CN107239486A - A kind of data characteristics storehouse method for building up and system - Google Patents
A kind of data characteristics storehouse method for building up and system Download PDFInfo
- Publication number
- CN107239486A CN107239486A CN201710256767.4A CN201710256767A CN107239486A CN 107239486 A CN107239486 A CN 107239486A CN 201710256767 A CN201710256767 A CN 201710256767A CN 107239486 A CN107239486 A CN 107239486A
- Authority
- CN
- China
- Prior art keywords
- data
- model
- client
- index
- public client
- 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
Classifications
-
- 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
- G06F16/21—Design, administration or maintenance of 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
- G06Q10/10—Office automation; Time management
-
- 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
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Abstract
The invention discloses a kind of data characteristics storehouse method for building up and system, wherein, methods described includes:According to the characteristic index obtained to common information and empirical analysis of the same trade or business to public client, business implication and processing rule further according to the characteristic index carry out theme to the characteristic index and dimension are divided, it is determined that the use demand of each index;According to the use demand of each index and processing rule, the feature database Data Mart to public client is built;Based on the feature database Data Mart, public client is analyzed and model training described, and mold curing and operation are carried out according to the data conversion rule and parameter configuration after the model training, and according to the data conversion rule.So as to realize the aimed management scheme based on client, proficient service is lifted, and more perfect data are provided and is supported.
Description
Technical field
The present invention relates to data processing field, more particularly, it is related to a kind of data characteristics storehouse method for building up and system.
Background technology
With continuing to develop that bank data is managed, bank customer relationship management (CRM) system (to public CRM system) is implementing
On the basis of industry business development strategy, client's strategy, with information technology, integrated customer relationship management, business activities management, pipe
The big module of decision support three is managed, client's identification and evaluation, business activities Whole Course Management and examination, management level decision assistant is met
Support to need, realize the business management model for core " with client driven ".But current CRM system is deposited
In Railway Project:First, lack management and decision tool, lack to business activities process control, cause manager much to manage
Means can not all be applied;Secondly, the Reference News's deficiency provided for customer manager, customer analysis screening capacity is inadequate, to client
Business opportunity digging on industry chain (supply chain) not enough, all constrains the development of my industry business;Finally, " quantitative is evaluated to customer manager's marketing
It is few ", lack the data accumulation of marketing process, the problems such as " with feeling " is evaluated service team.
The content of the invention
In view of the drawbacks described above of prior art, embodiment of the present invention provides a kind of data characteristics storehouse method for building up and is
System, can effectively solve the problem that the multinomial problem of management of existing customer relationship management system.
Specifically, embodiment of the present invention provides a kind of data characteristics storehouse method for building up, and it includes:
According to the characteristic index obtained to common information and empirical analysis of the same trade or business to public client, refer to further according to the feature
The business implication and processing rule of item are marked, theme is carried out to the characteristic index and dimension is divided, it is determined that each index makes
Use demand;
According to the use demand of each index and processing rule, the feature database data to public client are built
Fairground;
Based on the feature database Data Mart, public client is analyzed and model training described, and according to the mould
Data conversion rule and parameter configuration after type training, and mold curing and operation are carried out according to the data conversion rule.
Correspondingly, embodiment of the present invention additionally provides a kind of data characteristics storehouse and sets up system, wherein, the system bag
Include:
Characteristic index analysis module, for obtaining referring to the feature of public client according to common information and empirical analysis of the same trade or business
Item is marked, business implication and processing rule further according to the characteristic index carry out theme and dimension to the characteristic index
Divide, it is determined that the use demand of each index;
Data Mart builds module, for the use demand according to each index and processing rule, builds institute
State the feature database Data Mart to public client;
Analyzing and training module, for based on the feature database Data Mart, being analyzed public client and model described
Training;
Mold curing module, for according to the data conversion rule and parameter configuration after the model training, and according to institute
State data conversion rule and carry out mold curing and operation.
There are following beneficial effects by using embodiment of the present invention:The aimed management scheme based on client of realization,
Proficient service is lifted, and more perfect data are provided and is supported.
Brief description of the drawings
Fig. 1 is a kind of schematic flow sheet of data characteristics storehouse method for building up according to embodiment of the present invention;
Fig. 2 is according to table structural representation in embodiment of the present invention;
Fig. 3 is the graph of a relation according to width table in embodiment of the present invention;
Fig. 4 is the Organization Chart that system is set up according to a kind of data characteristics storehouse of embodiment of the present invention.
Embodiment
For the ease of understanding the various aspects, feature and advantage of technical solution of the present invention, below in conjunction with the accompanying drawings to this hair
It is bright to be specifically described.It should be appreciated that following various embodiments are served only for for example, not for the limitation present invention's
Protection domain.
The title or term that may relate to according to the present invention are explained first.
ETL:Extract-Transform-Load is used for describing from data processing of the source terminal through extracting, changing, loading
Process;
SAS:Statistics Analysis System mono- are used for the software of processing and statistical analysis.
Embodiment 1:
Fig. 1 is a kind of schematic flow sheet of data characteristics storehouse method for building up according to embodiment of the present invention.Reference picture 1,
Specific embodiment is as follows:
Methods described includes:
Step S1, according to the characteristic index obtained to common information and empirical analysis of the same trade or business to public client, further according to institute
The business implication and processing rule of characteristic index are stated, theme is carried out to the characteristic index and dimension is divided, it is determined that each
The use demand of index;
Step S2, according to the use demand of each index and processing rule, builds the spy to public client
Levy storehouse Data Mart;And
Step S3, based on the feature database Data Mart, is analyzed public client and model training described, and according to
Data conversion rule and parameter configuration after the model training, and mold curing and throwing are carried out according to the data conversion rule
Production.
The present invention sets up feature database based on global data warehouse, and analyst can be changed by the subdivision to client with group point
It is apt to me and manages it corporate client's weak foundation, promotes differential, proficient service of the client structure optimization lifting based on customer segmentation;It is logical
Cross excavation marketer hangar increase business opportunity sources and lift business opportunity and obtain efficiency, perfect single visitor is provided for customer manager
Family analyze data;Lift the process management of business opportunity sale;Data analysis basis is provided for management level decision support.Therefore, this hair
The bright feature database Data Mart by building to public client, so that the multinomial management for solving existing customer relationship management system is asked
Topic, realizes the aimed management scheme based on client, lifts proficient service, and provides more perfect data and supports.
Embodiment 2:
In the step S1, theme and dimension division are carried out to the characteristic index to be included:To the characteristic index
Carry out the multi-level division of data major class, class of service and item of information.Specifically, with classical ten large-sized models of Data Warehouse In Bank
Subject area (group, region, agreement, event, product, assets, marketing activity, channel, finance, application, model) is accurate as data
Standby basis;To promote the customer account management that becomes more meticulous based on subdivision, promote business opportunity and client's major event Whole Course Management as taking
The business demand driving of Data Mart is built, the data item that analysis feature database fairground needs includes:
1st, client's essential information:Accounts information, client's signing information, enterprise's incidence relation, financial information, Transaction Information,
Product information;
2nd, client characteristics information:Channel transaction feature, upstream and downstream feature, foreign trade feature, profit contribution feature, visitor
Family populational subdivision feature, cash flow are special.
(1) the demand data business major class for meeting customer character analyzing is analyzed with this:
(2) define after demand data business major class, according to demand business rule of the big alanysis from the access of system in row
Then, index classification is further refined:
3rd, last demand analysis teacher
Various specifying informations described in demand parameter classification after refinement are further decomposed, derived, repeatedly
Generation.
For example, the index classification of company's storage client's essential information to be subdivided into the tissue Date of Incorporation, the registration money of client
Gold, registered address, tissue property etc.;The real-time deal flowing water of account loans and deposits is counted with different time sections such as the moon, season, years
Accumulating sum and carry out com-parison and analysis etc., with this determination feature database index item.
Now feature database index item has been the complete description of the various basic datas needed for business demand, and each index
Can description most accurate, without half-heartedness explain a certain specific transactions implication, provided for the follow-up processed and applied to data
It is stable to support.Index item analysis result is exemplified below:
Embodiment 3:
In another embodiment of the invention, methods described in addition to above-mentioned processing mode, wherein, institute in step S2
Stating the structure feature database Data Mart to public client includes:Scene and practical application, the spy are excavated according to follow-up data
Data Mart is levied using width table mode.
Specifically, the storage of client feature library Data Mart refines processing client characteristics data, mainly face by depth
Comprehensive client characteristics data needed for providing from analysis mining to analysis model.Include according to client characteristics classification:Storage company visitor
Family feature database, potential corporate client's feature database, standard analysis dimension data.Using feature database index item, client characteristics classification as base
Plinth, consider the actual landing of index item safeguard, inquiry use, the scene such as analysis mining, build can be physico concept mould
Type.Its feature database index item inventory, processing frequency, conversation strategy, each that should include of each entity table of Definitions of Conceptual Model
The information such as index item data mart modeling path, inter-entity logical relation.
Data mart modeling personnel receive the characteristic index demand and conceptual model of demand analysis's teacher analysis design, according to demand
Analyze the process data path of each characteristic index and change physical model.According to the good entity table rule of Definitions of Conceptual Model,
Split conceptual model is actual or is integrated into and can create, workable physical model excavates for follow-up data and provides data
Source.Data mart model has Star Model, snowflake model etc..In the present invention, considered according to follow-up data excavation scene,
And considering to efficiency, storage, speed, easy-to-use degree etc. in actual use, Data mart model structure uses width
Table mode.
It is divided into Data Mart by data content:Knot is excavated in storage client feature library, potential customers' feature database, customer analysis
Fruit write-back;It is divided into current table, month to date table, year progressive schedule (table structure such as Fig. 2 institutes by index time scope of statistics and storage cycle
Show).
Wherein the relation of width table is as shown in figure 3, wherein:The data source of wide table in narrow table, be the collecting of narrow table data,
Summarize;Narrow table is the necessary supplement of wide table data.Wide table:Conveniently, use rapidly, easily, but processing logic is complicated.Narrow table:Data
Abundant information, easy processing, but data volume is big, processing speed is slow.The selection of wide table and narrow table is efficiency, storage, speed, easy-to-use journey
Degree etc. consider.
In figure 3, add up to the main table of public client characteristics, to public client characteristics month to date, to public client characteristics year as wide table;
It is commissioned and pays detailed, client and public affairs are participated in relationship, client and detailed, client contract of being contracted to private participation relationship, client
Detailed, the end of month contract is detailed, the end of month signing detail, cash flow month to date, cash flow month to date, channel are merchandised month to date, hold
Product earnings information, cash fleeting time add up, the fund fleeting time is accumulative, channel transaction adds up as narrow table in year.
Narrow table is detailed extension and the supplement of wide table corresponding data items.For example, in the main table of public client characteristics, existing
Some data item of " incidence relation ", and the storage of wide table can be only the public client association of this pair under current state number of the enterprise
(and relationship is participated in public affairs) and legal representative's information (and relationship is participated in private), if thinking the bright of this affiliated enterprise of analysis mining
Thin information and the historical variations situation of legal representative, need to be to more fine-grained narrow table drilling analysis.
Embodiment 4:
In another embodiment of the invention, methods described in addition to above-mentioned processing mode, wherein, it is described to described
Public client is analyzed and model training includes:Extemporaneous inquiry and thematic model are carried out to the public client using SAS instruments
Training.
Specifically, feature based storehouse Data Mart, analyst carries out the analysis mining to public client using SAS instruments,
Analysis mining application scenarios can be divided to two major classes:Extemporaneous inquiry, thematic model training.Extemporaneous inquiry is typically to be directed to row leader and right
Hot spot service problem, customers' strategy study task that public bar line management layer is proposed, quantify to public customer analysis Shi Caiyong
The mode of analysis, manages it True Data based on me, utilizes Data Mining Tools (such as SAS):1st, carry out extemporaneous analytic activity, and utilize
Excavate achievement and write specialist paper, to support management level to become more meticulous customers' policy development.2nd, the excavation achievement structure based on summary
The Marketing Model of interim form is built, and exports provisional special topic marketing list, to support to carry out precision marketing.Thematic model training
Include but are not limited to:1st, it is based on a line marketer interview, branch to public customer analysis teacher to feed back, by the battalion of certain time
Pin experience accumulation, periodically concludes typical to public client's marketing model and rule, and form thematic sales service specification.2nd, it is right
Public customer analysis teacher uses the mode of quantitative analysis, and based on real customer data, the product of thematic Customer is covered
The performance such as lid, profit contribution carries out trend analysis, assesses the effect of special topic marketing, model optimization is periodically carried out, constantly to be lifted
The precision of special topic marketing.
The output of extemporaneous inquiry, which is generally, writes analysis report, and thematic model training output is model rule.The following is point
The other illustration to two kinds of application scenarios:
(A) scene title:Extemporaneous analysis mining (customers' analysis of strategies)
Scene is defined:Hot spot service problem, client typically for row leader and to public bar line management layer proposition
Group's strategy study task, to the mode of public customer analysis Shi Caiyong quantitative analysis, manages it True Data based on me, utilizes data mining
Instrument (such as SAS EG), carries out extemporaneous analytic activity, and writes specialist paper using achievement is excavated, to support management level to become more meticulous
Customers' policy development.Operation flow:
Step1. management level assign hot spot service problem or customers' strategy study analysis task:Row is led and to public bar line
Management is proposed under hot spot service problem and strategy study task, line or passed through in customer account management and policy-making process
Reached under other routine work platforms to public customer analysis teacher.Output:To public customers' strategy study task description.
Step2. analyst carries out analytic activity:The analysis task that management level are assigned is directed to public customer analysis teacher, is based on
Newest True Data before my trade, using Data Mining Tools (such as SAS), analyze data performance, Rule Summary, and by model
The key chart of output is output to locally, to support follow-up report to write.Output:1st, the chart of model output, 2, it is extemporaneous
Analysis model.
Step3. analyst writes and submitted management level to report:To public customer analysis teacher according to analysis mining achievement and model
The chart material of output, writes analyst's report, by other routine work platforms using the OFFICE instruments such as PPT/WORD
Hand over to management level.Output:Analyst reports (form such as WORD/PPT)
(B) scene title:Special topic marketing modeling
Scene is defined:It is based on a line marketer interview, branch to public customer analysis teacher to feed back, by the battalion of certain time
Pin experience accumulation, periodically concludes typical to public client's marketing model and rule, and form thematic sales service specification;Herein
On the basis of, by the way of data quantify, corresponding Marketing Model is built, while model database management personnel (technology) will solidify and pass through
The model for crossing examination & verification is deployed to operation environment using certain way (SASCode or SQL) and applied.Operation flow is as follows:
Step1. analyst is based on a line interview and branch feeds back, and summarizes and write thematic sales service specification:Analysis
Teacher is based on a line interview result and branch and fed back, and periodically combs typically to public client's marketing model and rule, is locally writing
Thematic sales service specification (form such as EXCEL or PPT).Output:Thematic sales service specification.
Step2. analyst builds thematic Marketing Model and writes model operation instructions:Analyst is based on special topic marketing industry
It is engaged in specification, real data and Data Mining Tools (such as SAS EG) is manageed it using me, to client's screening rule of Marketing Model
Analyzed and summarized, and the thematic Marketing Model based on SAS is built in SAS training patterns storehouse, while according to unified mould
Type issues standard process, writes thematic Marketing Model operation instructions (and model development statement of requirements book), and by model and phase
Close and submit part to be committed to the model database management instrument under modeling environment in the lump.Output:1st, thematic Marketing Model, 2, special topic marketing mould
Type operation instructions, 3, thematic Marketing Model exploitation statement of requirements book.
Step3. model database management personnel carry out examination & verification confirmation to model and operation instructions, and in model database management instrument
It is middle to be issued, form the model version that examination & verification passes through.The thematic Marketing Model submitted for analyst, according to corresponding flow
Specification is audited, and examination & verification forms the model version that examination & verification passes through by that can be distributed to model database management instrument;Examination & verification is obstructed
Cross then by under line or other routine work platforms to analyst propose feedback opinion, and to the mould in model database management instrument
Type is modified or deleted.Output:Nothing.
Step4. model database management personnel (technology) carry out thematic Marketing Model deployment and gone into operation:
(1) thematic Marketing Model is deployed to SAS operation model libraries;
(2) if SAS operations model library uses SASCode forms, model database management personnel (technology) submit Step3
SASCode models are deployed to operation environment and carry out operation application;
(3) if SAS operations model library is using other forms such as SQL, model database management personnel (technology) are sent out based on Stp3
The thematic Marketing Model exploitation statement of requirements book of cloth, changes into the forms such as SQL, and is deployed to operation environment and applied.Output:
Corresponding SQL of model etc..
Step5. model operation operation monitoring personnel is monitored to model running situation:Model administrative staff after operation
(monitoring) is monitored and managed to the running of operation environmental model.Output:The operation report of operation environmental model.
Embodiment 5:
In another embodiment of the invention, methods described in addition to above-mentioned processing mode, wherein,
Based on the feature database Data Mart, public client is analyzed and model training described, in specific embodiment party
In formula, in addition, analyst's model after repetition training, will such as dispose the further testing model effect of the model in production environment,
Need to be through operation mold curing step.Analyst is enforceable processing rule the model conversion after training, and provides regular institute
The parameter configuration threshold values needed;Data mart modeling personnel are converted to the batch processing job that can be landed for the process requirements received, wherein
Parameter will use configurableization mode, support to change after going into operation.Solidification is a kind of analysis mining class mould for acting excavation and producing
The further rule optimization of detailed design key element that type and event class model are included, data mart modeling path are defined as repeatable realization
The process of rule model, is that model landing is deployed into operation environment subsequently through technical approach (SASCode or SQL) to be tied
Fruit output provides processing foundation with application.Wherein, analysis mining class model is included:Loyalty-value Segmentation Model, self-defined visitor
Family Segmentation Model, the distribution characteristics of customer groups analysis model, product coverage are with holding product Characteristics of Distribution model, capital chain
Bar upstream and downstream Characteristic Analysis Model, trading activity and channel preference Characteristic Analysis Model, cash flow Characteristic Analysis Model, customers
Product holds associated analysis model, trading activity and holds analyzing product association model, profit contribution Characteristic Analysis Model, visitor
The overall monitoring management model of family subdivision colony, customer segmentation population size variation analysis model, the product covering of customer segmentation colony
Spend variation analysis model, customer segmentation migration and churn analysis model and customers point colony's profit contribution variation analysis mould
Type.
Event class model is included:Marketing class event model, wherein, CE1:Settlement accounts fund fluctuation, CE2:I manages it
The expansion of foreign trade business upstream and downstream potential customers, CE3:Client handles opening up for the bank acceptance artificial potential customers of gathering
Exhibition, CE4:Client applies for be commissioned payment, CE5:I manage it core enterprise's upstream and downstream Customer mining, client alert class event model,
CE6:Fixed deposit is drawn in advance, CE7:Client hold finance product expire, CE8:Client's wholesale fund produce to him manage it is of the same name
Account, CE9:Active account number is persistently reduced, CE10:Account frequency of use is remarkably decreased.
Detailed design key element is included:1st, model data input, output characteristic field data source Pre-Evaluation;2nd, correlation analysis is set
Count (analysis path/rule);3rd, model output is designed with other models or function pages coupled modes;4th, model running mode
(artificial, Automatic dispatching etc.).
In addition, methods described also includes:Output result to operation model after the mold curing flows after data transfer
It is given to foreground application function.It is exemplified below:
Using one:Customer loyalty-value subdivision
Based on selected a certain class customers, analyzed according to loyalty-value combination dimension, form corresponding client
Segment colony.In recognizing and excavating high pay-off target customer group, the Strategic Proposals for target customers are formed, are supported
Decision-making level understands banking company's client's constitutive characteristic in depth, is supported to formulate and optimize customer value Strategies of Promoting.
Using two:Corporate client's cash flow signature analysis
The treasury trade analyzed between client and client is merchandised, and excavates client in bank capital stream feature and by clearing product
The clients fund stream feature of distribution, for Gong Getiao lines Marketing group is provided the objective group of target in the cash flow feature reference of bank and
Explanation.
Using three:Product coverage is with holding product distribution
Each subdivision colony exported based on storage client's hierarchical classification model and each Segmentation Model, further analyzes the colony
Client bank product distribution characteristics, including product coverage, hold product type, product quantity, product trading frequency and
Amount of money etc., is needed with the product for supporting total, corporate finance of branch Business Management Team to understand each high value subdivision colony client in depth
Rule is sought, it is accurate to formulate the product strategies segmented market for target customer.
5, the operation model having been carried out at present, operation model output and application relation are as shown in the table:
Fig. 4 is the Organization Chart that system is set up according to a kind of data characteristics storehouse of embodiment of the present invention, as illustrated, described
System includes:
Characteristic index analysis module 100, for according to the spy obtained to common information and empirical analysis of the same trade or business to public client
Levy index item, business implication and processing rule further according to the characteristic index, the characteristic index is carried out theme and
Dimension is divided, it is determined that the use demand of each index;
Data Mart builds module 200, for the use demand according to each index and processing rule, builds
The feature database Data Mart to public client;
Analyzing and training module 300, for based on the feature database Data Mart, being analyzed public client and mould described
Type training;
Mold curing module 400, for according to the data conversion rule and parameter configuration after the model training, and according to
The data conversion rule carries out mold curing and operation.
The present invention is by building the feature database Data Mart to public client, so as to solve existing customer relationship management system
Multinomial problem of management, realizes the aimed management scheme based on client, lifts proficient service, and provide more perfect data
Support.
In another embodiment of the present invention, the theme and dimension division of being carried out to the characteristic index includes:
The multi-level division of data major class, class of service and item of information is carried out to the characteristic index.
In another embodiment of the invention, the structure feature database Data Mart to public client includes:Root
Scene and practical application are excavated according to follow-up data, the characteristic fairground uses width table mode.
It is described public client to be analyzed and model training includes to described in a further embodiment of the present invention:Profit
Extemporaneous inquiry and thematic model training are carried out to the public client with SAS instruments.
In last embodiment of the present invention, the system also includes:Model output module, for the mould
The output result of operation model circulates after data transfer after type solidification gives foreground application function.
It should be noted that each embodiment of system is set up in above-mentioned data characteristics storehouse and the data characteristics storehouse is set up
The corresponding technology contents of method are completely the same, and in order to avoid repeating, this is not repeated here.
Through the above description of the embodiments, those skilled in the art can be understood that the present invention can be by
The mode of software combination hardware platform is realized.Understood based on such, technical scheme makes tribute to background technology
That offers can be embodied in the form of software product in whole or in part, and the computer software product can be stored in storage and be situated between
In matter, such as ROM/RAM, magnetic disc, CD, including some instructions are to cause a computer equipment (can be individual calculus
Machine, server, or network equipment etc.) perform method described in some parts of each of the invention embodiment or embodiment.
It will be appreciated by those skilled in the art that disclosed above is only embodiments of the present invention, certainly can not
The interest field of the present invention is limited with this, the equivalent variations made according to embodiment of the present invention still belong to the claims in the present invention
The scope covered.
Claims (10)
1. a kind of data characteristics storehouse method for building up, it is characterised in that methods described includes:
According to the characteristic index obtained to common information and empirical analysis of the same trade or business to public client, further according to the characteristic index
Business implication and processing rule, theme is carried out to the characteristic index and dimension is divided, it is determined that the use of each index is needed
Ask;
According to the use demand of each index and processing rule, the feature database data set to public client is built
City;
Based on the feature database Data Mart, public client is analyzed and model training described, and instructed according to the model
Data conversion rule and parameter configuration after white silk, and mold curing and operation are carried out according to the data conversion rule.
2. the method as described in claim 1, it is characterised in that described that theme and dimension division are carried out to the characteristic index
Including:
The multi-level division of data major class, class of service and item of information is carried out to the characteristic index.
3. the method as described in claim 1, it is characterised in that the structure feature database Data Mart bag to public client
Include:
Scene and practical application are excavated according to follow-up data, the characteristic fairground uses width table mode.
4. the method as described in claim 1, it is characterised in that described to be analyzed public client and model training bag described
Include:
Extemporaneous inquiry and thematic model training are carried out to the public client using SAS instruments.
5. the method as described in any one of Claims 1-4, it is characterised in that methods described also includes:
The output result of operation model after the mold curing is circulated after data transfer and gives foreground application function.
6. system is set up in a kind of data characteristics storehouse, it is characterised in that the system includes:
Characteristic index analysis module, for according to the characteristic index obtained to common information and empirical analysis of the same trade or business to public client
, business implication and processing rule further according to the characteristic index carry out theme to the characteristic index and dimension are drawn
Point, it is determined that the use demand of each index;
Data Mart builds module, and for the use demand according to each index and processing rule, it is described right to build
The feature database Data Mart of public client;
Analyzing and training module, for based on the feature database Data Mart, being analyzed public client and model training described;
Mold curing module, for according to the data conversion rule and parameter configuration after the model training, and according to the number
Mold curing and operation are carried out according to transformation rule.
7. system as claimed in claim 6, it is characterised in that described that theme and dimension division are carried out to the characteristic index
Including:
The multi-level division of data major class, class of service and item of information is carried out to the characteristic index.
8. system as claimed in claim 6, it is characterised in that the structure feature database Data Mart bag to public client
Include:
Scene and practical application are excavated according to follow-up data, the characteristic fairground uses width table mode.
9. system as claimed in claim 6, it is characterised in that described to be analyzed public client and model training bag described
Include:
Extemporaneous inquiry and thematic model training are carried out to the public client using SAS instruments.
10. the system as described in any one of claim 6 to 9, it is characterised in that the system also includes:
Model output module, circulates to foreground for the output result to operation model after the mold curing after data transfer
Application function.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710256767.4A CN107239486A (en) | 2017-04-19 | 2017-04-19 | A kind of data characteristics storehouse method for building up and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710256767.4A CN107239486A (en) | 2017-04-19 | 2017-04-19 | A kind of data characteristics storehouse method for building up and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107239486A true CN107239486A (en) | 2017-10-10 |
Family
ID=59983067
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710256767.4A Pending CN107239486A (en) | 2017-04-19 | 2017-04-19 | A kind of data characteristics storehouse method for building up and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107239486A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107784520A (en) * | 2017-10-16 | 2018-03-09 | 徐欣 | A kind of airline's marketing data integration system and method |
CN109360004A (en) * | 2018-09-25 | 2019-02-19 | 电子科技大学 | A kind of client relation management method and system |
CN110490729A (en) * | 2019-08-16 | 2019-11-22 | 南京汇银迅信息技术有限公司 | A kind of financial user classification method based on user's portrait model |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030208468A1 (en) * | 2002-04-15 | 2003-11-06 | Mcnab David Boyd | Method, system and apparatus for measuring and analyzing customer business volume |
CN101566981A (en) * | 2008-04-24 | 2009-10-28 | 长沙创智天马财务软件有限公司 | Method for establishing dynamic virtual data base in analyzing and processing system |
CN103747022A (en) * | 2011-12-31 | 2014-04-23 | 华茂云天科技(北京)有限公司 | Customer relation management system |
US20160300250A1 (en) * | 2015-04-07 | 2016-10-13 | Xerox Corporation | Methods and systems of forecasting customer demand in a print production environment |
CN106327325A (en) * | 2016-08-26 | 2017-01-11 | 北京元丁科技有限公司 | Bank big data operation management system and method |
-
2017
- 2017-04-19 CN CN201710256767.4A patent/CN107239486A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030208468A1 (en) * | 2002-04-15 | 2003-11-06 | Mcnab David Boyd | Method, system and apparatus for measuring and analyzing customer business volume |
CN101566981A (en) * | 2008-04-24 | 2009-10-28 | 长沙创智天马财务软件有限公司 | Method for establishing dynamic virtual data base in analyzing and processing system |
CN103747022A (en) * | 2011-12-31 | 2014-04-23 | 华茂云天科技(北京)有限公司 | Customer relation management system |
US20160300250A1 (en) * | 2015-04-07 | 2016-10-13 | Xerox Corporation | Methods and systems of forecasting customer demand in a print production environment |
CN106327325A (en) * | 2016-08-26 | 2017-01-11 | 北京元丁科技有限公司 | Bank big data operation management system and method |
Non-Patent Citations (2)
Title |
---|
上海市金融学会: "《深化改革中的货币政策、风险管理与金融发展》", 31 October 2014, 上海:学林出版社 * |
暨南大学教务处: "《暨南大学2006届本科生校级优秀毕业设计(论文)汇编(上)》", 31 December 2006 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107784520A (en) * | 2017-10-16 | 2018-03-09 | 徐欣 | A kind of airline's marketing data integration system and method |
CN109360004A (en) * | 2018-09-25 | 2019-02-19 | 电子科技大学 | A kind of client relation management method and system |
CN110490729A (en) * | 2019-08-16 | 2019-11-22 | 南京汇银迅信息技术有限公司 | A kind of financial user classification method based on user's portrait model |
CN110490729B (en) * | 2019-08-16 | 2022-11-18 | 南京汇银迅信息技术有限公司 | Financial user classification method based on user portrait model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Bertani et al. | The complexity of the intangible digital economy: an agent-based model | |
Abdeldayem et al. | Entrepreneurial finance and crowdfunding in the Middle East | |
Foroghi et al. | Applying decision tree to predict bankruptcy | |
CN107239486A (en) | A kind of data characteristics storehouse method for building up and system | |
Andreeva et al. | Competency-based management in a system of sustainable development of banks, financial and technology companies | |
Jeje | Intensive growth strategies and outreach performance of Tanzania-based savings and credit cooperative societies | |
Lestiyawati | Analytic Network Process and Tawhidi String Relation in Managing Business of Sharia Commercial Bank | |
Yangibaevich | Transformation of the regulation of commercial banks in the conditions of the development of the digital economy of Uzbekistan | |
Sajjadi et al. | A developing model for clustering and ranking bank customers | |
Muriuki | Effect of technology adoption on agency banking among commercial banks in Kenya | |
Mishra et al. | Business intelligence and analytics: Paving way for operational excellence in indian banks | |
Shawaqfeh | Contribution of Commercial Banks in Financing Small and Medium Enterprises. Applied Study on Commercial Banks in Jordan | |
Hasan et al. | The impact of big data on banking operations | |
Njuguna et al. | Quality management initiatives and performance of commercial banks in Makueni County, Kenya | |
Medennikov | The impact of digital transformation on the competitiveness of small and medium agro-industrial enterprises | |
Tuya | Improving costing methodology based on commercial bank restructuring | |
Wang et al. | A financial assets and liabilities management support system | |
KULCHITTIVEJ et al. | Credit management guidelines to strengthen Thai industrial sector | |
Esther et al. | Current trends in information communication technology (ICT) and the Accounting profession | |
Panja et al. | A Review Report on the Evolution and Implementation of Business Intelligence Technique in the Banking Sector | |
Yashin et al. | Contradictions of Corporate Top Management Motivation by Economic Value Added | |
Kalinin | The role and analysis of venture financing in the digitalization of the economy | |
SADIGOVA et al. | ECONOMICS AND ADVANCED CALCULATION METHODS OF FINANCIAL ENGINEERING | |
Mante et al. | Creating Competitive Advantage Through Service Innovation within the Banking Sector of Ghana: The Strategic Implementation of Corporate Social Responsibility | |
B. Gabbasova et al. | Digital technologies in the system of relations competitiveness of banks |
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 |
Application publication date: 20171010 |
|
RJ01 | Rejection of invention patent application after publication |