commercial bank is continuously increased. Meanwhile, the special treatment business is influenced by the global new epidemic situation, and the special treatment business is in the situation of increased traffic and more difficult treatment. The traditional manual collection system has difficulty in meeting the requirements of modern banking on efficient and intelligent treatment of special asset businesses, and the urgent information enabling of the special asset businesses improves the convenience of business disposal and improves the efficiency of customer adjustment.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide an intelligent clearing and collecting system and method for special assets of a bank.
The invention provides an intelligent clearing and collecting system for special assets of a bank, which comprises the following modules:
the customer portrait construction module: integrating data related to a specified special asset customer by integrating data in and out of a line, constructing a special asset enterprise customer portrait, and providing customer portrait presentation, portrait tagging and customer monitoring;
and the intelligent decision module: carrying out characteristic quantization on the treatment scheme through data modeling, and giving quantization treatment and reference recommendation of case similarity;
and (5) expanding property clue module: displaying and managing suspected property clue data of a user client, and exploring special property clues through artificial intelligence and machine intelligence;
an estimation center module: performing intelligent evaluation on the special assets, wherein the evaluation center module comprises an evaluation pricing model, an integrated evaluation tool and an evaluation report;
and a system security module: different security measures are implemented for different security threats, including identity authentication, access control, security audit, communication security, and software fault tolerance.
Preferably, the portrait tagging includes configurable tags and custom tags; the configurable label carries out quantization judgment according to the label value, and is obtained through corresponding fields, logical formulas and update frequency calculation of the configurable label; the custom tag is customized by acquiring the access authority user.
Preferably, the intelligent decision module is realized by an expert matching algorithm based on a label weight or an expert matching algorithm based on a machine learning model;
the expert matching algorithm based on the tag weight is to recommend expert opinion for a user by adjusting the tag weight, count the same tag by taking the tag of the portrait of the scheme to be treated as a quantization index of matching degree, recommend the reference of the scheme to be treated by the weighting value of the tag, and calculate the corresponding tag weight according to the weight set by the tag;
the expert matching algorithm based on the machine learning model is a matching algorithm based on the machine learning model, the matching degree is calculated through a matching formula obtained through data set training, the similarity value of the target client and the treatment case is obtained, and the recommended case is obtained through sorting according to the similarity value.
Preferably, the extended property cue module includes:
a clue mining and pushing tool module: the method comprises a traditional mode and a high-order mode, wherein the traditional mode is used for primarily filtering external data through labeling processing external information, and pushing property clues to corresponding households according to different assets; the high-order mode is based on the traditional mode, and the available property clues are further mined through stock penetration analysis and suspected real control people mining;
standard business approval automation module: after business data of traditional retail business enter the system, firstly, according to characteristic analysis of the data, including main body and debt; automatically identifying the data according to the rules, and classifying and approving; the system examines and approves the data according to the rule base, if the rule base is met, the examination and approval is completed, and if the rule base is not met, the manual examination and approval is carried out; aiming at complex nonstandard businesses, the system completes business approval through a multidimensional scoring mechanism and dynamically adjusts according to real-time treatment cases;
intelligent recommended treatment operating scheme module: based on the internal historical transaction cases and the external transaction data, key indexes under different treatment or operation schemes are quantified, and an analysis tool or model is built to generate a clearing scheme.
Preferably, the security audit is that after a user logs in the system, the system automatically records operation marks of the user, including operation users, operation time and IP addresses; the operation system and the database update log information along with the operation of the user, so that the traceability of the operation is ensured.
Preferably, the communication security includes: an HTTPS is used for establishing an information security channel, data verification is carried out in the data transmission process, a double-file storage mode is adopted for data files from a source data area to the system, namely, one data file and one mark file are adopted, verification information of the data files is recorded in the mark file, a special asset collection system carries out data verification on all the data files according to the verification information, files passing verification can be adopted, and errors are reported when verification fails, and the files wait for processing.
Preferably, the software fault tolerance comprises: the special asset collection system application adopts cluster deployment, and when a single application fails, the system automatically identifies and judges and directly forwards the application to an available application server for operation; and providing a main and standby mechanism for the database, and switching the database to the standby machine for operation when the main server fails.
The invention provides an intelligent clearing and collecting method for special assets of a bank, which comprises the following steps:
the customer portrait construction step: integrating data related to a specified special asset customer to construct a special asset enterprise customer portrait by integrating data in and out of a row, and providing customer portrait presentation, portrait tagging and customer monitoring;
intelligent decision step: carrying out characteristic quantization on the treatment scheme through data modeling, and giving quantization treatment and reference recommendation of case similarity;
expanding property clues: displaying and managing suspected property clue data of a user client, and exploring special property clues through artificial intelligence and machine intelligence;
asset assessment: performing intelligent evaluation on the special assets, wherein the evaluation center step comprises an evaluation pricing model, an integrated evaluation tool and an evaluation report;
a security evaluation step: different security measures are implemented for different security threats, including identity authentication, access control, security audit, communication security, and software fault tolerance.
Preferably, the portrait tagging includes configurable tags and custom tags; the configurable label carries out quantization judgment according to the label value, and is obtained through corresponding fields, logical formulas and update frequency calculation of the configurable label; the custom tag is customized by acquiring the access authority user.
Preferably, the intelligent decision step is realized by an expert matching algorithm based on a label weight or an expert matching algorithm based on a machine learning model;
the expert matching algorithm based on the tag weight is to recommend expert opinion for a user by adjusting the tag weight, count the same tag by taking the tag of the portrait of the scheme to be treated as a quantization index of matching degree, recommend the reference of the scheme to be treated by the weighting value of the tag, and calculate the corresponding tag weight according to the weight set by the tag;
the expert matching algorithm based on the machine learning model is a matching algorithm based on the machine learning model, the matching degree is calculated through a matching formula obtained through data set training, the similarity value of the target client and the treatment case is obtained, and the recommended case is obtained through sorting according to the similarity value.
Compared with the prior art, the invention has the following beneficial effects:
1. the special asset intelligent collection system provided by the invention greatly improves the data integration speed and the service processing speed through the information means, the big data analysis, the artificial intelligence and other technologies, can integrate the internal data and the external data and provide technological energization for special asset disposal service, thereby improving the processing efficiency of the special asset service.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the detailed description of non-limiting embodiments, given with reference to the accompanying drawings in which:
FIG. 1 is a technical frame diagram of a special asset intelligent collection system of the invention;
FIG. 2 is a block diagram of the intelligent collection system for special assets of the present invention;
FIG. 3 is a flow chart of a particular asset collection system deployment of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the present invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications could be made by those skilled in the art without departing from the inventive concept. These are all within the scope of the present invention.
Example 1:
the invention discloses an intelligent collection system for special assets of a bank, which comprises the following modules:
the customer portrait construction module: and integrating data related to the specified special asset clients by integrating data in and out of the line to construct the special asset enterprise client image. Through the data of the customer portraits (special resources), business personnel can conveniently check the relevant information of special resource customers, including customer basic information, contract information, guarantee information, (cash) account information, risk early warning information, debt information, mortgage property information, recent transaction opponent information, asset stripping information, (past) collection disposal information, judicial case information, external data and other dimensional information. And meanwhile, the client portrait module supports the image presentation of the enterprise and gold clients, the image tagging of the enterprise and gold clients and the monitoring of the enterprise and gold clients.
1. Image labelling method for enterprise and gold clients
The support of the labeling of the enterprise-gold customer portrait refers to the process of carrying out portrait quantization labeling on the enterprise-gold customer portrait. The collection system provides support for labeling enterprise gold images, and in principle, the system carries out labeled quantitative judgment through index values; for complex logic labeling, the system supports labeling customization. Enterprise gold customer portrait tagging supports 2 types of tags:
1) The label can be configured, the system provides a configuration entry of the label, and the system calculates the portrait label of the client through corresponding fields of the configuration label, related logic formulas and update frequency. For example:
A. defining a label named as 'high-frequency high-volume transaction';
B. the user builds a label, the label is associated with an index of 'average transaction amount of the last month', and the numerical value of the 'transaction amount of the last month' is judged, for example, the transaction amount is more than 10 ten thousand; meanwhile, the user who is associated with the average trading times in the last month, such as the trading times >10 times, can be identified as an index of the average trading amount in the last month;
C. each tag index is provided with a weight, and the setting range is an integer between 1 and 100;
D. the service manager can generate a label configuration after completing the configuration.
2) And customizing the label, wherein the collection system supports label customization for the configurable label which does not support index labeling. The system marks the corresponding label content on the portrait data through batch processing according to the available labels. The business personnel who obtain the access authority can access the label management function to carry out personalized customization on the labels.
The collection system manages the portrait tags of the enterprise through a tag management interface, and the management interface provides functions of adding, deleting, modifying and checking the tags. When the user performs the operations of deleting and changing, the system prompts the bullet phone frame to perform secondary reminding. When the user "looks" at the tag details, the system provides 2 pieces of information for the customer portrait: 1. the definition details of the tag comprise information such as configuration personnel of the tag; 2. the label acts as an image data volume.
The management interface provides a navigation bar, and the labels are positioned through the names, the identifications, the states, the time and the like of the search labels.
2. Enterprise and gold customer image presentation
The special fund intelligent collection system provides a main entry page of portrait information and supports the searching and displaying of the portrait of the enterprise and the client. The user selects and positions specific portrait information to enter a portrait information display page. When prompting list information, the information to be prompted comprises a client name, a whole row unified client information identification number, belonging branches and the like. Each user can only view the client portrait information of the organization.
After clicking on the corresponding customer information, the user may enter detailed data of the corresponding portrayal information through a "show details" button on the list. The user can enter the detailed information page of the customer portrait through the list information, and the user can see the detailed information of the customer portrait on the page. The detail page of the customer portrait is mainly composed of 2 parts: first, a summary information page of a customer portrait, the main information is shown in table 1; and secondly, a detailed information dimension navigation page of the customer portrait, wherein main information is shown in table 2.
The user can enter information detailed display of each dimension through navigation (information dimension), the displayed information dimension comprises each information dimension of the enterprise and gold portrait, and the information of each dimension has a detailed display page of the information so as to be convenient for the user to view. For unavailable information latitude (no corresponding information), the icon is displayed in gray and the active information icon is displayed in color.
TABLE 1
TABLE 2
Information dimension | Involving rights |
Customer information | Visible to the user |
Contract information | Visible to the user |
Guarantee information | Visible to the user |
Early warning information | Visible to the user |
Liability information | Visible to the user |
Property clue | Specified special resource client manager is visible |
Transaction adversary information | Specified special resource client manager is visible |
Bad treatment information | Visible to the user |
Judicial case information | Visible to the user |
External data information, e.g. auction information (temporary list) | Visible to the user |
3. Enterprise and gold customer monitoring
The special resource collection system can monitor the customer portrait in real time. The user can add the customer to the user's personal monitoring list through the [ concerns ] button (customer level) on the presentation page of the enterprise silhouette. The user can manage the monitored client objects in the personal monitoring list, and can add or remove the monitored objects or add the monitored objects under his own name (enterprise gold clients).
When the monitoring data information (default is property clues) of enterprise and gold clients in the monitoring list changes, the clearing and collecting system informs the user in the mode of OA to-do matters (special asset management) or short messages, and the user can check the configuration. Examples of notifications:
the number is: the property clues of the enterprise customers of XXXXX are changed, and the enterprise customers are requested to log in the system in time for checking.
And the intelligent decision module: and carrying out characteristic quantification on the treatment scheme through data modeling, and giving quantification treatment and reference recommendation of case similarity. The system supports multiple modes of policy matching, and a user can search for similar cases by selecting different policy matching values for the user to refer to and then make decisions. The intelligent decision is mainly realized based on two algorithms: expert matching algorithm based on label weight and expert matching algorithm based on machine learning model.
1. Expert matching algorithm (priority: high) based on tag weights
The expert matching algorithm based on the tag weight recommends proper expert opinion for the user by adjusting the tag weight so as to provide service for the user. The label-based treatment scheme takes the label of the image of the treatment scheme as a quantization index of the matching degree, the same label is counted, the reference recommendation of the treatment scheme is carried out through the weighting value of the label, and the corresponding label weight is calculated according to the weight set by the label to calculate. The algorithm is as follows:
1) The user inputs a certain enterprise gold client K, and a label set is obtained based on the portrait label of the client: { X1, X2, X3}, each tag carries a weight q1, q2, q3;
2) Acquiring a tag set of a disposal client object K1 corresponding to a case A1 in a case library: { X1, X3, X4, X5}, each tag carries a weight q1, q3, q4, q5;
3) Taking the same tag as 2: and X1 and X3, calculating corresponding weights and taking the weights as matching degrees: p=q1+q3;
4) The traversal is all the same as the condition, a matching degree value p is obtained through calculation, and the ranking is performed according to the p value from high to low, namely the recommended ranking;
5) The reason for the matching is illustrated in the show as the tag: x1 and X3 are matched.
2. Expert matching algorithm (priority: middle) based on machine learning model
The matching algorithm based on the machine learning model needs to obtain an optimal matching formula through data set training and then calculate the matching degree. The enterprise-gold case recommendation algorithm needs to be trained on the basis of the information dimension of the case-corresponding customer portraits. The information dimension table is shown in table 3.
TABLE 3 Table 3
And the system obtains a similarity value m of the target client and the treatment case according to the dimension information and combining a machine learning algorithm, and sorts the similarity value m to obtain a recommended case. The system gives a similar judgment description (combined with model output) at the time of presentation, such as presentation of similar contract properties, similar property clues and similar judicial conditions.
And (5) expanding property clue module: in order to further discover special asset and property clues, restore real values of assets and avoid clue risks, the special asset intelligent collection system combines 'personal intelligence' and 'machine intelligence' to develop market discovery and value discovery. The property cue expanding module comprises:
a clue mining and pushing tool module: the clue mining and pushing tool module has two modes: a legacy mode and a higher order mode. The traditional mode carries out preliminary filtering on external data by labeling processing external information. And then, aiming at different assets (such as cash in a row and real estate under a name), the property clues are pushed to corresponding householder.
The high-order mode is based on the traditional mode, and available property clues are further mined through tools requiring cross verification capability such as equity penetration analysis, suspected real control person mining and the like.
Standard business approval automation module: standard business approval is for traditional retail businesses and complex non-standard businesses.
A traditional retail business
After business data enter the system, firstly, carrying out characteristic analysis on the data, wherein the characteristic analysis mainly comprises a main body and debt; then automatically identifying the data according to the rules, and classifying and approving; and the system examines and approves the data according to the rule base, if the rule base is met, the examination and approval is completed, and if the rule base is not met, the manual examination and approval is carried out.
B complex nonstandard business
For complex nonstandard businesses, the system completes business approval through a multidimensional scoring mechanism, and approval contents comprise risk preference, market environment, inventory condition, supervision policy and the like, and are dynamically adjusted according to real-time treatment cases.
Intelligent recommended treatment operating scheme module: based on the internal historical transaction cases and the external transaction data, key indexes under different treatment or operation schemes are quantized by combining 'old teacher' experience, an analysis tool or model is built, intelligent scheme recommendation is realized, and business personnel are assisted to draw a more comprehensive and better clear collection scheme.
An estimation center module: in order to reduce the labor cost, the judgment of the special assets is more comprehensive and fair, the fully-adjusted price setting burden is effectively reduced, and the special asset collection system builds an evaluation center module to carry out intelligent evaluation on the special assets. The estimation center mainly comprises an estimation pricing model, an integrated estimation tool and an estimation report.
1. Valuation pricing model
The valuation model is formed by integrating internal historical transaction data and external collection transaction/litigation data based on valuation methodology and comprises an asset model, a creditor model and a small enterprise batch automatic recovery prediction model. And the estimation is more accurate through dimension model verification and discrete machine learning.
2. Integrated estimation tool
In order to meet the requirements of fine management such as the fit degree of the valuation tool and the flow, the application difference of parameters of different transaction scenes needs to be considered.
1) Data source
Aiming at the data of the internal business system and the price and cost data of the external market, the valuation center can normalize asset valuation measurement by a self-selection valuation method. The basis for valuation is mainly derived from an already online asset case library, such as villa, residence, office building, etc.
2) Application scenario
The estimation center is matched with different application systems according to different application scenes. Aiming at mortgage assets, using a deposit valuation model, and performing systematic evaluation on the user mortgage assets through a value index model, a big data drive and a look-ahead valuation model; aiming at the purchased assets, an acquisition valuation system is used for evaluating the asset acquisition value by integrating asset attenuation factors, discount rates and compensatory capacities; for the asset to be treated, a treatment valuation system is used to give reasonable treatment suggestions in combination with treatment policies, lifetime, budget treatment amount.
3) Mechanism matching
In order to make the valuation more accurate, the valuation center adopts an internal valuation and external valuation mode to comprehensively evaluate the assets to be evaluated. The internal valuation adopts a standardized valuation tool to carry out scientific evaluation on the assets to be evaluated, and the external evaluation carries out comprehensive evaluation on the assets to be evaluated by means of professional skills. And the inner and outer combination and the division work cooperate to improve the accuracy of the evaluation.
3. Valuation reporting
The project with successful evaluation can be summarized and counted and displayed in different dimensions through the evaluation center module; recording original data including estimated data and deleted data when the user uploads the data; and support the generation of test reports in an automated manner from the predicted outcome information and the associated data information. The main operation flow is as follows:
1) Checking an evaluation report, wherein a user can summarize and count and display items with successful evaluation in different dimensions through an evaluation center, and a system provides that a branch person can only check the evaluation report generated by uploading data information;
2) Clicking [ valuation result inquiry and downloading ] to select [ valuation report downloading ], wherein the page is in an input format;
3) Clicking the [ download ] button of the page, namely, deriving related estimation type information in a word form;
4) And (3) displaying all the items by default, clicking [ query ] after inputting the query condition, and displaying the item information meeting the condition by the page. Clicking [ reset ], clearing the query content, and displaying all items on the page;
5) If multiple valuation reports need to be downloaded, the valuation reports are downloaded in the form of compressed packets, and each valuation report only contains one item. The derived file name format is: estimate report + estimate type + estimate method + year month day (accurate to time-division seconds);
6) If the interface inquiry has a value, displaying an estimation result as [ download ]; if the interface query does not return a value, displaying an estimated result as [ in the estimation ]; if no value is returned by the interface three times in succession, the evaluation result is displayed as [ evaluation failure ].
And a system security module: different security measures are implemented for different security threats, including identity authentication, access control, security audit, communication security, and software fault tolerance.
1. Identity authentication
And uniformly verifying the users logging in the special resource collecting system through the industrial portal platform system. The user who fails authentication is not allowed to log in the system, and log information of login failure is recorded.
2. Access control
And verifying the user logged in the system through the portal platform system of the industry, and determining whether the special resource collecting system can be logged in by the system. User role information is maintained by the portal and rights are controlled.
3. Security audit
After the user logs in the system, the system can automatically record the operation trace of the user, including the operation user, the operation time, the IP address and the like. Meanwhile, the operation system and the database can update log information along with the operation of the user, so that the traceability of the operation is ensured.
4. Communication security
In order to ensure the communication safety of the user accessing the special resource collecting system, the system establishes an information safety channel by using HTTPS. Meanwhile, data verification is required in the data transmission and transfer process so as to ensure the integrity and safety of the data. The data files from the source data area to the system adopt a double-file storage mode, namely a data file and a flag file, wherein the flag file records verification information of the data file, including information such as file name, data date, file size, record number and the like. After the files arrive at the system, the special data collection system performs data verification on all the data files according to the verification information, and only files passing the verification are adopted by the system, otherwise errors are reported and waiting for processing.
The database is set differently for different access users. The data are stored according to the hierarchy, and the data which need to be displayed by the front-end application are arranged at the top layer. For the back-end ETL processing, an ETL user (with relatively large authority) is set, a front-end user is set for front-end application access, and the front-end user can only perform read-only access on result data displayed by the front end, so that the authority of different users on data access is strictly controlled, and the data security is ensured in all directions.
5. Software fault tolerance
The special resource collection system application adopts cluster deployment, and when a single application fails, the system can automatically identify and judge and directly forward the application to an available application server for operation. A main and standby mechanism is provided for the database, and when the main server fails, the database can be switched to the standby machine to run, so that the stability of the whole running of the clearing and collecting system is ensured.
Example 2
This embodiment is a preferred embodiment of embodiment 1.
Referring to fig. 1-3, the special asset intelligent collection system is composed of a client, an external service, an access service, a database, and the like. The project is based on a new generation JAVA development basic platform JAVA Unified Platform, JUP for short. The platform is a multi-layer architecture (VUE/AJAX+spring+Hibernatis/iBatis/JDBC) based on J2EE, takes Acegi as a security control architecture, builds a framework on the mature open source technical result, provides mechanisms such as thing management, log management, exception management, log management, cache management, internationalization support and the like, and supports general functions such as user login, user management, authority management, system monitoring and the like. Meanwhile, the platform is compatible with a common database such as Oracle, mySql, informix, DB and the like and a common application server such as JBoss, webSphere, webLogic and the like.
A basic deployment
The special asset collection system is independently deployed based on container cloud, a kylin operating system is adopted, a data real-time wind control system is called, the calculation requirements of various valuation models of special assets are met, and the special asset collection system is in butt joint with an external data integration platform and a big data base platform. The external data integration platform is responsible for real-time interface call of external data and GUT standardized conversion of external data files; the big data basic platform is responsible for unified storage of external data, and unified processing is carried out on the external data and the in-line data through a data integration layer of the big data basic platform. The special asset intelligent collection system supports three working modes of real-time, batch and tenant, and simultaneously services such as log, cache, full-link tracking, registration configuration center, distributed monitoring analysis and the like of a station in the access technology provide data support for collection decision assistance. Middleware uses JUP built-in integration, and a database is built by OpenGauss. Meanwhile, the valuation models of bonds, physical objects, equity and other assets of the same bank can be opened to other system clothes for use through the data wind control system.
2. System optimization
The intelligent collection system for special assets belongs to a general level system with a security level of 1, and has more batch processing steps and daily data processing capacity of more than 10G. In order to improve the operating efficiency of the system, the system will be deployment optimized from the following aspects.
1. Improving the efficiency of the system in batch processing and data processing
1) The processing servers are deployed in a cluster mode, so that the parallel processing capacity is improved, and the processing efficiency is improved;
2) Splitting and refining batch processing steps, and improving concurrency;
3) The dependence among batch processing steps is optimized, the batch processing time is shortened, and the data providing time is advanced.
2. Front-end data query optimization
1) Storing the data table according to time partition, and storing the historical data and the current data in the data partition;
2) Reasonably creating an index;
3) The complex inquiry is preprocessed and then the result is stored, so that the time is replaced by space, and the inquiry efficiency is improved.
3. Extensibility of
1) If the data processed by the system are related to the large data platform, the tenant can be applied to the large data base platform, and the data are processed by utilizing the calculation force of the large data base platform and then provided for the system;
2) The functional module follows the principle of high cohesion and low coupling, thereby improving the expandability of the software.
4. Data quality
An independent data quality inspection module is constructed, measures such as data verification are adopted to check the accuracy, the integrity, the consistency and the like of the data, the data quality problem is captured in time, and the data quality is improved.
3. Specific measures of RPO, RTO index
1. RPO specific countermeasures
1) Backing up data: incremental sealing is carried out every day, and full backup is carried out once a week;
2) Data backup location: carrying out data disaster recovery in a same city disaster recovery center;
3) Failure/disaster recovery protocol: the system is provided with a fault/disaster recovery plan after complete testing and exercise.
2. RTO specific countermeasures
1) The deployment mode is as follows: adopting deployment modes such as hot standby or clusters, and the like to prevent single-point faults of an application server or a database server;
2) Standby network system: the data center is provided with a standby communication route and redundant network equipment;
3) Standby infrastructure: setting up disaster recovery environments in the same city/different place machine room, and enabling the environments to be in a ready or running state for supporting recovery of disaster faults;
4) Operation maintenance management capability: the system has medium access, verification and restocking management system, and signs an emergency supply protocol meeting the disaster recovery time requirement with relevant manufacturers;
5) Disaster recovery plans: the system is provided with a fault/disaster recovery plan after complete testing and exercise.
The beneficial effects brought by the system are as follows:
1. response time
For administrative class users: under the maximum transaction concurrency of the production environment, the probability that the response time of the system management and parameter setting type request is less than 3s is more than 80%.
For query class users: under the production environment, the probability that the response time of the general query request is less than 3s is more than 80 percent; the probability that the response time of the complex type query request is less than 5s is more than 80%.
2. Maximum time allowed for system service interruption
The longest tolerance time for single fault system recovery is less than or equal to 24 hours, namely RTO fault recovery < = 24 hours;
3. maximum time allowed for system data loss
The longest tolerance time for lost data from system fault to recovery is less than or equal to 24 hours, namely RPO fault recovery < = 24 hours;
4. traffic handling capacity
Average value: 7000 pen/day peak: 8000 pens/day
5. Data throughput
Average value: 20G/day peak: 24G/day
6. Data storage amount
Mean 758G/day peak: 1533G/day
Those skilled in the art will appreciate that the invention provides a system and its individual devices, modules, units, etc. that can be implemented entirely by logic programming of method steps, in addition to being implemented as pure computer readable program code, in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units for realizing various functions included in the system can also be regarded as structures in the hardware component; means, modules, and units for implementing the various functions may also be considered as either software modules for implementing the methods or structures within hardware components.
The foregoing describes specific embodiments of the present invention. It is to be understood that the invention is not limited to the particular embodiments described above, and that various changes or modifications may be made by those skilled in the art within the scope of the appended claims without affecting the spirit of the invention. The embodiments of the present application and features in the embodiments may be combined with each other arbitrarily without conflict.
Claims (10)
1. The intelligent collection system for the special assets of the bank is characterized by comprising the following modules:
the customer portrait construction module: integrating data related to a specified special asset customer by integrating data in and out of a line, constructing a special asset enterprise customer portrait, and providing customer portrait presentation, portrait tagging and customer monitoring;
and the intelligent decision module: carrying out characteristic quantization on the treatment scheme through data modeling, and giving quantization treatment and reference recommendation of case similarity;
and (5) expanding property clue module: displaying and managing suspected property clue data of a user client, and exploring special property clues through artificial intelligence and machine intelligence;
an estimation center module: performing intelligent evaluation on the special assets, wherein the evaluation center module comprises an evaluation pricing model, an integrated evaluation tool and an evaluation report;
and a system security module: different security measures are implemented for different security threats, including identity authentication, access control, security audit, communication security, and software fault tolerance.
2. The bank specific asset intelligent collection system according to claim 1, wherein: the portrait tagging comprises configurable tags and custom tags; the configurable label carries out quantization judgment according to the label value, and is obtained through corresponding fields, logical formulas and update frequency calculation of the configurable label; the custom tag is customized by acquiring the access authority user.
3. The bank specific asset intelligent collection system according to claim 1, wherein: the intelligent decision module is realized through an expert matching algorithm based on a label weight or an expert matching algorithm based on a machine learning model;
the expert matching algorithm based on the tag weight is to recommend expert opinion for a user by adjusting the tag weight, count the same tag by taking the tag of the portrait of the scheme to be treated as a quantization index of matching degree, recommend the reference of the scheme to be treated by the weighting value of the tag, and calculate the corresponding tag weight according to the weight set by the tag;
the expert matching algorithm based on the machine learning model is a matching algorithm based on the machine learning model, the matching degree is calculated through a matching formula obtained through data set training, the similarity value of the target client and the treatment case is obtained, and the recommended case is obtained through sorting according to the similarity value.
4. The bank specific asset intelligent collection system according to claim 1, wherein: the extended property cue module comprises:
a clue mining and pushing tool module: the method comprises a traditional mode and a high-order mode, wherein the traditional mode is used for primarily filtering external data through labeling processing external information, and pushing property clues to corresponding households according to different assets; the high-order mode is based on the traditional mode, and the available property clues are further mined through stock penetration analysis and suspected real control people mining;
standard business approval automation module: after business data of traditional retail business enter the system, firstly, according to characteristic analysis of the data, including main body and debt; automatically identifying the data according to the rules, and classifying and approving; the system examines and approves the data according to the rule base, if the rule base is met, the examination and approval is completed, and if the rule base is not met, the manual examination and approval is carried out; aiming at complex nonstandard businesses, the system completes business approval through a multidimensional scoring mechanism and dynamically adjusts according to real-time treatment cases;
intelligent recommended treatment operating scheme module: based on the internal historical transaction cases and the external transaction data, key indexes under different treatment or operation schemes are quantified, and an analysis tool or model is built to generate a clearing scheme.
5. The bank specific asset intelligent collection system according to claim 1, wherein: the security audit is that after a user logs in the system, the system automatically records the operation trace of the user, including the operation user, the operation time and the IP address; the operation system and the database update log information along with the operation of the user, so that the traceability of the operation is ensured.
6. The bank specific asset intelligent collection system according to claim 1, wherein: the communication security includes: an HTTPS is used for establishing an information security channel, data verification is carried out in the data transmission process, a double-file storage mode is adopted for data files from a source data area to the system, namely, one data file and one mark file are adopted, verification information of the data files is recorded in the mark file, a special asset collection system carries out data verification on all the data files according to the verification information, files passing verification can be adopted, and errors are reported when verification fails, and the files wait for processing.
7. The bank specific asset intelligent collection system according to claim 1, wherein: the software fault tolerance includes: the special asset collection system application adopts cluster deployment, and when a single application fails, the system automatically identifies and judges and directly forwards the application to an available application server for operation; and providing a main and standby mechanism for the database, and switching the database to the standby machine for operation when the main server fails.
8. The intelligent clearing and collecting method for the special assets of the bank is characterized by comprising the following steps of:
the customer portrait construction step: integrating data related to a specified special asset customer to construct a special asset enterprise customer portrait by integrating data in and out of a row, and providing customer portrait presentation, portrait tagging and customer monitoring;
intelligent decision step: carrying out characteristic quantization on the treatment scheme through data modeling, and giving quantization treatment and reference recommendation of case similarity;
expanding property clues: displaying and managing suspected property clue data of a user client, and exploring special property clues through artificial intelligence and machine intelligence;
asset assessment: performing intelligent evaluation on the special assets, wherein the evaluation center step comprises an evaluation pricing model, an integrated evaluation tool and an evaluation report;
a security evaluation step: different security measures are implemented for different security threats, including identity authentication, access control, security audit, communication security, and software fault tolerance.
9. The bank special asset intelligent collection method according to claim 8, wherein: the portrait tagging comprises configurable tags and custom tags; the configurable label carries out quantization judgment according to the label value, and is obtained through corresponding fields, logical formulas and update frequency calculation of the configurable label; the custom tag is customized by acquiring the access authority user.
10. The bank special asset intelligent collection method according to claim 8, wherein: the intelligent decision step is realized through an expert matching algorithm based on a label weight or an expert matching algorithm based on a machine learning model;
the expert matching algorithm based on the tag weight is to recommend expert opinion for a user by adjusting the tag weight, count the same tag by taking the tag of the portrait of the scheme to be treated as a quantization index of matching degree, recommend the reference of the scheme to be treated by the weighting value of the tag, and calculate the corresponding tag weight according to the weight set by the tag;
the expert matching algorithm based on the machine learning model is a matching algorithm based on the machine learning model, the matching degree is calculated through a matching formula obtained through data set training, the similarity value of the target client and the treatment case is obtained, and the recommended case is obtained through sorting according to the similarity value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211345603.6A CN115994818A (en) | 2022-10-31 | 2022-10-31 | Intelligent collection system and method for special assets of bank |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211345603.6A CN115994818A (en) | 2022-10-31 | 2022-10-31 | Intelligent collection system and method for special assets of bank |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115994818A true CN115994818A (en) | 2023-04-21 |
Family
ID=85994437
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211345603.6A Pending CN115994818A (en) | 2022-10-31 | 2022-10-31 | Intelligent collection system and method for special assets of bank |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115994818A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116341879A (en) * | 2023-05-26 | 2023-06-27 | 杭州度言软件有限公司 | Overdue asset collection intelligent case division method and system |
CN116800792A (en) * | 2023-06-25 | 2023-09-22 | 福建润楼数字科技有限公司 | Method for building intelligent routing platform system with multiple partners |
-
2022
- 2022-10-31 CN CN202211345603.6A patent/CN115994818A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116341879A (en) * | 2023-05-26 | 2023-06-27 | 杭州度言软件有限公司 | Overdue asset collection intelligent case division method and system |
CN116341879B (en) * | 2023-05-26 | 2024-05-31 | 杭州度言软件有限公司 | Overdue asset collection intelligent case division method and system |
CN116800792A (en) * | 2023-06-25 | 2023-09-22 | 福建润楼数字科技有限公司 | Method for building intelligent routing platform system with multiple partners |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20220343432A1 (en) | Machine learning architecture for risk modelling and analytics | |
CN110796470B (en) | Data analysis system for market subject supervision and service | |
Messier Jr et al. | Inducing rules for expert system development: an example using default and bankruptcy data | |
CN102930025B (en) | For the treatment of attribute and the interval system and method for large data | |
US11532040B2 (en) | International cash management software using machine learning | |
CN115994818A (en) | Intelligent collection system and method for special assets of bank | |
US8359284B2 (en) | Organization-segment-based risk analysis model | |
US20060085235A1 (en) | Inventory mitigation and balancing system for dynamically and iteratively tracking, matching, and exchanging inventory excess and storage | |
Rishehchi Fayyaz et al. | A data-driven and network-aware approach for credit risk prediction in supply chain finance | |
US9721294B1 (en) | Apparatus and method for evaluating and presenting supply chain condition of an enterprise | |
WO2011142988A1 (en) | Risk element consolidation | |
WO2011142985A1 (en) | Technology infrastructure-change risk model | |
US8984022B1 (en) | Automating growth and evaluation of segmentation trees | |
CN111160745A (en) | User account data processing method and device | |
CN112183916B (en) | Land reserve life cycle management system | |
Mossalam et al. | Using artificial neural networks (ANN) in projects monitoring dashboards’ formulation | |
US11995622B2 (en) | Method of international cash management using machine learning | |
CN113609193A (en) | Method and device for training prediction model for predicting customer transaction behavior | |
CN110766462A (en) | Intelligent panoramic customer portrait linkage method and system based on streaming platform | |
CN115619571A (en) | Financing planning method, system and device | |
CN111415067A (en) | Enterprise and personal credit rating system | |
Wang et al. | Multiview Graph Learning for Small‐and Medium‐Sized Enterprises’ Credit Risk Assessment in Supply Chain Finance | |
Polder et al. | Complementarities between Information Technologies and Innovation Modes in the Adoption and Outcome Stage: A MicroEconometric Analysis for the Netherlands. CAED conference | |
CN107844874A (en) | Enterprise operation problem analysis system and its method | |
KR101909138B1 (en) | Receivable recovery support system for medium-small enterprise account receivable bond decrease and bad debt prevention based on big data |
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 |