CN1347529A - Method of Visualizing Information in Data Warehouse Environment - Google Patents

Abstract

According to the invention, techniques for visualizing customer data (103) contained in databases (6), data marts and data warehouses (8). In an exemplary embodiment, the invention provides a method for graphically analyzing relationships in data (103) from one or more data sources of an enterprise. The method can be used with many popular visualization tools (21), such as a On Line Analytical Processing (OLAP) tools (2) and the like. The method is especially useful in conjunction with a meta-model (103) based technique for modeling the enterprise data. The enterprise is typically a business activity (21), but can also be other loci of human activity (10). Embodiments according to the invention can display data from a variety of sources in order to provide visual representations of data in a data warehousing environment (8).

Description

Method for visualizing information in data warehousing environment

The application requires the right of priority of following U.S. Provisional Patent Application, and disclosing of this application comprises all appendix and all attached documents, includes on the whole in this, supplies the various uses reference:

The title that Li-Wen Chen was accepted on January 15th, 1999 is 60/116, No. 016 temporary patent application of the U.S. (agency's examination 19608-000200US) of " is the method and apparatus of the integrated and application integration processes user data of OLAP according to contrary star pattern ".

The following the application's of comprising the common pending application of owning together is accepted simultaneously, and other application is also included in this for the various uses reference.

The title of Li-Wen Chen be " method for visualizing information in data warehousing environment " the U.S. _ _ _ number patented claim (agency examination 19608-000210US);

The title of Li-Wen Chen for U.S. of " making the equipment of information visualization under the data warehouse environment " _ _ _ number patented claim (agency's examination 19608-000220US); And

The title of Li-Wen Chen for U.S. of " making the system of information visualization under the data warehouse environment " _ _ _ number patented claim (agency's examination 19608-000230US).

Background of invention

Relate generally to computer database system of the present invention relates in particular to the visual method that is stored in the information with dynamic format in the data warehouse environment.

Only go back the few several years ago and can predict computer technology rapid development.Now computing machine is in our family, in the office, in the school, even in our briefcase and in the knapsack.Since computer automation in daily life more and more earth effect we, government, enterprises and individuals have sought help from database technology and have helped them and tackle " information explosion ", promptly must be classified on continuous basis, the exponential form of digestion and information of managing increases.An importance in the database design field is that the data model that is used for database application is selected.

The structure or the tissue of the data of storing in data model representation database.It makes and can use data under some form and may use under other forms by restricting data.Different application needs different data models usually.Can there be many different data models, and their marks differently usually each other.Typically, database application is to the customization of the particular data model of certain database.Usually, these application must realize again to each database, and are identical even the function of using keeps.

Recently, the database development personnel have sought help from data warehouse technology and have solved the data management requirement that often clashes.Traditional data warehouse method is absorbed in decision support application, and the latter emphasizes to have the summary information of static format.Although there is tangible advantage, an inherent defect of these systems is that lost customer identifies.In addition, on classic method, by can improve the availability on the customer data analytical technology to client's display analysis result and data-base content.

Needed be the customer data that comprises in a kind of analytical database, data center (data mart) and the data warehouse and under dynamic format the method for exhibition information.

Summary of the invention

According to the present invention, be provided for the technology of the customer data that contains in visible database, data center and the data warehouse.In an example embodiment, the invention provides a kind of method, figure ground is analyzed from the relation in the data of one or more data sources of an enterprise.This method can with many universal visualization tools for example on-line analysis handle (OLAP) instrument etc. and use together.This method combines with technology based on meta-model, and is particularly useful to the data modeling.A kind of typically commercial activity of enterprise, but also can be other mankind's activity rail track.Can show data according to various embodiments of the present invention, thereby provide visual representation the data in the data warehouse environment from various data sources.

Provide a kind of method according to a first aspect of the present invention, be used for the relation of figure ground analysis from the data of one or more data sources of an enterprise.This method comprises various key elements, for example receives the definition of at least one the client's profile in a plurality of client's profiles (profile) group.User's input that reception indicates the form selection of configuration also is the part of this method.Based on this form selection of configuration and this information creating at least one the first dimension table and the fact table of at least one also be the part of this method.The first dimension table and the form that constitutes of at least one fact table that provides by at least one also is provided this method.Each embodiment can utilize any technology in the following multiple technologies that form is provided, that is: quantitative analysis tech, for example number percent, accumulation, the close classification of classification; Statistical technique, for example Pareto method, histogram, bell formula curve and linear regression; Data mining technology, for example affine rule, neural network, decision tree and fragment how much, financial analysis technology, for example net present value (NPV) and time series technology, for example mobile average.

Some embodiment can also create customer list and form the client segmentation assembly of building in the meta-model for each client's profile for each the client's profile in described a plurality of client's profile groups.This meta-model can be Call center's model, for example reverse star schema.

In one embodiment, the invention provides the method that a kind of foundation is used for client's profile form of Business Performance mensuration etc.In this embodiment, the first dimension table of setting up at least one comprises and sets up client's profile level; And the described fact table of setting up at least one also comprises according to this client's profile level total Business Performance to be measured.

In another embodiment, the invention provides the method that a kind of foundation is used for the operation form of Business Performance mensuration etc.In this embodiment, the fact table of setting up at least one comprises and adds up to Business Performance to measure and filter client's profile.

In an embodiment again, the invention provides the method that a kind of foundation is used for client's behavior form of user logging etc.In this embodiment, the described first dimension table of setting up at least one comprises and sets up client's profile; And the described fact table of setting up at least one comprises according to these client's profiles collects user logging.

Provide a kind of handle to be sent to method according to the tissue of the second optional data model according to the information of the first data model tissue according to a second aspect of the present invention, wherein this second data model serves as the basis of analyzing these data.This method comprises various key elements, for example receive as input to the definition of first data model and receive definition to second data model as input.Set up a mapping so that the also part of this method of the data conversion from first data model to second data model to be provided.Therefore, can according to this mapping data from first database migration to second database.

Provide the method for a kind of analysis according to third aspect present invention from the information of database.Can organize this database according to first data model.This method can comprise the definition to second data model that receives as input.The mapping of foundation from first data model to second data model also can be the part of this method.Then this method can be according to second data model and these map analysis data.

Each embodiment can be data organization in various data models, to support any model of multiple analytical model, for example: the scoring model comprises profit table, recently and the analysis of frequency amount of money and freedom and mobility analysis; Parted pattern is as demographic geographic distribution, behavior, purchase intention, termination tendency; And distributed model, as client's life time value, sequence analysis and affinity analyzing.

Can obtain surpassing many benefits of routine techniques by the present invention.The present invention can provide various for satisfy different commercial require customization make the visual technology of data relationship.In addition, use some embodiment to be used to solve the customer data problem analysis according to technology of the present invention and data model.Many embodiment can be the user ability of their data presentation of customization to use together with customer data analytic function general and that can re-use are provided.Each embodiment can provide client's dynamics and commercial dynamic (dynamical) dynamic observing.Many embodiment can easier, the faster and commercial application of foundation than hitherto known method with having bigger extendability.The benefit of these and other is described by this instructions.Can further understand essence of the present invention and advantage by remainder and each accompanying drawing of consulting this instructions.

Description of drawings

Illustrate according to data analysis of the present invention and visual representative architecture with 1A;

Figure 1B illustrates and is applicable to the representative computer system that realizes according to a specific embodiment of the present invention;

Fig. 2 A-2D illustrates according to the data model example in each specific embodiment of the present invention;

Fig. 3 A-3D illustrates according to the representative data model in each specific embodiment of the present invention;

Fig. 4 A-4D illustrates according to the representative visualization technique in each specific embodiment of the present invention;

Fig. 5 A-5B illustrates the treatment step process flow diagram according to the simplification of one embodiment of the invention; And

Fig. 6 A-6D illustrates according to each the representative data model in the specific embodiment of the present invention.

Embodiment

Can provide the various technology that are used for carrying out customer data analysis and visual analyzing result according to various embodiments of the present invention according to the data of company information disposal system.

Recently, enterprise can have one or more decision support system (DSS) (DDS), and DDS is devoted to provide information to support various enterprise operations to the decision maker.This information is typically with whole entity, for example public entity of enterprise, government organs etc., operation performance and efficient be the center.Yet it is on the example at center that traditional commercial decision system is based upon usually with the operation, and this has limited their serviceabilities in application customer-centric.Recently, analyze the ability of client's information,, become decision maker's important information source promptly about the information of client's profile, client's behavior and client's commercial activity.

Legacy system can comprise a plurality of infosystems discrete or networking, and each system supports a concrete commercial operation and business process.For example, so-called " distributed key task systems " supports daily commercial operation.They can comprise operation system and transaction system in company or the government " rear office ", and for example application, human resources application or the like are used, made to Enterprise Resources Plan.The system that disposes in so-called " the place ahead office " uses is used by the personnel such as representative of sales ﹠ marketing, Customer Service Representative or marketing personal, being customer service daily.For example, sale force automation system (SFA), promote automated system, the desktop of seeking help etc. is representational the place ahead office system.The place ahead office application and rear office application all provide support to the daily operation of enterprise.

Decision support system (DSS) provides analysis to summary information etc. according to the user at the needs of analyzing and understand operation performance and efficient.Decision support system (DSS) can comprise data warehouse, data center or the like.Decision support system (DSS) is mainly according to multidimensional model, and some client's intelligence analysis function can be set.Decision support system (DSS) can provide runing " macroscopic observation " of performance to the decision maker.Typically, decision support system (DSS) is made of the data model with the business scope that defines during system design.

Figure 1A illustrates the system architecture according to the decision support system (DSS) customer-centric in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not cause restriction.The insider can recognize other modification, modification and substitute.Fig. 1 illustrates a plurality of assemblies, and for example the OLAP control desk 1.1 pair of analysis of OLAP control desk provides interface with the user of the data that the OLAP form presents.1 butt joint of OLAP control desk is subjected to the OLAP instrument of user instruction input that " front end " is provided.An olap server 2 is imported the data base querying 4 of generation to database 6 from user's instruction.The user can utilize input command inputs such as graphic user interface (GUI) or command link interface.For retrieving information can be applied to database 6 to the order input, utilize GUI to be illustrated on the OLAP controller 1 or and represent that parts can provide this information to the user with other.OLAP instrument (not shown) for example utilizes GUI to can be used for manipulation data according to multidimensional model 5.With discussed in detail, olap server utilizes user's multidimensional data model mapping to convert user's instruction input to data base querying as below.

OLAP metadata 3 can comprise 5 to for example mappings of database 6 of database model of multidimensional data model from the user.For example OLAP metadata 3 can provide the table and attribute respectively the tieing up and true conversion to database 6 ofmultidimensional model 5.

The inquiry 4 that OLAP generates for example can be generated to retrieve one group of result from database 6 by olap server 2.Multidimensional model 5 can be made of conceptual model, and they for example can provide observes " macroscopic view " of Business Performance.These models can help business analysis to understand commercial operation performance etc.

The process of (3,5,6,12) such as the 21 management definition of decision support manager, establishment and generation data model, form, metadata, tables.Keeper's control desk 20 can make user's plan of fighting to thefinish support manager 21 definition of data models, form etc.This control desk for example can be to user's video data model in GUI.

Database 6 is set up by decision support manager 21.The user can be used for the pattern (schema) of database 6 from 20 definition of keeper's control desk.In a specific embodiment, keeper's control desk 20 can with OLAP control desk 1 in same position.The user can utilize the data definition multidimensional model of keeper's control desk 20 for them.At user definition behind their model,decision support manager 21 can automatically be set up database according to the user's data model definition, for example database 6.Decision support manager 21 for example can generate and the star schema data model of multi-dimension data cube (cube) 5 correspondences.Supervisory routine 20 can generative process and mapping ruler (mapping rules), so that use the data breeding database from data warehouse 8.

Metadata repository 22 is from storing the metadata of being used by decision support manager 21.This metadata can comprise the definition to data model, form, table, managing process, such as the process of the operation metadata of progress msg, or the like.

When the user asked form,decision support management 21 can generate middle table, and for example qualified target customer's table can reuse in other inquiry after them.Can generate middle table by the process that the usefulness that is positioned at same place withdecision support manager 21 generates temporary table.But the middle table in middle table high-speed cache 7 diode-capacitor storages.Middle table high-speed cache 7 can be according to any algorithm in a plurality of cache replacement algorithms, LRV for example, the intermediate result in the maintenance memory.In a present preferred embodiment,decision support manager 21 can comprise a management process that is used to manage middle table high-speed cache 7.

Data warehouse 8 is made of a data storage bank that is used for customer data, commercial operation data etc.Data warehouse 8 for example can comprise a database, and can be used as an operation data storage.One at present and in the preferred embodiment, data warehouse 8 can have a sign centre data tissue, and wherein the center identifies the composition customer information.Sometimes this data organization is referred to as " reverse star schema " data model.This data organization comprises data warehouse 8 " back end layers " (" back-end tier "), and its front end layer is based on the data model that is called " star schema ".

Data source 9 is made up of the raw data source.In a representative embodiment, data source 9 is operating systems of an enterprise, for example is used to manage daily commercial operation etc.Data from data source 9 can be transformed and move before use.For example, be transformed into reverse star schema data model to data storage to it in data warehouse 8 time so long.

Can generateanalytical statement 10 from the content of data warehouse 8.The user can be from keeper's control desk 20request forms.Manager 21 can be for the inquiry of retrieve data proposition to database 6.Manager 21 can send to the user to data as form.

Application 11 is waited by various commercial application to be formed, and the data of storage are worked together in this application and the data warehouse 8.Use 11 and can comprise sale group automation application, market automation application, E-business applications etc.The user can utilize keeper's control desk 20 to select interested will being input to use target customer's section of 11.Manager 21 can automatically generate the table that is used for each application integration.

Target customer's table 12 contains relevant target customer's information.Can generate these tables by catalogue listing or artificially as required.For example can be mutually integrated with decision support system functions customer-centric according to making commercial application from the table 12 that data warehouse 8 generates

For example can asmultidimensional data model 5, generate the OLAP metadata according to defineddata model.Manager 21 can utilize the metadata breeding 0LAP metadata 3 that derives from the user's data model definition.Like this, can automatically generate OLAP metadata 3.

Figure 1B illustrates and is applicable to realization one representative computer system according to specific embodiments of the invention.This figure only is exemplary and the scope of claims of this paper is not imposed restriction.The insider can recognize other modification, modification and substitute.Figure 1B illustrates each basic subsystem that is suitable for the computer system used together with the present invention.In Figure 1B,computer system 113 comprisesbus 115, the subsystem that this bus interconnection is main, for examplecentral processing unit 114,system storage 116, I/O control 118, such as theremovable dish machine 136 that holdsremovable dish 138 in the external unit (not shown) of printer,indicator screen 124,serial port 128, keyboard 132,fixed disk drive 144 and the operation through display adapter 126.Can connect many miscellaneous equipments, for example the Geniusmouse 146 or thenetwork interface 148 that are connected withserial port 128 of the scanner (not shown) by I/O controller 118.Can similarly connect many miscellaneous equipments or subsystem (not shown) under the mode.In addition, as discuss the back, realize that need not there be all devices shown in Figure 1B in the present invention.Can be to be different from the mode shown in Figure 1B interconnect each equipment and subsystem.For example the operation of the computer system shown in Figure 1B is that the insider is known, thereby does not go through in this application.Implement to be deployed in thesystem storage 116 in the source code of the present invention operation or be stored on the storage medium such asshaft collar 144 orremovable dish 138.

In according to present preferred enforcement of the present invention, a kind of method that is used to provide the customer data analysis ability of prior art the unknown can be set in thesystem storage 116 of thesystem 113 of Figure 1B or shaft collar in the operation.The customer data analysis can include, but not limited to decision-assisting analysis that business decision and client's behavior are associated.The customer data analytical applications can be analyzed data and client activities, incident, transaction and state and customer ID are associated according to customer ID.Also can use technology such as decision support application, summary technology or the like, this does not depart from the scope of the present invention.

Data organization in the data model descriptive data base.Select data model to go up the use of convenience data in some respects but use of possible restricting data on aspect other.Different application typically needs different data model very inequality each other usually.Thereby database application normally is that the data model that this database adopted customizes.These application can be embodied as the database that is used to have the different pieces of information model, although the basic logic of using may be similar.Meta-model is a kind of abstract data model, between the different entities in this model description data model or the relation between the different entities group.Application according to a specific meta-model design and exploitation can be re-used in other analogue.Can be easily by customizable details application code be set fourth as metadata, customized application.Thereby by following the relation that illustrates in the meta-model, application can customize the digital-to-analogue type.

Application developer can be considered trading off between the competition factor when setting up meta-model.Some competition factors are: 1) the customization data model is to adapt to the ability of different commercial application requirements; 2) dirigibility of the design application code that can re-use; And 3) availability that when using, should use with certain specific data models.

Can be written as database application consistent and come with reference to detailed data model by means as the data dictionary with meta-model.Application code is become to re-use according to the present invention's these and other technology.An example of data model/meta-model combination is star schema/multidimensional model combination of using in the data warehouse applications.In this combination, star schema comprises one " generic data model " and multidimensional model comprises a meta-model.

Data warehouse can make the commercial operation generalities with the multidimensional model such as meta-model.Such meta-model can be simplified analysis and understand the performance of commercial operation or the task of efficient.Multidimensional model for example, can provide " macroscopic view " of Business Performance to observe.These observations can be high level overviews, so that to corporate executive officer commercial clear " big picture " is shown.Can be by every data of index such as a plurality of business methods, aspects, thus to action officer data are shown from different visual angles.

Fig. 2 A is illustrated in the expression according to the data model in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not impose restriction.The insider can recognize other modification, modification and substitute.Fig. 2 A illustrates adata model 201 with star schema structure.The insider can recognize easily that other does not deviate from the structure of the scope of the presentapplication.Data model 201 comprises a fact table 202 and a plurality of dimension table 204,206,208 and 210.These multidimensional tables can be mapped to index of each dimension in the multidimensional model of this database, and the record in the fact table can be mapped to tolerance (measure) or data point in the multidimensional model of this database.

Fig. 2 B illustrates cube and expresses 220, and its expression is according to the dimension table 204,206,208 in Fig. 2 Adata model 201 of a specific embodiment of the present invention, 210 and the tissue of fact table 202.This figure only is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.The cube that can define when setting up this database among Fig. 2 B is expressed each dimension of 220.The dynamically data cube expression of the foundation of the data from data warehouse such ascube 220 of alternate embodiment of customer data analysis but, is provided according to client's dynamic behaviour.These embodiment can make information can be user capture by demand.

Fig. 2 C illustrates the conceptual model such as meta-model is mapped to according to the data model in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 2 C illustrates the relationship map that defines on the meta-model level to a plurality of relational data models by a plurality of cube representatives.In one embodiment, these relation data moulds can have " star schema " tissue, and this tissue is applicable to the runnability of the decision support mechanisms of for example describing enterprise-level.The star schema data organization can further be discussed in the back.The insider can set up many modification of star schema and multidimensional model, supports various application, instrument, system and framework under the scope that does not deviate from the present application.In an alternate embodiment, can utilize the multidimensional model that is not mapped to data model to realize the multi-dimensional database system.

Fig. 2 C illustrates from generating the multidimensional data model according to the meta-model the specific embodiment of the present invention.This figure only is an example, and the claims to this paper do not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 2 C illustrates meta-model 230.Meta-model 230 has tissue customer-centric.But, can use other tissue and not deviate from the scope of the present application.

Fig. 2 C illustrates meta-model 230 and is mapped to a plurality of data models of representing with cube 270,260,250 and 240.Cube 240 and 250 representatives have client's dimension and dynamic customer section level based on operation properties data tissue.What have client's dimension can illustrate the Business Performance tolerance relevant with client characteristics based on operation properties data tissue.Fig. 2 C also illustratescube 260, and it is to have the representative example based on operation properties data tissue that client's section filters.Based on operation properties data tissue some Business Performance tolerance can be shown, for example by the sales volume of client's section, press product line and marketing channel is measured the income of the customers of certain product of purchase in special time period.Fig. 2 C also illustrates cube 270, and it is the representative example based on the data organization of client's behavior with a plurality of client's sections dimension.The commercial activity that client's behavior tissue can utilize different classification or different section that the client is shown distributes.

Fig. 2 D is illustrated in according to the simplification to the meta-model 30 of Fig. 2 C in the specific embodiment of the present invention and expresses.This figure only is illustrative, and the scope to this paper claims does not apply restriction.The insider can recognize other modification, modification and substitute.

The data model of Fig. 2 A-2D and the insider said that already very clearly various modifications and other data model can provide the basis for a plurality of application in the foundation various embodiments of the present invention utilizes commercial accounts' data warehouse illustrates in greater detail with reference to Fig. 3 A-3D as the back.

Fig. 3 A-3D illustrates the entity relationship diagram of the simplification that is used for representational example information relation or " pattern ", and these information relationships or " pattern " can be supported according to the client's intelligence analysis in the various embodiments of the present invention.Fig. 3 A illustrates according to a data model 340 in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Data model 340 is representative example by the operation performance data tissue shown in thecube 240 of Fig. 2 C.Fig. 3 A illustrates data model 340 and comprises that being used to analyze the client that the client sells ties up 342.Data model 340 also comprises client's profile entity 344, so that use the sale of determining client's section according to profile.Data model 340 can form client's profile according to a plurality of predefined static natures or predefined behavioral characteristics in the profile entity 344.In this example, client's profile entity 344 comprises predefined profile attributes: the average profit 345 of client's purchase frequency grade 341, average bill grade 343 and every month.Data model 340 just can be used for an example in the several data model that customer data analyzes.The insider can add other attribute and use other structure, and does not deviate from the scope of claims of invention.

Fig. 3 B illustrates according to of the present invention one concrete harmful second data model of executing in the example 350.This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 3 B illustrates the fact of the data model 340 of collecting Fig. 3 A, so that directly obtain the sales volume of client's section.Data model 350 is representative example by the operation performance data tissue with client's peacekeeping dynamic customer section level shown in thecube 250 of Fig. 2 C.Data model 350 comprises client's profile entity 352 bags as a dimension that is used for the fact 354.But can be for will data model 350 being set with client's section gold, silver, copper definition or that define with any other receiving mode.Can collect (aggregate) along the level of one or more dimensions.For example, can collect sales volume to other client of every kind of level.

Fig. 3 C illustrates the data model 360 according to a specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Data model 360 shown in Fig. 3 C is by the representative example based on operation properties data tissue that client's section filters that has shown in thecube 250 of Fig. 2 C.

Fig. 3 D illustrates according to the data model 370 in the specific embodiment of the present invention.This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 3 D illustrates data model 370, and it can utilize the profile determination method to be used for analyzing client's commercial activity.Data model 370 is the represented representative example based on the data organization of client's behavior with client's section dimension of the cube 270 of Fig. 2 C.The commercial activity that client's behavior tissue can utilize different classification or different sections that the client is shown distributes.Application can utilize these data models to obtain can have such as the client what have some feature the information of certain performance profile.For example, performance profile can comprise the client's profile based on the performance metric of collecting.This information is to be provided by the profile of data model 370 dimension 374, and this information is based on client characteristics in the data warehouse.

Other data model that the data model of Fig. 3 A-3D and insider understand can serve as the basis of the multiple analysis in the foundation various embodiments of the present invention, as following illustrate in greater detail with reference to Fig. 4 A-4D.

Fig. 4 A illustrates according to the data model 340 of Fig. 3 A in the specific embodiment of the present invention representative analytical technology of (also being represented by thecube 240 of Fig. 2 C).This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 4 A illustrates the cube of have the client (customer name) dimension, sales promotion peacekeeping time (moon) dimension and expresses 440 data models of representing 340.Table 401 provides each dimension, a Business Performance tolerance (being the average consumption volume in this example) and formula that calculates this Business Performance tolerance of cube 440.The business measurement that form 405 illustrates the example relevant with client characteristics is the average consumption volume.

Fig. 4 B illustrates according to the data model 350 of Fig. 3 B in the of the present invention one concrete enforcement the representative analytical technology of (also being represented by thecube 250 of Fig. 2 C).This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 4 B illustrates the data models 350 that the cube of have the client (consumption section) dimension, sales promotion peacekeeping time (stage) dimension is expressed 450 representatives.Table 411 provides each dimension, a Business Performance tolerance (being the average consumption volume in this example) and formula that calculates this Business Performance tolerance of cube 450.The business measurement that form 415 illustrates with gold, silver, example that copper client section is relevant is the average consumption volume.

Fig. 4 C illustrates according to the data model 360 of Fig. 3 C in the specific embodiment of the present invention representative analytical technology of (also being represented by thecube 260 of Fig. 2 C).This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 4 C illustrates the data models 360 that the cube of have channel (channel group) dimension, product (product group) peacekeeping time (stage) dimension is expressed 460 representatives.Table 421 provides cube to express each dimension, a Business Performance tolerance (being to sell to distribute in this example), a formula and a filtration dimension (being the client in this example) of calculating this Business Performance tolerance of 460.Form 425 illustrates for the business measurement of different client's sections example relevant with each product group and promptly sells distribution.

Fig. 4 D illustrates according to the data model 370 of Fig. 3 D in the specific embodiment of the present invention representative analytical technology of (also being represented by the cube 270 of Fig. 2 C).This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Fig. 4 D illustrates the data models 370 that the cube of (consumption section) the peacekeeping sales promotion that has client dimension is expressed 470 representatives.Table 431 provides each dimension, a Business Performance tolerance (being client's counting in this example) and formula that calculates this Business Performance tolerance of cube 470.It is that the client counts that form 435 illustrates in this example with gold, silver, example business measurement that each client's section of copper is relevant.

Fig. 5 A illustrates the reduced graph according to the representative client intelligence analysis in the specific embodiment of the present invention.This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Some embodiment do not comprise all by the analysis shown in Fig. 5 A.Other embodiment may comprise the analysis of other type simultaneously.The information of analyzing is from the customer data library searching, and this customer database can be such as the data warehouse of the data warehouse 8 of Figure 1A or the part of data center.In many examples, directly or indirectly the customer ID with concrete is relevant for this information.Can merge and construct this information in following step, handling subsequently.In a representative embodiment, merge and tectonic information can comprise and being transformed into from the source database information extraction, it and the form of basic meta-model (for example can be reverse star schema) compatibility and then data being encased in this meta-model.

In step 504, according to one or more features to client segmentation (or segmentation).Feature can comprise static nature and behavioral characteristics.That static nature can comprise directly and business management is associated but from understanding and predicting that the commercial viewing angle of client's commercial activity is useful client properties.For example, client's demography, geography information etc. are static natures.Behavioral characteristics can comprise the commercial activity that the client shows in the business management process.For example, client activities, incident, transaction and the information of therefrom deriving are behavioral characteristics.

Definition client profile in step 506.Can form client's profile from the top useful client segmentation of user behavior that step 504, forms for understanding under the commercial operation environment.Can define a plurality of client's profiles, make each profile can comprise that one group of business decision person is relatively and can discovery be useful client characteristics in related.For example, can be that the client definition in existing 3 to 5 years of 40,000 to 70,000 U.S. dollars and the account that had is a profile for age group, the income level that constituted by 20 years old to 30 years old.

In step 508, carry out client's behavior trend analysis.Can collect such as the Business Performance data of sales volume, income, client's quantity etc. and and one or more client's profile (for example client's profile that forms in the step 506) be associated.In a preferred embodiment, the Business Performance data analytic trend of client's profile can be used for.Can show the result so that the understanding to client's commercial activity and trend thereof to be provided to the user, for example the canned data 435 of Fig. 4 D.

In step 510, can be according to the understanding prediction client behavior that in step 508, forms to client's behavior trend.Can be according to the client properties in client's attribute or signatures to predict future.For example, in market index analysis (Market Basket Anlysis), can use one group of affine rule to determine purchasing model according to user profiles.

Fig. 5 B is illustrated in the foundation representative embodiment of the present invention to carrying out the reduced graph that the process steps of multidimensional form is set up in client's intelligence analysis.This figure is illustrative, and the scope to claims of this paper does not apply restriction.The insider can be appreciated that other modification, modification and substitutes.In step 550, receive the definition of at least one the client's profile in a plurality of client's profile groups.Then, in step 551, judge whether to need the profile that continues.If the profile that need to continue is carried out optional step 552, so that set up a customer list for each client's profile of client's profile group.Then, in another optional step 553, each customer list of setting up in the step 552 is cached in the middle table high-speed cache 12.Then, in optional step 554, set up the client segmentation assembly that stands in the meta-model for each client's profile.Otherwise perhaps under anything part, execution in step 556 is so that receive the input of indication form selection of configuration from the user.In a present embodiment, the user can select from the various report form type that comprise client's profile level, client's behavior and client's section and filtration.The insider can easily create various embodiment so that comprise the subclass of these report form type and the report form type that comprises the scope of claims that other does not deviate from this paper.

Can set up client's profile form for information according to embodiments of the invention, wherein this information for example is the Business Performance tolerance in data warehouse 8, data center or other database that is stored in data warehouse such as Figure 1A.Fig. 5 B is depicted as thestep 558 that client's profile form is set up the first dimension table of at least one.In step 560, for each dimension table of setting up in thestep 558 is set up client's profile level and attribute.Then, instep 562, set up the fact table of at least one.Can be according to client's profile level, fact table is created as has client's dimension, perhaps the set of Business Performance tolerance, the perhaps out of Memory in the data warehouse.At last, instep 564, provide a form of forming by one or more dimension tables and one or more fact table.

Can set up client's behavior form for information according to embodiments of the invention, wherein this information for example is to be stored in user logging in data warehouse 8, data center or the database of data warehouse such as Figure 1A etc.Fig. 5 B is shown in the step 568 that client's profile of using in client's behavior form is set up the first dimension table of at least one.These dimension tables can comprise surrogate (surrogate key) etc.Then instep 572, set up the fact table of at least one for client's behavioural information by collecting user logging according to client's profile.At last, instep 574, provide a form of forming by one or more dimension tables and one or more fact table.

Can utilize client's section and filter or the like to set up for information and run form according to embodiments of the invention, wherein this information for example is the various Business Performance tolerance of the data warehouse, data center or other database that are stored in data warehouse such as Figure 1A etc.Fig. 5 B is depicted as the step 578 that the operation form is set up the first dimension table of at least one.Then instep 582, measure and filter, set up at least one fact table by collecting Business Performance according to client's profile.At last, instep 584, provide a form of forming by these one or more dimension tables and one or more fact table.

Fig. 6 A illustrates the simplification entity relationship diagram according to the representative meta-model with star schema tissue of a specific embodiment of the present invention.This figure is illustrative, and the scope to claims of this paper does not apply restriction, and the insider can recognize other modification, modification and substitute.1203 representatives of business measurement group are used for the Business Performance of compiling of a commercial operation and measure.Business measurement group 1203 can comprise one or more " true components ", for example true component 1200.True component is represented the concrete measurement in each commercial subject areas that the corporate decision maker wishes to analyze.For example, common true component can comprise that many Business Performances measure, for example sales volume, always lay in sales volume etc.Fig. 6 A describes a plurality of commercial operation groups 1204, and they are represented index or describe the business process or the operation of each tolerance in the business measurement group 1203.Embodiment can have any amount of commercial operation group, for example each commercial operation group 1204 of Fig. 6 A.Commercial operation group 1204 comprises that one or more dimension components 1201, one or more dimension search component 1202 and other component.Dimension component 1201 is represented the commercial operation of the data in the true component of concrete sign.For example, can be to be used to sell each true dimension such as projects such as product, marketing channels.Dimension is searched the details that component 1202 is described relevant dimension component 1201.For example, can be to search component such as projects such as product category, product types with the dimension of product dimension component correspondence.

For the macroscopic view prospect that commercial operation is provided, it is useful adopting the embodiment of star schema data model.Provide the embodiment of this macroscopic view prospect that decision support system (DSS) is provided " big picture " information relevant with decision-making as guide.Thereby, can provide static solution based on each predefined dimension and summary data based on the embodiment of star schema.The data aggregate that provides by the embodiment with star schema can provide high layer analysis prospect because of the character of multidimensional model.

Fig. 6 B illustrates being used in the specific embodiment and carries out the representative meta-model of the reverse star schema analyzed according to customer data of the present invention.This figure is illustrative, and the scope to the book of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.Client's incident or the activity association from the different commercial operations in one or more zones of business activity got up in many customer data analyses.In one embodiment, customer data is analyzed different client activities components, the client activities component 1215 of Fig. 6 B for example, and interior different activity associations get up.These analyses can associate client activities by a plurality of core components such as core component 1212.Can be used to the data in the core component 1212 discern entity from the client activities of different client activities components.In certain embodiments, in the enterprising line data analysis of level of the more details of client activities.These embodiment search component for the one or more activities of more detailed client activities data search, and for example component 1214 is searched in activity.Data in the client segmentation component provide different classification client's method or provide client's different commerce is observed.For example, can pass through geographic area, demography etc. to client segmentation.The mode that comes in handy of utilizing the embodiment of one or more types in these client segmentation component types can provide multiple observation customer data branch to harden fruit.The embodiment of employing reverse star schema provides the observation to the data level of detail, and this observation provides the ability of analyzing according to notion (for example customer data, client activities and their correlativitys on transaction or incident level).

Fig. 6 C illustrates according to the common version after the simplification of the representative data model in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims of this paper does not apply restriction.The insider can recognize other modification, modification and substitute.The data model of Fig. 6 C is a data model with reverse star schema tissue.Fig. 6 C illustrates a representative data model that comprisesfocal group 1210, and thisfocal group 1210 comprises that at least one component at least one component in a plurality ofcore components 1212, at least one component in a plurality ofclient segmentation component 1213, at least one group in a plurality ofcustomization group 1211, a plurality ofclient activities component 1215 and a plurality of activity search at least one component in the component 1214.Also can comprise other component, and needn't exist the institute that illustrates important in certain embodiments, this does not deviate from scope of the present invention.

Focal group, for examplefocal group 1210, comprise each component of descriptor, and this information for example is client characteristics, profile and commercial relevant classification, client role, the definition etc. in the different commercial operations field.Fig. 6 C illustrates two types component in the focal group 1210.Can comprise other component, this does not deviate from this scope that has.Fig. 6 C illustratescore component 1212 and client segmentation component 1213.Core component 1212 comprises client entity (CC1) and other identification data relevant with the client that indicates with (CC2-CCn).Information such as name such as account identification symbol, social insurance number, encryption are the examples of these customer ID data.When carrying out the analysis of client's event correlation, these entities are useful especially.Client segmentation component 1213 describe the role of relevant client in establishment or process in or the information of position.These illustrative components can with client's pattern of trade or organize relevant, for example, perhaps relevant with client characteristics such as the information of zone, channel, marketing team etc., for example business profile, demography, current profile etc.

Each component ofcustomization group 1211 and various forms of commercial operation transaction are corresponding.Because event transactions is disperseed in time, these components constitute the group of business measurement and attribute.These incidents can be independent each other, but also can be to rely on each other.Can utilize certain concrete sequence of events to describe the different phase of client activities.For example, in a concrete time phase, a client may go through such sequence of events: order＞make out the bill＞pay＞promote＞Pricing Program change＞service call＞cancellation.Each incident can relate to the different business processes or the operation of a plurality of reflection customer life cycles.

Customization group 1211 comprises that a plurality ofclient activities components 1215, a plurality ofactivity search component 1214 etc.Client activities component 1215 can be represented the event transactions or the tolerance of relevant client activities.These entities can comprise one or more attributes, for example type of transaction, exchange hour marking or the like.When definitionclient activities component 1215, select to be used for a thresholding of type of transaction.Type of transaction is an attribute useful to the incident correlation analysis.In one embodiment, the user can be by selecting attribute from a plurality of pre-established attributes, each client activities component of definition such as client activities component 1215.Some embodiment can also provide the ability of adding attribute defined by the user.Many embodiment provide the ability of definition client activities entities (for example CAC3 among Fig. 6 C).The entity thatcomponent 1214 is represented the feature of each refinement client event transactions is searched in activity.For example, can be stored as each activity that is used to analyze to the product of buying in the transaction, store locations of purchase or the like andsearch entity 1214.

Fig. 6 D illustrates the simplification entity relationship diagram according to a representative example of the data model with reverse star schema tissue in the specific embodiment of the present invention.This figure only is illustrative, and the scope to claims does not apply restriction.The insider can recognize other modification, modification and substitute.In the sample data model of Fig. 6 D, the core component group of Fig. 6 C comprises a client entity 1220 and an account entity 1222.Many-one relationship between arrow 1221 expression client entities 1220 and the account entity.The entity of existence such as client entity 1220 andaccount entity 1222 can makedata model 1213 that the account levels notion is provided, in order to inquire about the relevant client in the business prototype of being considered.Theclient segmentation component 1213 that Fig. 6 D illustrates Fig. 6 C comprises a plurality of entities, and wherein four are shown: marketing channel entity 1228, client region entity 1230, client'sprofile entity 1224 and demographic entity 1226.Many embodiment can comprise other entity, perhaps comprise the part of these entities rather than all, and this does not deviate from scope of the present invention.

In a concrete representative embodiment, population system body entity 1226, client'sprofile entity 1224 and client region entity 1230 have relation with client entity 1220.For example, as shown in Fig. 6 D, the many-to-one relationship between arrow 1223 expression client entities 1220 and the demographic entity 1226.Similarly, the many-to-one relationship betweenarrow 1225 expression client entities 1220 and the client'sprofile 1224; Many-to-one relationship betweenarrow 1227 expression client entities 1220 and the regional entity 1230.In addition, in this specific embodiment, shown in the arrow among Fig. 6D 1229, marketing channel entity 1228 andaccount entity 1222 have the relation of one-to-many.

In this concrete representative embodiment shown in Fig. 6 D, a plurality of different entities are formed theclient activities component 1215 of Fig. 4 A.These entities comprise; Bill business entities 1232, purchase/return transaction entity 1234, Service events entity 1236,marketing activity entity 1240,sales promotion entity 1242 and user-defined incident 1238.In addition, Fig. 6 D illustrates an activity that is made of product entity 1244 and searches component.Can also comprise not shown or unaccounted other entity in the foundation some embodiments of the present invention here, in addition, some embodiment may not comprise the entity that all are described herein, and this does not deviate from scope of the present invention.

Although according to various particular systems the present invention has been described generally above, it is much extensive that application of the present invention is wanted.Especially, the present invention is not subject to the data pattern of its particular type, but need can be applicable to the analysis of improved or optimization so that and data warehouse customer-centric and any data model of using under the occasion of using together.Thereby in certain embodiments, technology of the present invention provides all types of visits of different traditional commerce, government and educational database in a large number.Certainly, the insider can recognize other modification, modification and substitute.

Claims (37)

1. the information from the data of at least one is set up the method for multidimensional form, and described method comprises:

Reception is to the definition of at least one client's profile of a plurality of client's profile groups;

Receive the input of indication form selection of configuration from the user;

Set up at least one first dimension table according to described form selection of configuration and described information;

Set up at least one fact table according to described form selection of configuration and described information; And

Provide by described at least one the first dimension table and the report that constitutes of described at least one fact table.

2. the process of claim 1 wherein that described form is client's profile form, described information also comprises Business Performance tolerance, and the described first dimension table of setting up at least one also comprises:

Create client's profile level; And described at least one fact table of setting up comprises:

According to the described Business Performance tolerance of described client's profile level polymerization.

3. the process of claim 1 wherein that described form is the operation form, described information also comprises Business Performance tolerance, and the described fact table of setting up at least one also comprises:

The described Business Performance tolerance of polymerization; And

Filter described client's profile.

4. the process of claim 1 wherein that described form is client's behavior form, described information also comprises user logging, and the described first dimension table of setting up at least one also comprises:

Set up client's profile; And the described fact table of setting up at least one also comprises:

According to described client's profile polymerization user logging.

5. the method for claim 1 also comprises:

For each the client's profile in described a plurality of client's profile groups is set up customer list; And

Be that each client's profile is set up upright client segmentation component in meta-model.

6. the process of claim 1 wherein that described information comprises telecom information.

7. the process of claim 1 wherein that described information comprises Financial Information.

8. the process of claim 1 wherein that described information comprises retail market information.

9. the process of claim 1 wherein that described information comprises insurance information.

10. the process of claim 1 wherein that described information comprises healthcare information.

11. the information from the database of at least one is set up the method for multidimensional form, described method comprises:

Reception is to the definition of at least one the client's profile in a plurality of client's profile groups:

For each client's profile of described a plurality of client's profile groups is set up a customer list;

Be that each client's profile is set up upright client segmentation component group in a meta-model;

Receive the input of indication form selection of configuration from the user;

From described at least one database information extraction;

Set up at least one first dimension table according to described form selection of configuration;

Set up the fact table of at least one according to described form selection of configuration and described information; And

Provide by described at least one the first dimension table and the form that constitutes of described at least one fact table.

12. the method for claim 11, wherein said form are client's profile forms, described information also comprises Business Performance tolerance, and the described first dimension table of setting up at least one also comprises:

Set up client's profile level; And the described fact table of setting up at least one also comprises:

According to the described Business Performance tolerance of described client's profile level polymerization,

13. the method for claim 11, wherein said form are the operation forms, described information also comprises Business Performance tolerance, and the described fact table of setting up at least one also comprises:

The described Business Performance tolerance of polymerization, and

Filter described client's profile.

14. the method for claim 11, wherein said form are client's behavior forms, described information also comprises user logging, and the described first dimension table of setting up at least one also comprises:

Set up client's profile; And the described fact table of setting up at least one also comprises:

According to described client's profile polymerization user logging.

15. the method for claim 11, wherein said information comprises telecom information.

16. the method for claim 11, wherein said information comprises Financial Information.

17. the method for claim 11, wherein said information comprises retail market information.

18. the method for claim 11, wherein said information comprises insurance information.

19. the method for claim 11, wherein said information comprises healthcare information.

20. one kind is sent to the method for second database to the information from first database, described first database root is according to the first data model tissue, and described second database root is according to the second data model tissue as the basis of data analysis, and described method comprises:

Receive first data model definition as input;

Receive second data model definition as input;

Set up a mapping, described mapping provides the data conversion from described first data model to described second data model;

According to described mapping, the described data migtation from described first database is arrived described second database.

21. the method for claim 20, wherein said first data model comprises star schema.

22. the method for claim 20, wherein said second data model comprises to be designated the data organization at center.

23. the method for claim 22, wherein said sign is a customer ID.

24. the method for claim 20, wherein said information comprises telecom information.

25. the method for claim 20, wherein said information comprises Financial Information.

26. the method for claim 20, wherein said information comprises retail market information.

27. the method for claim 20, wherein said information comprises insurance information.

28. the method for claim 20, wherein said information comprises healthcare information.

29. the method for claim 20, wherein said second data model comprises reverse star schema.

30. an analysis is from the method for the information of database, described database root is according to the first data model tissue, and described method comprises:

Receive definition to second data model as input;

The mapping of foundation from described first data model to described second data model;

According to described second data model and the described data of described map analysis.

31. the method for claim 30, wherein said first data model comprises to be designated the data organization at center.

32. the method for claim 31, wherein said sign is a customer ID.

33. the method for claim 30, wherein said information comprises telecom information.

34. the method for claim 30, wherein said information comprises Financial Information.

35. the method for claim 30, wherein said information comprises retail market information.

36. the method for claim 30, wherein said information comprises insurance information.

37. the method for claim 30, wherein said information comprises healthcare information.

Images