MANAGING CLASSIFICATION HIERARCHIES IN MASTER DATA MANAGEMENT ENVIRONMENTS
FIELD
[001] The present disclosure pertains to apparatus and methods for analyzing  modifying  and viewing hierarchies  records  and metadata for master data management (MDM) systems.
BACKGROUND
[002] Master data management (MDM) technology typically includes a set of processes and tools that define non-transactional entities of an organization in a persistent manner. These non-transactional entities can include customers  products  items  vendors  employees  users  or other suitable non-transactional entities. These non-transactional entities  also referred to as master data  are important assets for a company. The scope of Master Data Management includes creation as well as maintenance of such master data.
[003] There are a number of commercial MDM products in the market  which are supplied by vendors such as IBM (e.g.  IBM-WPC)  SAP (e.g.  SAP-MDM)  and Tibco (e.g.  Tibco-CIM). These existing commercial MDM products do not completely facilitate the master data management process. For example  users of these commercial MDM products desire degrees of customization and flexibility that are not completely satisfied by any of these products.
[004] Currently-available commercial MDM products come with their own bundled user interface (UI) and provisions to model MDM data according to a particular MDM system’s requirements. The products can provide a limited set of features that facilitate processes related to item creation  item-search  item-copy  search  data quality  data enrichment  and other suitable features. Further  the user interface bundled with MDM products is typically not easily customizable to include new screens and functionality. This restricts the business process and use cases of MDM users  forcing them to change or limit their business processes. Data models supported by MDM products also have limited flexibility when it comes to modeling various MDM entities and relations.
SUMMARY
[005] Apparatus  computer-readable storage media  and methods are disclosed for rapid development of standardized data classification hierarchies for use with Master Data Management (MDM) products  rapid development of mapping tables for data integration  including field-value level semantic mapping  and rapid mapping of a data classification hierarchy to multiple presentation views  allowing various categories of users to view MDM data in a manner customized to a particular user’s role with an organization.
[006] In some examples  the disclosed techniques utilize various modules of a universal MDM PAIP (Product Abstraction and Integration Platform). Universal MDM PAIP components address limitations of off-the-shelf MDM products by extending available features of commercial MDM Products. The PAIP can serve as a standalone  low-cost alternative to jump-start MDM efforts (e.g.  evaluation or migration efforts) and can be used to help to users select a full-fledged MDM Product. The PAIP can also provide smooth integration of external tools lacking in many commercial MDM implementations  allowing reuse of software components.
[007] Combining a PAIP architecture with a Service-Oriented Architecture (SOA) increases reusability and flexibility. Consumers of MDM services are not exposed to the complexities of a service implementation or even a particular MDM vendor’s implementation. Thus  disclosed PAIP architectures are designed to provide an MDM-agnostic platform.
[008] Data models disclosed herein can be designed taking into consideration the various domain entities and relationships that are not used in existing MDM platforms. The data models can be realized on a variety of suitable database implementations  thereby enhancing the tools-agnostic attributes of PAIP implementations.
[009] In some examples of the disclosed technology  a computer-implemented method of developing master data classifications for an MDM environment includes receiving a classification hierarchy defining one or more dimensions and one or more relations between a set of hierarchy levels  each of the hierarchy levels representing a collection of related objects in the MDM environment  and each of the dimensions representing a set of valid values for one or more nodes in the classification hierarchy. The method can further include generating a dimensional level by associating at least one of the dimensions with a level of the set of hierarchy levels  the at least one dimension defining a set of valid values for nodes associated with the dimensional level  associating the dimensional level with one or more nodes  and generating master data classifications for the MDM environment based on the classification hierarchy and the associated nodes  the master data classifications including one or more objects assigned valid values based on the dimensional level and the associated nodes. In some examples  the dimensions can be independent dimensions  dependent dimensions  or both independent and dependent dimensions.
[010] In some examples of developing master data classifications  a method further includes storing the master data classifications in an MDM repository  generating a revised classification hierarchy based on the received classification hierarchy  the dimensional level  and the associated nodes  and storing the revised classification hierarchy in an extended repository.
[011] In some examples of the disclosed technology  a computer-implemented method of generating mappings between objects in an MDM repository and a master data classification hierarchy stored in an extended repository includes receiving the master data classification hierarchy  the classification hierarchy defining relations between data elements of the classification hierarchy and definitions of valid values for one or more of the data elements  and based on the master data classification hierarchy  generating one or more mappings between the data elements and a respective one or more object fields stored in the MDM repository to produce mapped object fields.
[012] In some examples  a method of generating mappings further includes selecting an object value associated with at least one of the mapped object fields  the object value being stored in the MDM repository and generating a mapping between the object value and one or more valid values for a data element mapped to the object field  the value values being determined by at least one of the relations of the master data classification hierarchy. In some examples  the mappings are stored in an extended repository.
[013] In some examples of the disclosed technology  a computer-implemented method of generating one or more presentation hierarchies for an MDM environment includes receiving a classification hierarchy for the MDM environment  generating one or more presentation hierarchies for the MDM environment  each of the presentation hierarchies being associated with a user role in the MDM environment  each of the presentation hierarchies including one or more presentation levels  selecting one of the presentation levels  selecting a level of the classification hierarchy to map to the selected presentation level  and generating a mapping between the selected level of the presentation hierarchy and one or more corresponding nodes of the selected classification hierarchy level.
[014] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter  nor is it intended to be used to limit the scope of the claimed subject matter. The foregoing and other objects  features  and advantages of the invention will become more apparent from the following detailed description  which proceeds with reference to the accompanying figures.
BRIEF DESCRIPTION OF THE DRAWINGS
[015] FIG. 1 is a block diagram illustrating a generalized example of a suitable Master Data Management (MDM) Product Abstraction and Integration Platform (PAIP) system in which described embodiments  techniques  and technologies can be implemented.
[016] FIGS. 2A–2E illustrate data elements  relations  and nodes at several different stages of performing a method of developing an example data classification hierarchy.
[017] FIG. 3 is a flow chart that illustrates a generalized example of a method of developing a master data classification hierarchy for an MDM environment.
[018] FIG. 4 is a flow chart that illustrates a generalized example of a method of generating data elements for a master data classification hierarchy.
[019] FIG. 5 is a flow chart that illustrates a generalized example of a method of generating relations for data elements for a master data classification hierarchy.
[020] FIG. 6 is a flow chart that illustrates a generalized example of a method of generating independent and dependent dimensions for a master data classification hierarchy.
[021] FIG. 7 is a flow chart that illustrates a generalized example of a method of generating nodes and node values for a master data classification hierarchy.
[022] FIGS. 8A–8D illustrate GUI windows that can be used with certain methods for developing a master data classification hierarchy for an MDM environment.
[023] FIG. 9 is a flow chart that illustrates a generalized example of a method of generating mappings for a master data classification hierarchy for an MDM environment.
[024] FIG. 10 is a flow chart that illustrates a generalized example of a method of generating mappings  including level and node mappings for a master data classification hierarchy.
[025] FIG. 11 is a flow chart that illustrates a generalized example of a method of mapping previously-unmapped fields and/or values for an MDM system to a master data classification hierarchy.
[026] FIG. 12 illustrates a generalized example of mappings between fields and data elements  and mappings between values and valid node values as can be used in some examples of the disclosed technology.
[027] FIG. 13 is a flow chart that illustrates a generalized example of a method of mapping a presentation hierarchy level to a master data classification hierarchy.
[028] FIG. 14 illustrates a generalized example of two different presentation hierarchies that have been mapped to a master data classification hierarchy.
[029] FIG. 15 illustrates a generalized example of a suitable computing environment in which described embodiments  techniques  and technologies can be implemented.
DETAILED DESCRIPTION
Introduction
[030] This disclosure is set forth in the context of representative embodiments that are not intended to be limiting in any way.
[031] As used in this application and in the claims  the singular forms “a ” “an ” and “the” include the plural forms unless the context clearly dictates otherwise. Additionally  the term “includes” means “comprises.”
[032] The systems  methods  and apparatus disclosed herein should not be construed as being limiting in any way. Instead  this disclosure is directed toward all novel and non obvious features and aspects of the various disclosed embodiments  alone and in various combinations and sub-combinations with one another. The disclosed systems  methods  and apparatus are not limited to any specific aspect or feature or combinations thereof  nor do the disclosed systems  methods  and apparatus require that any one or more specific advantages be present or problems be solved. Furthermore  any features or aspects of the disclosed embodiments can be used in various combinations and sub-combinations with one another. Furthermore  as used herein  the term “and/or” means any one item or combination of items in the phrase.
[033] Although the operations of some of the disclosed methods are described in a particular  sequential order for convenient presentation  it should be understood that this manner of description encompasses rearrangement  unless a particular ordering is required by specific language set forth below. For example  operations described sequentially may in some cases be rearranged  omitted  or performed concurrently. Moreover  for the sake of simplicity  the attached figures may not show the various ways in which the disclosed systems  methods  and apparatus can be used in conjunction with other systems  methods  and apparatus. Additionally  the description sometimes uses terms like “receive ” “produce ” “generate ” “associate ” “select ” “search ” and “provide” to describe the disclosed methods. These terms are high-level abstractions of the actual operations that are performed. The actual operations that correspond to these terms can vary depending on the particular implementation and are readily discernible by one of ordinary skill in the art.
[034] Any of the disclosed methods can be implemented with computer-executable instructions stored on one or more computer-readable storage media (e.g.  non-transitory computer-readable media  such as one or more volatile memory components (such as DRAM or SRAM)  or nonvolatile memory components (such as hard drives) and executed on a computer. Any of the computer-executable instructions for implementing the disclosed techniques as well as any data created and used during implementation of the disclosed embodiments can be stored on one or more computer-readable media (e.g.  non-transitory computer-readable media). The computer-executable instructions can be part of  for example  a dedicated software application or a software application that is accessed or downloaded via a web browser or other software application (such as a remote computing application). Such software can be executed  for example  on a single local computer (e.g.  any suitable commercially-available computer) or in a network environment (e.g.  via the Internet  a wide-area network  a local-area network  a client-server network (such as a cloud computing network)  or other such network) using one or more network computers.
[035] For clarity  only certain selected aspects of the software-based implementations are described. Other details that are well known in the art are omitted. For example  it should be understood that the disclosed technology is not limited to any specific computer language or program. For instance  the disclosed technology can be implemented by software written in C  C++  Java  JavaScript  Perl  Python  or any other suitable programming language. Likewise  the disclosed technology is not limited to any particular computer or type of hardware. Certain details of suitable computers and hardware are well-known and need not be set forth in detail in this disclosure.
[036] Furthermore  any of the software-based embodiments (comprising  for example  computer-executable instructions for causing a computer to perform any of the disclosed methods) can be uploaded  downloaded  or remotely accessed through a suitable communication means. Such suitable communication means include  for example  the Internet  the World Wide Web  an intranet  software applications  cable (including fiber optic cable)  magnetic communications  electromagnetic communications (including RF  microwave  and infrared communications)  electronic communications  or other such communication means.
[037] Theories of operation  scientific principles  or other theoretical descriptions presented herein in reference to the systems  methods  and apparatus of this disclosure have been provided for the purposes of better understanding and are not intended to be limiting in scope. The systems  methods  and apparatus in the appended claims are not limited to those systems  methods  and apparatus that function in the manner described by such theories of operation.
[038] Any trademarks used herein are used for illustrative purposes only and are the property of their respective owners.
Example MDM PAIP Architecture
[039] FIG. 1 illustrates an exemplary universal MDM PAIP architecture 100 that complements features of the existing MDM products by providing improved user interface usability  process automation capabilities  reporting and analyzing facilities  and other features. In some examples  a Service Abstraction Layer is provided to decouple presentation and/or data layers from MDM components. A product adapter interface layer is provided such that a number of MDM components can be integrated with the MDM PAIP.
[040] An exemplary high-level architecture 100 for an MDM PAIP system is illustrated in FIG. 1. As shown  the high-level architecture 100 includes a presentation layer 110  a functional layer 120  a service enabler layer 130  and a service abstraction layer (SAL) 140. As shown in FIG. 1  these layers 110  120  130  and 140 are added to interact with an MDM domain 180 and an extended domain 190.
[041] The MDM domain 180 includes at least one MDM Repository 181  which is specific to an MDM environment (e.g.  a commercial MDM implementation). Master data is stored in the MDM repository  and additional data can be accessed within the MDM domain  for example  E-mail data 187 (e.g.  stored on an SMTP server)  unstructured data 185 (e.g.  data stored in other databases or filesystems)  and network data 186 (e.g.  data stored on SMB or NFS network shares). Access to the MDM repositories 181 can be obtained through product-specific APIs 167 that communicate with the SAL 140. The MDM domain is accessed by standard MDM vendor products 188 using an MDM vendor UI 189  which are both typically provided by a commercial MDM vendor. Hence  the vendor products 188 and UI 189 can be referred to as “out-of-the-box” components.
[042] The Extended domain 190 includes at least one extended repository 191  which can be built based on a generic data model and used to store information outside of the MDM domain 180. For example  information specific to the client and/or other information that cannot be supported by the MDM product can be stored in the extended repository 191 in the extended domain 190. Additional data  such as search configuration data 195 and indexed content 196 can also be stored in the extended domain 190  and can be used  for example  to stored index data generated by an indexer 175 and to produce search results and query results to a search result assembler 176 and query builder 177  respectively. Search functionality can be configured with a search configuration engine 172  which stores search settings as search configuration data 195. Thus  at least a portion of the extended domain 190 includes computing resources not provided by an MDM system.
[043] The presentation layer 110 is a customized user interface (UI) used to present data to users  and can include the use of a service-based UI 111 (e.g.  a UI based on Java Servlets  JavaServer Pages (JSP)  and/or JavaServer Faces) and/or an AJAX-based (Asynchronous Java and XML) UI 112. The presentation layer 110 can be a different  customized user interface than the interface that is provided out-of-the-box by the MDM vendors (e.g.  MDM Vendor UI 189). In some examples  the functional layer 120 represents services specific to a client as per-use cases. The functional layer 120 can aggregate calls on an MDM core service layer. The functional layer 120 is responsible for handling method-level transactions.
[044] In some examples  a Service Enabler Layer (SEL) 140 allows MDM services to be exposed in a protocol-friendly manner. The SEL 140 enables MDM services to be accessible by consumers using multiple communication protocols based on suitability and design choice. For example  SOAP (Simple Object Access Protocol)  HTTP (Hypertext Transfer Protocol)  RMI (Remote Method Invocation)  DWR (Direct Web Remoting)  and Web Services (e.g.  Java API for XML Web Services (JAX-WS)) protocols are examples of suitable communication protocols for enabling such MDM services.
[045] In the example shown  the SEL 130 is the layer that interacts directly with the service consuming applications via the functional layer 120 and/or presentation layer 110. By using the SEL 130  there is no direct connection between service consumers and the SAL 140  thus hiding implementation complexity from the outside world.
[046] Specifics regarding how the SAL 140 is called and the type of data that is transferred can be specified by a service contract and/or data contract. Two approaches to implementing the SEL 130 include a Web Services Approach and a Delegate Approach.
[047] In a Web Services approach  a services contract is enabled as a web service that can be consumed by the consumers using SOAP request and responses.
[048] In a delegate approach  a delegate class uses a services contract to access an MDM repository (e.g.  MDM repository 181). Delegate class methods can be consumed by other applications through requests and responses.
[049] The MDM PAIP system 100 can provide a number of infrastructure services 145  such as shown in FIG. 1  including: error handling  transaction handling  logging  caching  and communication-level security and authentication. Infrastructure services 145 can be provided by an enterprise application server environment where one or more solution components are deployed and hosted. The Service Enabler Layer 130 can provide an abstraction to leverage these infrastructure services 145 using an application server.
[050] The MDM PAIP system 100 components provide functionality  including the ability to store structured and unstructured information  to update product releases (e.g.  to update standards for data synchronization)  and to use a metadata-driven flexible data model. The MDM PAIP system 100 components can be built considering Service-Oriented Architecture (SOA) technology.
[051] In some examples  the Service Abstraction Layer (SAL) 140 provides a set of interface methods that represent atomic services that can be exposed to MDM service consumers  thus abstracting implementation details of MDM services from the consumers. Examples of interfaces that are provided in the SAL 140 include: Entity Manager  Hierarchy Manager  Meta Data Manager  Relationship Manager  Service Manager  and Security Manager. The SAL 140 interface can be used to provide system contracts and can be implemented using Java classes.
[052] The SAL 140 provides a high level of abstraction to the atomic services that are exposed on one or more MDM repositories to users. The abstraction provided by the SAL 140 loosely includes an MDM service abstraction 150  which provides access to MDM services  and a search abstraction 170  which provides access to search and indexing services based on the extended domain 190. Thus  the search abstraction 170 can be used to search for data across multiple MDM domains (e.g.  MDM domain 180).
[053] In some examples  the SAL 140 can be further defined as providing three interface layers: a core service layer  a business layer and a data access layer. Service managers can execute business logic to transparently handle requests from service consumers.
[054] The core service layer provides access to a service contract that includes interfaces to operations exposed to consumers. The core service layer also provides access to atomic methods to create and update a master data entity. In some implementations  the core service layer communicates with the business layer.
[055] The business layer handles business implementation of services. Business rules to be applied before data is persisted can be implemented in the business layer.
[056] The data access layer can be implemented through Data Access Object (DAO) layers 171 that interact with a product adapter 166 to persist data to vendor databases or Object/Relational Mapping tools (e.g.  a Hibernate framework) to persist data to the extended repository. Using DAOs 171 introduces an abstraction between the data layer and the business layer  essentially hiding the database implementation details from the business layer. Thus  using DAOs 171 acts as a façade for data access to the MDM extended repository.
[057] Value types are objects that can be serialized and can be used for communication with services. In some examples  value types depict object state but not object behavior.
[058] Domain models are feature-rich models of value objects that can have additional methods  allows the domain models to depict both object behavior and state. Domain models also provide primary key-foreign key relationships between value objects.
[059] As shown  a data synchronizer 155 is activated upon receiving trigger events from a listener. Once triggered  an incremental synchronization process is initiated to update 156 the content indexes and the extended repository with data changes detected in the MDM Repository 181. The data synchronizer 155 thus ensures that the extended repository 191 is synchronized with the MDM repository 181.
[060] The data synchronizer 155 can be implemented as an asynchronous event handling mechanism wherein messages (e.g.  messages requesting an action) are put into a queue (e.g.  a Java Message Service (JMS) queue). The data synchronizer 155 can have a listener component listening to the queue. Upon arrival of a message  the data synchronizer 155 pulls 157 data from the MDM repository 181 and pushes data (e.g.  by sending data from the MDM repository or data representing differences in the data from the MDM repository) into the extended repository 191. In examples where data synchronization occurs asynchronously  it may not be possible to provide real-time synchronized data  which can compromise accuracy when performing search operations on the extended repository.
[061] An authorization module 152 provides secure and authorized access to access information in both the custom repository and the MDM repository. In some examples  authorization can be based at least in part on the role of a respective user. The SAL exposes the security context through a Security Manager  which can be used on the extended repository 191 or the MDM repository 181.
[062] When using MDM solution services  access to a data item is controlled by a role associated with an authorized user. Information on users and roles can be maintained by administrative services. Field-level authorization is ensured before displaying data to the user  so that secure data is exposed when authorized for a given role.
[063] When using authorization features provided by a third-party tool (e.g.  MDM vendor products such as IBM WPC)  an application can use security context exposed by the third party tool. This context can be accessed using wrapper APIs that provide hooks to vendor product specific APIs.
[064] An Event Manager 160 is used for handling the events that initiate synchronization of data between the MDM product repository and the extended repository to provide search results with near- to real-time data. The Event Manager 160 can have registered listeners that subscribe to changes events associated with an entity attribute or an entity itself  for example  item attribute change or addition. Upon adding or updating an entity in the MDM repository 181  the change listener receives a notification and sends 163 a trigger to the Synchronizer module 155.
[065] Product Adapter Interface (PAI) components 165 and 166 implement wrapper classes to allow access to the MDM repository 181 and the extended repository 191  respectively  using product APIs 167 and 168. For example  the MDM product API 167 (e.g.  proprietary Java APIs) is used to access data and functions available in the MDM domain 180. The Product API 168 used to access the extended domain 190 is selected based on the configuration used to implement the extended domain. Thus  these product adapters 165 and 166 allow the MDM PAIP system 100 to bypass the regular user interface provided by the MDM product “out of the box ” and allows transparent implementation of custom user interfaces.
[066] A Flexible Domain Data Model (FDDM) 173 is offered as part of the solution provides a relational representation of the various entities and relationships that exist across industries in a MDM context. In some examples  the FDDM 173 includes customized hierarchical MDM Retail and/or Customer Package Goods (CPG) Data Models (e.g.  providing roles  permissions  etc.). The FDDM 173 can also provide hierarchy management and category-specific attributes. The FDDM 173 can be used across both retail and CPG domains.
Example Data Elements and Classification Hierarchy
[067] FIGS. 2A–2E illustrate data elements at several different stages of performing a method of developing an example data classification hierarchy that includes a number of data elements. Data elements represent a collection of similar objects in an MDM environment. For example  for an MDM PAIP system implemented for a retail merchandiser  three objects Polo  Shirt  and T-shirt could be objects associated with the data element “Tops.”
[068] A data classification hierarchy (e.g.  an MDM data classification hierarchy) can be built by organizing data elements along parent-child relations. For example  in some hierarchies  one parent data element may have multiple children  and a child data element can belong to no more than one parent. In other examples  a child data element can have more than one parent. A data classification hierarchy will have two or more hierarchy levels. For example  an organization of data elements into a hierarchy  listed from higher to lower levels as follows: Segment  Department  Class  Sub-Class  and Style  can be used as a data classification hierarchy for a retail merchandise application. Data elements include hierarchy levels and dimensions.
[069] As used herein  a “node” is an instance of a hierarchy level. For example  the following nodes belong to the hierarchy levels (in parentheses) in the retail merchandiser example above: Apparel (Segment)  Men’s Apparel (Department)  Tops (Class)  and T-shirts (Sub-Class)  and Short Sleeve Round Neck T-shirts (Style).
[070] As used herein  a “relation” defines one or more types of dependencies between data elements and/or nodes. A node can reference  or be referenced by  other nodes.
[071] Nodes can also be assigned one or more node types. A child node is a node referencing an upper node in a relation. A parent node is a node referenced by a lower node in a relation. Other nodes types can be assigned. For example  a root node is a top node in a data classification hierarchy. A root node will not reference any other parent nodes. A leaf node is a bottom node in a hierarchy. A leaf node will not be referenced by any other child nodes.
[072] FIG. 2A is a diagram 200 that illustrates a generalized example of a number of data elements that can be organized into data classification hierarchies as will be described in further detail below. The diagram 200 of FIG. 2A illustrates a number of hierarchy levels  including dimensional levels (e.g.  dimensional levels 210–212)  regular levels (e.g.  regular levels 220–226)  and a base level 230.
[073] Base levels (e.g.  base level node 230) are hierarchy levels acting as containers for data objects at a level of granularity typically used in transactions. In a sense  the other levels (e.g.  regular levels 220–226 and dimensional levels 210–212) and dimensions in a classification hierarchy are created to better group  detail  and organize base levels. Base levels can be used to describe what objects are stored  sold  and forecasted. For example  in the apparel industry  forecasting  planning  and ordering is often done at a style/color level. Hence  nodes belonging to the style/color level will be designated as base nodes.
[074] Dimensional levels (e.g.  dimensional levels 210–212) are levels having nodes that are defined by combining values resulting from a set of “dimensions” referenced by the nodes. For example a dimensional node “Department” can be derived from a Segment dimension by adding a reference to a Gender dimension. If the Gender Dimension can be assigned to the following values: Mens  Womens  Childrens  then Department-level dimensional nodes Men’s Apparel  Women’s Apparel  and Childrens’ Apparel can be derived by combining the Segment and Gender dimensions. As described further below  dimensions can be independent or dependent.
[075] Regular levels (e.g.  regular levels 220–226) are levels that are not generated through a combination of dimensions. Values for nodes assigned to regular levels can be individually generated and assigned.
[076] FIG. 2A also illustrates a number of dimensions 240–245  which are a type of data element. As used herein  there are two types of dimensions  independent dimensions (e.g.  independent dimensions 240–244) and dependent dimensions. Independent dimensions have values that are independent of node values for a respective node. For the retail merchandiser example  Gender is an example of an independent dimension  because values for the dimension can be applied to nodes regardless of the node’s associated value.
[077] Dependent dimensions are dimensions that have values dependent on an upper node value. For example  a level Sub-Class that has values that depend on an upper level (e.g.  a Class-Level Node) can be defined using dimensions. The allowable values for dependent dimensions can vary based on the department or class node values. For example  the dependent dimension “Neck Shape” may not be relevant for Dress Shirts  but be relevant for T-Shirt. Similarly  the dimension “Collar Shape” may be relevant only to Dress Shirts. In some cases  a dimension may be relevant to multiple hierarchy levels  but the actual allowable (valid) values may differ.
Example Method of Data Classification Hierarchy Development
[078] FIGS. 3–7 illustrate a number of techniques that can be used to generate MDM data classification hierarchies  including methods for generating data elements  including hierarchy levels and dimensions  defining relations between data elements  defining dimensional levels  and generating nodes and valid values for data classification hierarchies. As will be readily understood to one of ordinary skill in the art  the disclosed techniques can be used together or separately  and can be used  for example  to manipulate the exemplary data classification hierarchies illustrated in FIGS. 2A–2E. Further  FIGS. 8A–8D illustrate example GUIs that can be used with the exemplary methods.
Example Method of Generating Data Elements
[079] FIG. 3 is a flow chart 300 that outlines an exemplary method of receiving a classification hierarchy  generating dimensional levels and associated nodes  and generating master data classifications based on the classification hierarchy and associated nodes  as can be used in certain embodiments of the disclosed technology. Master data classifications can be used to classify master data in an MDM system based on multiple criteria (e.g.  multiple classifications  such as by an organization responsible for a product line or according to a standard industry classification). Master data classification can be mapped to external master data (e.g.  master data generated by business partners  competitors  or acquired MDM systems).
[080] At process block 310  a classification hierarchy is received defining one or more dimensions and one or more relations between a set of hierarchy levels. Each of the hierarchy levels represents a collection of related objects in an MDM environment  for example  objects stored in an MDM repository  as discussed above regarding FIG. 1. Each of the dimensions represents allowable values that can be assigned to nodes in the classification hierarchy. For example  a hierarchy level that is associated with a Brand dimension might have allowable node values of “Chanel ” “Nike ” and “Gucci.” The relations can include references between hierarchy levels  to establish parent/child relations. The classification hierarchy can define hierarchy levels as being higher or lower in the hierarchy based on the relations  based on a hierarchy level being designated as a root level  or based on one or more hierarchy levels being designated as leaf levels.
[081] In some examples  the classification hierarchy received at process block 310 is defined by a user using  for example  a GUI interface. In some examples  the classification hierarchy is previously defined or stored as data in an extended repository. As discussed above regarding FIG. 2A  the classification hierarchy can include a number of different types of hierarchy levels  including regular levels  base levels  and dimensional levels. The classification hierarchy can also include a number of independent and/or dependent dimensions.
[082] At process block 320  one or more dimensional levels are generated by associating a dimension with one or more of the hierarchy levels. The dimension can be independent or dependent  and defines a set of allowable values for nodes associated with the dimensional level. For the retail merchandiser example  a dimensional level associated with a Brand and Gender dimension could have nodes with valid levels of “Mens Chanel ” “Mens Nike ” “Mens Gucci ” “Womens Chanel ” “Womens Nike ” and “Womens Gucci.” Thus  dimensional levels are distinguished from regular hierarchy levels in that allowable values for nodes associated with the dimensional levels can be generated based on the associated dimensions  and therefore do not require additional input to define the dimensional level.
[083] At process block 330  one or more nodes are associated with the dimensional levels generated at process block 320. For example  the six valid values mentioned above for the dimensional level associated with a Brand and Gender dimension could be assigned as the valid values for the dimensional level. Thus  allowable values for the associated nodes can be determined based on allowable values for a particular dimension.
[084] At process block 340  master data classifications are generated for the MDM environment based on the classification hierarchy and the associated nodes. Because the associated nodes are based on dimensional levels  the master data classifications include objects having allowable values that are based on the associated nodes and respective dimensions. For example  data records and data structured stored in an MDM environment can be assigned valid values based on the valid values for a dimensional level associated at process blocks 320 and 330  above.
[085] In some examples  the master data classifications are generated as schemas for one or more MDM environments. In some examples  the master data classifications are generated as database records in the MDM environments. It should be noted that multiple master data classifications can be generated for multiple MDM environments based on a single classification hierarchy and associated nodes. Thus  the classification hierarchy acts as a golden hierarchy that can be used to generate classifications across multiple different MDM environments  which might come from different MDM vendors  and include a variety of hardware  software  and capabilities. Further  different MDM data classifications can be generated for each MDM environment  thereby allowing automatic tailoring of MDM classification data based on a single classification hierarchy
Example Method of Defining Hierarchy Levels and Dimensions
[086] FIG. 4 is a flow chart 400 of another  more detailed  exemplary method of generating hierarchy levels and dimensions for a classification hierarchy as can be used in certain embodiments of the disclosed technology. An exemplary GUI form 800 suitable for entering and selecting data as shown in the flow chart 400 is depicted in FIG. 8A.
[087] At process block 410  a data element name is entered  using  for example  a string field in a GUI form. The data element name can be arbitrary  and does not necessarily need to match object names used in the corresponding MDM environments. As shown in FIG. 8A  a data element name “Division” has been entered.
[088] At process block 420  a data element type is selected using  for example  a GUI form. As shown  a data element can be selected to be a level  and independent dimension  or a dependent dimension. As shown in FIG. 8A  a data element type “Level” has been selected. At process block 425  a determination is made whether a level or a dimension (e.g.  a dependent or independent dimension) as selected. If a level was selected  the method proceeds to process block 430  otherwise  the method proceeds to process block 440.
[089] The method proceeds to process block 430  where data is received selecting a hierarchy level type. For example  as shown in FIG. 8A  a hierarchy level type “Regular” is selected using radio buttons in a GUI. Other level types that can be selected include base levels and dimensional levels. At process block 435  the method determines whether a base level is selected and if so  proceeds to process block 470. If the level selected at process block 430 was not a base level  the method proceeds to process block 450.
[090] The method proceeds to process block 440  if a dimension data type was selected at process block 420  where it determines whether the dimension is an independent dimension or dependent dimension. If the selected dimension is an independent dimension  the method proceeds to process block 445. Otherwise  the method proceeds to process block 450.
[091] At process block 445  additional input is received to enter a number of valid values for the selected independent dimensions. For example  as shown in FIG. 8A  a user adds a new value for the selected dimension by typing the value into a GUI form. A user can also remove dimension values by selecting an existing independent dimension value and selecting a button. After receiving one or more valid dimension values for the independent dimension  the method proceeds to process block 450.
[092] At process block 450  the method determines whether the newly added data element (e.g.  a non-base level hierarchy level  an independent dimension  or a dependent dimension) will affect the nodes of a base node. Non-base levels can be identified as part of a unique combination defining the base level nodes. For example  if the newly added data element is a parent level of a base level that is used to create a new combination of values for base level nodes  then the newly added data element is determined to impact the base node numbers. For example  as shown in FIG. 8A  a user can select “Yes” in the GUI form to indicate that the new data element will affect base node numbers. If the new data element will affect base node numbers  the method proceeds to process block 460. If the new data element will not affect base node numbers  the method proceeds to process block 470.
[093] At process block 460  a level or dimension is designated as part of a unique base nodes combination for a base level. When data elements or associated nodes designated as part of a unique based nodes combination are changed  this indicates that one or more associated base nodes are to be changed as well. For the retail merchandiser example  if a new model name is assigned for a style/color code  then a change in the model name (part of a particular unique based nodes combination) implies that the style/color code should be changed as well.
[094] At process block 470  the method checks to determine whether additional data elements are to be added to the classification hierarchy. For example  a user can select the “Add Element” button in the GUI form 800 shown in FIG. 8A to indicate that additional elements are to be added. If additional data elements are to be added  the method proceeds back to process block 410  otherwise  the method proceeds to process block 480.
[095] At process block 480  entering of data elements is completed. For example  the GUI form 800 shown in FIG. 8A can be closed  and the MDM PAIP system can generate and store schema and/or other records for the classification hierarchy (e.g.  the defined levels  dimensions  and values that were generated and/or selected) in an extended repository.
[096] Some of the data elements that can be defined using the method outlined in the flow chart 400 is shown in FIG. 2A. As shown  a number of dimensions (e.g.  Brand dimension 240 and Gender dimension 241) and hierarchy levels (e.g.  regular level Enterprise 220  base level Style/Color 230  and dimensional level Sub-Class 211) have been defined. In some examples  data elements can be shown on a computer display as they are created. For example  the name of the data elements can be displayed on shapes that are color-coded or shaded to indicate the type of the data element.
[097] It should be noted that the GUI form windows 800  810  820  and 830 shown in FIGS. 8A–8D do not necessarily require execution of associated operations in a particular sequence. For example  after generating new data elements using the form shown in FIG. 8A  a user can select the “Build” tab and build hierarchy level relations  as discussed regarding FIG. 8B  then select the “Associate” tab to generate associations between levels  as discussed below regarding FIG. 8C  and then return back to the form shown in FIG. 8A to add additional data elements. Thus  the user can iteratively build a classification hierarchy.
Example Method of Defining Relations in a Data Classification Hierarchy
[098] FIG. 5 is a flow chart 500 that outlines an exemplary method of defining hierarchical relations between two or more levels of the classification hierarchy as can be used in certain embodiments of the disclosed technology.
[099] For example  by using the GUI form 820 depicted in FIG. 8B  a user can select any of the previously defined hierarchy levels (e.g.  hierarchy levels defined using the method described above regarding FIG. 4) and then establish or eliminate a relation to a different level. A user can initiate the GUI form by  for example  selecting the “Build” tab as shown in FIGS. 8A–8D.
[0100] At process block 510  a previously-defined hierarchy level is selected using  for example  a GUI screen to select from a list of hierarchy levels. After selecting a hierarchy level  the method proceeds to process block 520.
[0101] At process block 520  a relation between the selected hierarchy level and another hierarchy level defined in the classification hierarchy can be created or removed. For example  a user can select levels to add or remove as child or parent relations by using the GUI form 810 shown in FIG. 8B. As shown  unrelated levels in a list of valid levels can be selected for adding  while levels that have an existing child or parent relationship with the selected hierarchy level can be selected for removal. Depending on the relationship action selected  the method proceeds to one of the appropriate sub-process blocks 521–524 to create or remove the selected child or parent relationship.
[0102] At process block 530  the method checks to determine whether additional relations are to be created for the selected hierarchy level. For example  the user can select additional relations to add or remove for the selected hierarchy level using a GUI form as shown in FIG. 8B. If additional relations are to be created  the method proceeds to process block 520. If no additional relations are to be created for the selected level  the method proceeds to process block 540.
[0103] At process block 540  the method checks to determine whether relations are to be added for additional levels of the classification hierarchy. For example  a user can select another level using a GUI form. If additional relations are to be added  the method proceeds to process block 510. If no additional relations are to be added  the method proceeds to process block 550.
[0104] At process block 550  generation of relations is completed. For example  the GUI form 810 shown in FIG. 8B can be closed  and the MDM PAIP system can generate and store schema and/or other records for the relations defined for the classification hierarchy as entries in an extended repository.
[0105] A set of example relations that have been defined using the method outlined in the flow chart 500 is shown in the class hierarchy illustrated in diagram 201 of FIG. 2B. As shown  a number of relations have been defined between hierarchy levels. For example the Style level 224 has two parent levels  Sub-Class 211 and Sub-Division 212  and one child base level  Style/Color 230. In some examples  relations and their associated data elements can be shown on a computer display as they are created. The class hierarchy further includes one leaf level named SKU 226 and two root levels  named Category 210 and Enterprise 220. Thus  the class hierarchy includes two alternative hierarchies. Whether to use a particular root level as the top of a hierarchy (or  alternatively  more than one root level) can be determined by the role of an MDM user. For example  a user operating in a bricks-and-mortar role might use the first hierarchy having a root level Category 210  while a user operating in an E-commerce role might use the second hierarchy having a root level Enterprise 220.
Example Method of Associating Dimensions in a Data Classification Hierarchy
[0106] FIG. 6 is a flow chart 600 that outlines an exemplary method of associating dimensions with hierarchy levels as can be used in certain embodiments of the disclosed technology.
[0107] For example  by using the GUI form 820 depicted in FIG. 8C  a user can select any of the previously defined hierarchy levels (e.g.  hierarchy levels defined using the method described above regarding FIG. 4) and then establish or eliminate relations to independent or dependent dimensions. A user can initiate the GUI form 820 by  for example  selecting the “Associate” tab shown in FIGS. 8A–8D.
[0108] At process block 610  a hierarchy level is selected using  for example  a GUI screen to select from a list of hierarchy levels. In some examples  hierarchy levels are designated as dimensional levels when the respective level is created. In other examples  hierarchy levels can be designated as dimensional levels by associating a dimension with the level below at process block 620. After selecting a hierarchy level  the method proceeds to process block 620.
[0109] At process block 620  an association between the selected dimensional level (or alternatively  a regular hierarchy level that is changed to a dimensional level based on creating an association) and a dimension in the classification hierarchy can be created or an existing association can be removed. By associating a dimension to a dimensional level  a user informs the MDM PAIP system that nodes belonging to the selected hierarchy level should be defined by the associated dimension values. Hence  valid values for a level can be generated based on the dimension  instead of being designated or entered manually.
[0110] For example  a user can create or remove dimension associations using the GUI form 820 shown in FIG. 8C. As shown  a level (Subclass) and dimension (Sleeve Length) are selected. If the action of removing a dimension is selected  the method proceeds to sub process block 630  and the association between the selected dimension and hierarchy level is removed from the class hierarchy. If the action of adding a dimension is selected  the method proceeds to sub-process block 640.
[0111] At sub-process block 640  a dimension is selected from a list of dimensions that have been previously defined (for example  by creating data elements as discussed above regarding FIG. 4). The dimension can be selected by using a GUI form  or by selecting a representation of the dimension on a computer display of the class hierarchy. After selecting a dimension  the method proceeds to sub-process block 650.
[0112] At sub-process block 650  the method determines whether the selected dimension is an independent or dependent dimension. If the selected dimension is dependent  the method proceeds to sub-process block 660  otherwise  the method proceeds to sub-process block 670.
[0113] At sub-process block 660  a reference level for the selected dependent dimension can be selected. For example  the reference hierarchy level can be selected from the “Select Definition Level” list shown in the GUI form 820 of FIG. 8C. The purpose of selecting a reference level for the dependent dimension is to establish distinct dimension values for groups of nodes having the same parent node. For example  subclasses derived from different respective parent nodes can have different associated dimensions. After selecting a reference level  the method proceeds to sub process block 670.
[0114] At sub-process block 670  the method determines whether there are additional dimensions to be associated with the selected hierarchy level. If so  the method proceeds to process block 620  and additional dimensions can be selected at sub-process block 640 or dimensions can be removed at sub-process block 630. Otherwise  the method proceeds to process block 680.
[0115] At process block 680  the method determines whether there are additional hierarchy levels to be associated with dimensions. If there are additional levels  the method proceeds to process block 610 so that additional levels can be selected. Otherwise  the method proceeds to process block 690.
[0116] At process block 690  association of dimensions is completed. For example  the GUI form 820 shown in FIG. 8C can be closed  and the MDM PAIP system can generate and store schema and/or other records for the relations defined for the classification hierarchy as entries in an extended repository. Further  the user can later return to the GUI form 820 shown in FIG. 8C to add associate additional dimensions.
[0117] A set of example associations that have been defined between dimensions and hierarchy levels using the method outlined in the flow chart 600 is shown in the class hierarchy illustrated in diagram 202 of FIG. 2C. As shown  a number of dimensions are now associated with hierarchy levels. For example  relations are created such that the Segment dimension 242 is associated with the Category dimensional level 210 and the Sub-Division dimensional level 212. Thus  values for nodes of the dimensional levels can be defined using the associated dimensions. Also indicated by the dashed arrows are optional associations. For example  the dimension Channel 245 has an optional association with the Sub-Division level 212. Assignment of values to nodes from the optional associated dimension is optional  so that nodes associated with a level (e.g.  nodes associated with Sub-Division level 212) do not need to have a value assigned for the associated dimension.
Example Method of Populating Nodes in a Data Classification Hierarchy
[0118] FIG. 7 is a flow chart 700 that outlines an exemplary method of populating nodes in a data classification as can be used in certain embodiments of the disclosed technology. This includes assigning and/or defining nodes belonging to hierarchy levels  as well as defining valid values for dependent dimension. Populating nodes starting with root-level nodes of the hierarchy facilitates the creation of lower-level nodes  which can be related to previously defined nodes at higher levels of the class hierarchy.
[0119] For example  by using the GUI form 830 depicted in FIG. 8D  a user can select levels and/or parent nodes for a node  select dimension values for a previously-defined dependent dimension  and/or add node components and nodes to the class hierarchy. A user can initiate the GUI form 830 by  for example  selecting the “Populate” tab shown in FIGS. 8A–8D.
[0120] At process block 710  a hierarchy level for a node to be populated is selected using  for example  a GUI screen to select from a list of hierarchy levels. For example  FIG. 8D illustrates a set of GUI controls 831 that allows a user to identify the level for the new node by selecting a previously defined level. After selecting a level for the node  the method proceeds to process block 720.
[0121] At process block 720  the method determines whether the selected hierarchy level for the selected node is a root level. If a root level was selected  the method proceeds to process block 730  otherwise the method proceeds to process block 725.
[0122] At process block 725  a reference node in a parent level of the selected non-root node is selected. The reference node can be selected using  for example  a GUI form 830 to select from a list of parent nodes. For example  a user can select a level and reference nodes using a control 831 including a list of levels and parent nodes  as shown in FIG. 8D. Subclasses derived from different respective parent nodes can have different associated dimensions. For example  subclasses derived from a parent node “Shirt” will have different dimensions than subclasses derived from a parent node “Polos.”
[0123] At process block 730  the method determines whether a dimensional or regular level has been selected. If the selected level is a regular level  the method proceeds to process block 740 to be defined individually based on  for example  user input. If the selected level is a dimensional level  then the method proceeds to process block 750 to be defined based on a previously-selected set of dimensional values. Base levels can be processed as regular levels (e.g.  and be defined individually by proceeding to process block 740) or dimensional levels (e.g.  and be defined by selecting from a set of values by proceeding to process block 750).
[0124] At process block 740  a name for the selected (regular) node is entered. This name will be added to the list of valid values for the reference level associated with the selected node. For example  a user can directly enter a new node name using an entry field 832 in a GUI form 830.
[0125] At process block 750  a number of nodes are generated be selecting valid combinations of dimensions associated with a hierarchy level. For example  a user can select a dimension (e.g.  a sleeve length dimension) and a valid value (e.g.  long sleeve) by selecting from a list interface element 833 in a GUI window 830. Hence  values to be associated with a dimension can be selected based on previously defined valid values for the dimension. Thus  for dimensional levels  the node values will be based on a combination of different values associated with one or more dimensions. This allows a user to select specific dimension values for the new node. These dimension values can be saved even if they are not used in the final node name. Not all possible combinations or dimensions associated with a hierarchy level have to be used as valid levels for a given node. Once valid combinations have been selected  and node can be added using a control 836 as shown in FIG. 8D.
[0126] At process block 760  the method determines whether a reference level is to be added to the dimension definition. If a reference level is to be added to the dimension definition  the method proceeds to process block 770. Otherwise  the method proceeds to process block 780.
[0127] At process block 770  a number of valid dependent dimension values can be entered for the selected node. For example  a user can select a defined dependent dimension  and enter or remove new dimension values for the dimension using controls 834 of a GUI window 830. For example  the defined dimension “sleeve length” can apply to multiple nodes  but may have different sets of values for nodes having the parent node “shirt” vs. nodes having the parent node “Polos.”
[0128] For dependent dimensions  defining dimension to definition-level relation is for guidance. The actual relation is defined at the node level. Similarly  even when the same dimension is associated to multiple definition nodes  it can include different values based on the node. After adding or removing dependent dimension values  the method proceeds to process block 775.
[0129] At process block 775  the method determines whether additional node values are to be added  and if so  returns to process block 770. Otherwise  the method proceeds to process block 780.
[0130] At process block 780  a name for the new node is created. For example  the new name can be created as a combination of upper node and selected dimension values. In some examples  a user can select values to be used for a name using GUI controls 835 of a GUI form 830. A user can select node components for a name  or select previously defined node names from a nodes list. If a user determines a name that is unique or a name is becoming too long  at least some of the values for the name combination can be ignored or overridden. In this case  other node  value  and level associations are still stored for reference.
[0131] At process block 785  the method determines whether additional nodes are to be added  using for example  input received using a GUI form 830. If additional nodes are to be added  the method proceeds to process block 730. Otherwise  the method proceeds to process block 790.
[0132] At process block 790  generation of valid node values is completed. For example  the GUI screen shown in FIG. 8D can be closed  and the MDM PAIP system can generate and store schema and/or other records for the node values defined for the classification hierarchy as entries in an extended repository.
[0133] A set of example dependent dimension values that have been defined using the method outlined in the flow chart 700 is shown in the class hierarchy illustrated in diagram 203 of FIG. 2D. Some hierarchy levels and dimensions that were shown in FIG. 2C are not shown for clarity. As shown  a number of valid node values are now associated with a number of hierarchy levels. For example  the dependent dimension “Sleeve Length” 250 is associated with the Class hierarchy level 221 and has valid node values Short 251  Long 252  and ¾ 253  which are associated with the Sub-Class hierarchy level 211. Specific values of the dependent dimension Sleeve Length 250 vary based on the Class Node selected. Similarly  the dependent dimension “T-shirt” 260 is associated with the Class hierarchy level and has valid node values Crew Neck 261  V Neck 262  and A-Shirt 263. Values associated with the independent dimensions are based on the associated dependent dimension and are not assigned as with the dependent dimensions.
[0134] Also shown in FIG. 2D is an arrow 265  which indicates that the dimension “Design Features Groups” is defined at the category level. Thus  different values associated with Design Features Groups can be applied to different Categories.
[0135] A set of valid independent dimension values and valid combinations that are generated using the method outlined in the flow chart 700 is depicted in the diagram 204 of FIG. 2E. As shown  a number of values have been generated for each independent dimension (e.g.  valid values 270 Men’s  Women’s  Boys  Girls  and Genderless for the independent dimension Gender 241) using the techniques described above regarding process block 740. A number of valid combinations 275 of the valid values 270 have also been generated (e.g.  the combination of values of Chanel  Women’s  and Apparel from the independent dimensions Brand  Gender  and Segment  respectively  have been designated as a valid combinations). These combinations 275 can be generated using the techniques described above regarding process block 750.
[0136] It should be noted that the method of designating valid values and combinations described regarding the flow chart 700 can be an iterative process. For example  after generating the valid combinations 275  additional valid dimension values can be specified. As shown in FIG. 2E  a set 280 of valid values have been generated for nodes in the hierarchy level Category 210. Further  the valid value “Bottoms” 281 is depicted as assigned to the valid combination “Chanel Women’s Apparel” 276. The Category value Tops has been defined to have a number of valid values 285 in the Class category 221. Next  additional valid combinations of the additional values 280 and 285 can be designated as valid combinations for the Sub-Division 212 and Sub-Class 211 hierarchy levels  respectively. These are possible node names  or nodes values from source or destination systems. As an example  an analysis tool reads attribute names for a third party data extract and allows mapping to the newly-defined structure. Similarly  actual field values can be mapped to the predefined dimension value set.
[0137] Thus  an interactive and iterative method of defining a class hierarchy for an MDM PAIP environment is disclosed. The disclosed methods encourage the use of dimensional levels  which can in turn encourages MDM users to develop more standardized hierarchies that can be used in varying business roles in an MDM system. The resulting class hierarchy can be used in a number of ways to define and use data in one or more MDM environments  and several examples are provided below.
Example Method of Generating Data Feed / Master Data Mappings
[0138] FIG. 9 is a flow chart 900 that outlines an exemplary method for developing mapping tables for data integration using data classification hierarchies  including the use of field value level semantic mapping  as can be used in certain embodiments of the disclosed technology. The mapping can be used with “main” classification hierarchies  which are designated to be used for describing the structure that MDM and external repository data will be stored in  and “alternative” (or “presentation”) classification hierarchies  which can be used to create data views for specific groups of stakeholders or users. Alternative data classification hierarchies can be mapped to a main data classification hierarchy. In some examples  a main classification hierarchy can also be used as a presentation hierarchy.
[0139] At process block 910  a master data classification hierarchy including relations between one or more data elements of the hierarchy is received. The data classification hierarchy includes valid values for at least one of the data elements. The master data classification hierarchy can be generated in a number of different ways. For example  the master data classification hierarchy can be generated using the exemplary techniques described above regarding FIGS. 3–7.
[0140] The master data classification hierarchy received at process block 910 includes a description of the hierarchy levels for the MDM system  and can include  for example  data elements such as hierarchy levels and dimensions  as well as nodes and valid values and combinations for the nodes. In some examples  additional format data is received from an extended MDM repository that describes data formats  attributes  and/or fields in an MDM repository or data formats  attributes  and/or fields in an extended MDM repository. After receiving the master data classification hierarchy  the method proceeds to process block 920.
[0141] At process block 920  a number of mappings between the data elements and object fields stored in an MDM repository are generated. In some examples  at least some of the mappings are generated automatically (e.g.  by analyzing an MDM repository  MDM extended repository  or using a semantic dictionary). In some examples  at least some of the mappings are generated using user input generated using a GUI. The mappings can include mappings from the data feed to the MDM system data elements and/or mappings from the MDM system to the data feed. In some examples  the mappings describe transforms from data elements in the classification hierarchy to objects stored in the MDM repository  or vice versa  or both.
[0142] As shown in FIG. 12  a number of data feed fields (e.g.  data fields 1210 and 1211) have been mapped to a number of dimension-based fields (e.g.  for a classification in an MDM system. For example  data in the MDM system stored as records corresponding to the object field “Label” 1210 are mapped to the “Brand” data element 1220 in the classification hierarchy. In this way  multiple MDM systems can be mapped to the correct corresponding fields in a generic classification hierarchy. The mappings between fields and values are not necessarily one-to-one. For example  the object field “Department” 1211 is mapped to both the “Gender” data element 1221 and the “Segment” data element 1222. Mappings between data feed fields and dimension-based fields can be provided to analysis tools  which examine descriptions associated with the mapped fields and recommend values for the corresponding attributes. For example  source data fields may not have attributes for fit  neck  or color  but instead these attributes are included in the data field description. The analysis tools tool can parse the data field description (e.g.  using string parsing routines) and provide a preliminary allocation to attributes. The generated mappings can then be reviewed and approved by users.
[0143] In some examples  a user interface (e.g.  an interface implemented using the Ajax Reach UI Module) guides users through generation of mappings. For example  user interface users can import a source or destination Format and selects a field or data element from a list of available fields. Then  for the selected field or data element  the user can select a matching hierarchy level to map. In some examples  the user can “drag and drop” icons representing fields onto icons representing data elements  or vice versa  or both  to generate mappings. In some examples  a user interface is used to generate a semantic mapping dictionary  which can be used to map field values. After generating the mapping(s)  the method proceeds to process block 930.
[0144] At process block 930  an object value associated with an object field mapped at process block 920 is selected. For example  the object value (which is stored in an MDM repository) can be selected using a GUI or a semantic mapping dictionary. In some examples  a user can drag and drop icons representing the values to generate the mappings. After selecting an object value  the method proceeds to process block 940.
[0145] At process block 940  a mapping between the object value and at least one of the valid values for a data element mapped to the object field. For example  the user can enter or select possible values and map them to nodes of a previously mapped hierarchy level in the master data classification hierarchy. Thus  the method can be used to capture semantically equivalent terms or commonly misspelled words.
[0146] The mappings generated at process blocks 910–940 can be stored for future use as entries in an extended repository.
Example Method of Generating Data Feed / Master Data Mappings
[0147] FIG. 10 is a flow chart 1000 of another  more detailed  exemplary method of developing data element mappings and semantic matches using data classification hierarchies  as can be used in certain embodiments of the disclosed technology.
[0148] At process block 1010  source field data and destination field data is extracted from one or more MDM systems (e.g.  from an MDM repository). The source and destination fields extracted are to be mapped to a classification hierarchy. In this way  fields for generating mapping from/to source/destination fields in a single MDM system  or source/destination fields across multiple MDM systems  can be generated. In some examples  a filter can be applied to extract a subset of the available source/destination fields. In some examples  the extracted fields can be displayed as a selectable list  or as icons representing the extracted fields. For example  extracted fields can be represented as rectangles  as shown in FIG. 12. After extracting source and destination field data  the method proceeds to process block 1020.
[0149] At process block 1020  one or more of the source and/or destination fields extracted at process block 1010 are selected. For example  if the fields are displayed as a list or a number of icons in a GUI  a user can select one or more fields from the list  or click on icons for one or more fields. After selecting the source and/or destination fields  the method proceeds to process block 1030.
[0150] At process block 1030  one or more data elements in a data classification hierarchy are selected to map to the source and/or destination fields selected at process block 1020. For example  as shown in FIG. 12  the Label source field 1210 has been mapped to the Brand data element 1220. In some examples  the data elements are displayed as icons  and arrows showing the relations are displayed after selecting a data element to map. In some examples  the mapping can be performed based on a text search performed based on identified matches and semantic rules. After selecting one or more data elements  the method proceeds to process block 1040.
[0151] At process block 1040  one or more of the source and/or destination values corresponding to the source and/or destination fields selected at process block 1020 are selected. In some examples  a GUI can be used to selected values in a similar manner to that discussed above regarding process block 1020. After selecting one or more source and/or destination values  the method proceeds to process block 1050.
[0152] At process block 1050  one or more nodes in a data classification hierarchy are selected to map to the source and/or destination values selected at process block 1040. For example  as shown in FIG. 12  the Short Sleeve source value 1230 has been mapped to the S/S node 1240. In some examples  the values are displayed as icons  and arrows showing the mappings are displayed after selecting a value to map. In some examples  the nodes to match to can be automatically selected by a matching algorithm (e.g.  a string matching algorithm) that determines likely matches for the source and/or destination values. After mapping one or more values  the method proceeds to process block 1060.
[0153] At process block 1060  a number of acceptable semantic matches are entered. For example  during the matching process described above regarding process blocks 1010–1050  a user is matching imported/exported fields to a solution level and a node. This matching can result in complex node names (e.g.  “Shirt V-neck Long Sleeve . . .”)  so there is a possibility that multiple matches from external sources may have to be mapped to the defined set. After identifying new possible valid matches for a particular node name  they can be entered in a matching engine  thereby training the matching engine. After entering an acceptable semantic matching  the method proceeds to process block 1070.
[0154] At process block 1070  the method determines whether additional semantic matches are to be made. If so  the method proceeds to process block 1060. For example  if a new term or multiple matches are determined  additional semantic matches are generated by proceeding to process block 1060  otherwise  the method proceeds to process block 1080.
[0155] At process block 1080  the method determines whether there are additional source/destination values to map. If so  the method proceeds to process block 1040 to select additional values for the selected field/data element combination  otherwise  the method proceeds to process block 1090. For example  a user can select additional values to map using a GUI  or a non-empty list of values can be processed to determine whether there are additional values to map.
[0156] At process block 1090  the method determines whether there are additional source/destination fields to map. If so  the method proceeds to process block 1020 to select additional fields in an MDM environment. Otherwise  the method proceeds to process block 1095.
[0157] At process block 1095  entering of data elements and node mappings is completed. For example  the MDM PAIP system can generate and store schema and/or other records for the mappings generated at process block 1010–1090 in an extended repository. In some examples  a semantic dictionary is generated that includes the generated mappings.
Example Method of Mapping Unmapped Values and Nodes
[0158] FIG. 11 is a flow chart 1100 of another  more detailed  exemplary method of developing data element mappings and semantic matches by matching previously unmatched values and nodes  as can be used in certain embodiments of the disclosed technology. In particular  the flow chart 1100 outlines a method of mapping data import errors to a semantic dictionary. For example  a semantic dictionary can store a number of valid semantic matches determined at process block 1060 of the method outlined in FIG. 10.
[0159] For example  when MDM data is imported based on previously-defined class hierarchies and previously-defined semantic mappings  there can be some values that will have a valid match defined. For example  changes to source/destination fields or values  or changes to data elements or nodes in a class hierarchy  can result in undefined matches.
[0160] At process block 1110  an import/export routine is executed to process data based on previously-mapped source/destination fields and values in an MDM system. The import/export routine determines whether any fields or values are not matched  and if so  adds them to a list. After processing the import/export on an MDM system  the method proceeds to process block 1120.
[0161] At process block 1120  a number of values and/or fields that are not matched are displayed. For example  a list of values and corresponding records to be mapped can be displayed  with any unmatched values highlighted in a GUI.
[0162] At process block 1130  mappings for unmatched fields/values and corresponding data elements/nodes can be selected. For example  a user can select an unmatched value and select the appropriate matching level and node in a class hierarchy that is displayed using a GUI. After the proposed mapping(s) for unmatched fields/values are selected  the method proceeds to process block 1140.
[0163] At process block 1140  entering of the proposed mapping(s) generated at process block 1130 is completed. For example  the MDM PAIP system can generate and store schema and/or other records for the proposed mappings in an extended repository. In some examples  a semantic dictionary is generated that includes the proposed mappings. After storing data for the proposed mappings  the method proceeds to process block 1150.
[0164] At process block 1150  the method determines whether there are additional unmatched values based on the list that was generated at process block 1120. If there are additional unmatched values  the method proceeds to process block 1130 to select additional unmatched fields or values. Otherwise  the method proceeds to process block 1160.
[0165] At process block 1160  mapping of unmatched fields and/or values is completed. For the MDM PAIP system can generate and store schema and/or other records for the mappings in an extended repository.
Example Method of Mapping to Multiple Presentation Hierarchies
[0166] FIG. 13 is a flow chart 1300 that outlines an exemplary method of generating a presentation hierarchy for a number of presentation views and mapping corresponding nodes of a classification hierarchy to a level of the presentation hierarchy  as can be used in certain embodiments of the disclosed technology. Mapping a classification hierarchy to multiple presentation views allows various categories or roles of users to see data presented in a manner customized to their role.
[0167] A user interface guides a user through a group of activities  thus enabling the mapping of a classification hierarchy to multiple presentation hierarchies.
[0168] At process block 1310  a classification hierarchy for an MDM environment is received. The classification hierarchy can be defined using similar techniques as those described earlier in this regarding FIGS. 3–7. After receiving the classification hierarchy  the method proceeds to process block 1320.
[0169] At process block 1320  one or more presentation hierarchies are defined. For example  a user can define a desired presentation hierarchy using a GUI  including the use of similar techniques as those described above regarding classification hierarchies. After defining the presentation hierarchies  the method proceeds to process block 1330.
[0170] At process block 1330  a level of the presentation hierarchy is selected. For example  a user can select a presentation hierarchy level form a list or from a graphical representation of the presentation hierarchy defined at process block 1320. After selecting a presentation hierarchy level  the method proceeds to process block 1340.
[0171] At process block 1340  a matching hierarchy level of the classification hierarchy is selected. For example  a user can select a level of a main classification hierarchy  or a level of an alternative hierarchy level. A number of nodes associated with the selected level of the classification hierarchy can be associated to the selected presentation hierarchy level. In some examples  a user can select all the corresponding nodes  or only a subset of corresponding nodes  depending on the desired classification view. After the classification hierarchy level and corresponding nodes  the method proceeds to process block 1350.
[0172] At process block 1350  the method determines whether there are additional nodes or levels of the classification hierarchy to be mapped. For example  a user can enter input to indicate whether there additional nodes or levels to be mapped. If there are additional nodes or levels to be mapped  the method proceeds to process block 1340  otherwise the method proceeds to process block 1360.
[0173] At process block 1360  the method determines whether there are additional presentation levels of one of the presentation hierarchies generated at process block 1320 to be mapped. If there are additional presentation levels to be mapped  the method proceeds to process block 1330 to select additional presentation levels  otherwise the method proceeds to process block 1370.
[0174] At process block 1370  the MDM PAIP system can generate and store schema and/or other records for the presentation views as entries in an extended repository. The data for the presentation views can be used by a GUI to present MDM environment data to users based on a selected presentation view.
[0175] For example  FIG. 14 is a diagram 1400 that illustrates two presentation hierarchies 1410 and 1450. As shown  a first presentation hierarchy 1410 includes a root level “Mens Apparel” 1415 that has three child levels 1420–1422. Each of the child levels includes a number of values (e.g.  Polos 1430  Shorts 1431  and Vests 1432). This first presentation hierarchy 1410 was developed for use by a user in an internal product role within an organization. Also shown is a second presentation hierarchy 1450 developed for a use by a user in an E-commerce product role within the same organization. As shown  a root node Mens Apparel 1455 has five child levels 1460–1464  and includes nodes defined at a different level than the first presentation hierarchy 1410. For example  the node value “Polos” 1430 is located under the Tops level 1420 in the first presentation hierarchy 1410  but the same node value “Polos” is located under the Polos level 1461 in the second presentation hierarchy. In this way  the same underlying data in the MDM environment can be arranged as desired for different user roles without the need to redefine the underlying MDM hierarchy. For example  polo shirts (e.g.  shirts associated with the Polos value 1430) might be sold and tracked using a distinct category online (e.g.  a website developed based on the E-commerce Product hierarchy 1450) while the same product might be sold and tracked together with T-shirts and Sweaters at bricks-and-mortar locations.
Example Computing Environment
[0176] FIG. 15 illustrates a generalized example of a suitable computing environment 1500 in which described embodiments  techniques  and technologies may be implemented. For example  the computing environment 1500 can implement generating class hierarchies  mapping class hierarchies  and developing presentation hierarchies  as described above.
[0177] The computing environment 1500 is not intended to suggest any limitation as to scope of use or functionality of the technology  as the technology may be implemented in diverse general-purpose or special-purpose computing environments. For example  the disclosed technology may be implemented with other computer system configurations  including hand held devices  multiprocessor systems  microprocessor-based or programmable consumer electronics  network PCs  minicomputers  mainframe computers  and the like. The disclosed technology may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment  program modules may be located in both local and remote memory storage devices.
[0178] With reference to FIG. 15  the computing environment 1500 includes at least one central processing unit 1510 and memory 1520. In FIG. 15  this most basic configuration 1530 is included within a dashed line. The central processing unit 1510 executes computer-executable instructions and may be a real or a virtual processor. In a multi-processing system  multiple processing units execute computer-executable instructions to increase processing power and as such  multiple processors can be running simultaneously. The memory 1520 may be volatile memory (e.g.  registers  cache  RAM)  non-volatile memory (e.g.  ROM  EEPROM  flash memory  etc.)  or some combination of the two. The memory 1520 stores software 1580 that can  for example  implement the technologies described herein. A computing environment may have additional features. For example  the computing environment 1500 includes storage 1540  one or more input devices 1550  one or more output devices 1560  and one or more communication connections 1570. An interconnection mechanism (not shown) such as a bus  a controller  or a network  interconnects the components of the computing environment 1500. Typically  operating system software (not shown) provides an operating environment for other software executing in the computing environment 1500  and coordinates activities of the components of the computing environment 1500.
[0179] The storage 1540 may be removable or non-removable  and includes magnetic disks  magnetic tapes or cassettes  CD-ROMs  CD-RWs  DVDs  or any other medium which can be used to store information and that can be accessed within the computing environment 1500. The storage 1540 stores instructions for the software 1580 and image data  which can implement technologies described herein.
[0180] The input device(s) 1550 may be a touch input device  such as a keyboard  keypad  mouse  touch screen display  pen  or trackball  a voice input device  a scanning device  or another device  that provides input to the computing environment 1500. For audio  the input device(s) 1550 may be a sound card or similar device that accepts audio input in analog or digital form  or a CD-ROM reader that provides audio samples to the computing environment 1500. The output device(s) 1560 may be a display  printer  speaker  CD-writer  or another device that provides output from the computing environment 1500.
[0181] The communication connection(s) 1570 enable communication over a communication medium (e.g.  a connecting network) to another computing entity. The communication medium conveys information such as computer-executable instructions  compressed graphics information  video  or other data in a modulated data signal.
[0182] Some embodiments of the disclosed methods can be performed using computer-executable instructions implementing all or a portion of the disclosed technology in a computing cloud 1590. Storing and persisting MDM records and classification hierarchies can be performed on servers located in the computing cloud 1590.
[0183] Computer-readable media are any available media that can be accessed within a computing environment 1500. By way of example  and not limitation  with the computing environment 1500  computer-readable media include memory 1520 and/or storage 1540. As should be readily understood  the term computer-readable storage media includes the media for data storage such as memory 1520 and storage 1540  and not transmission media such as modulated data signals.
[0184] Any of the methods described herein can be performed via one or more computer-readable media (e.g.  storage or other tangible media) comprising (e.g.  having or storing) computer-executable instructions for performing (e.g.  causing a computing device to perform) such methods. Operation can be fully automatic  semi automatic  or involve manual intervention.
[0185] Having described and illustrated the principles of our innovations in the detailed description and accompanying drawings  it will be recognized that the various embodiments can be modified in arrangement and detail without departing from such principles. It should be understood that the programs  processes  or methods described herein are not related or limited to any particular type of computing environment  unless indicated otherwise. Various types of general purpose or specialized computing environments may be used with or perform operations in accordance with the teachings described herein. Elements of embodiments shown in software may be implemented in hardware and vice versa.
[0186] In view of the many possible embodiments to which the principles of the disclosed invention may be applied  it should be recognized that the illustrated embodiments are only preferred examples of the invention and should not be taken as limiting the scope of the invention. Rather  the scope of the invention is defined by the following claims. We therefore claim as our invention all that comes within the scope of these claims and their equivalents.

We claim:

1. A computer-implemented method of developing master data classifications for a master data management (MDM) environment  the method comprising:
receiving a classification hierarchy defining one or more dimensions and one or more relations between a set of hierarchy levels  each of the hierarchy levels representing a collection of related objects in the MDM environment  and each of the dimensions representing a set of valid values for one or more nodes in the classification hierarchy;
generating a dimensional level by associating at least one of the dimensions with a level of the set of hierarchy levels  the at least one dimension defining a set of valid values for nodes associated with the dimensional level;
associating the dimensional level with one or more nodes; and
generating master data classifications for the MDM environment based on the classification hierarchy and the associated nodes  the master data classifications including one or more objects assigned valid values based on the dimensional level and the associated nodes.

2. The method of claim 1  wherein the at least one dimension is an independent dimension being associable to a respective one or more of the hierarchy levels regardless of values for nodes associated with the respective hierarchy level.

3. The method of claim 1  wherein the at least one dimension is a dependent dimension  wherein valid values for nodes associated with a respective one or more of the hierarchy levels for the dependent dimension are based at least in part on a value for a different node in the classification hierarchy.

4. The method of claim 1  wherein the generating the master data classifications further comprises:
receiving first input indicating that one of the dimensions is an independent dimension;
receiving second input indicating valid values for the independent dimension;
defining one or more dimensional levels based on the indicated valid values assigned to the indicated independent dimension.

5. The method of claim 1  further comprising:
receiving first input designating a second node in the classification hierarchy as a dimensional node;
receiving selection input selecting two or more nodes to be associated with the second node as dimensional combinations; and
based on the selected nodes  selecting valid dimension values for the second node  and wherein the one or more master data classifications includes the valid dimension values for the second node.

6. The method of claim 1  further comprising defining the classification hierarchy by defining hierarchical relations between two or more levels of the classification hierarchy.

7. The method of claim 1  further comprising:
storing the master data classifications in an MDM repository;
generating a revised classification hierarchy based on the received classification hierarchy  the dimensional level  and the associated nodes; and
storing the revised classification hierarchy in an extended repository.

8. The method of claim 1  further comprising generating valid combinations for master data classification objects based on the dimensional level.

9. The method of claim 1  further comprising storing schemas  database records  or schemas and database records for the master data classifications in the MDM environment.

10. One or more computer-readable media storing computer-readable instructions that when executed by a computer  cause the computer to perform the method of claim 1.

11. A system  comprising:
one or more processors; and
one or more computer-readable media storing computer-readable instructions that when executed by the processors  cause the computer to perform the method of claim 1.

12. A computer-implemented method of generating mappings between objects in an MDM repository and a master data classification hierarchy stored in an extended repository  the method comprising:
receiving the master data classification hierarchy  the classification hierarchy defining relations between data elements of the classification hierarchy  the classification hierarchy including definitions of valid values for one or more of the data elements; and
based on the master data classification hierarchy  generating one or more mappings between the data elements and a respective one or more object fields stored in the MDM repository to produce mapped object fields.

13. The method of claim 12  further comprising:
selecting an object value associated with at least one of the mapped object fields  the object value being stored in the MDM repository; and
generating a mapping between the object value and one or more valid values for a data element mapped to the object field  the value values being determined by at least one of the relations of the master data classification hierarchy.

14. The method of claim 12  further comprising:
generating a set of unmapped object fields and/or unmapped object values stored in the MDM repository;
generating a set of unmapped data elements and/or unmapped valid values in the classification hierarchy;
receiving input indicating a mapping between at least one or more of the unmapped object fields and/or object values and at least one or more of the unmapped data elements and/or an unmapped valid values; and
based on the indicated mapping  generating a mapping between the indicated fields and/or values.

15. The method of claim 12  further comprising storing the mappings in an extended repository.

16. The method of claim 12  wherein the mappings are generated using drag-and-drop functionality of a graphical user interface.

17. The method of claim 12  further comprising storing object data in the MDM repository based on the mapping.

18. A computer-implemented method of generating one or more presentation hierarchies for an MDM environment  the method comprising:
receiving a classification hierarchy for the MDM environment 
generating one or more presentation hierarchies for the MDM environment  each of the presentation hierarchies being associated with a user role in the MDM environment  each of the presentation hierarchies including one or more presentation levels;
selecting one of the presentation levels;
selecting a level of the classification hierarchy to map to the selected presentation level; and
generating a mapping between the selected level of the presentation hierarchy and one or more corresponding nodes of the selected classification hierarchy level.

19. The method of claim 18  further comprising presenting data from the MDM environment based on a selected presentation hierarchy  the presentation hierarchy being selected based on a user role in the MDM environment.

20. The method of claim 18  wherein a first one of the presentation views corresponds to a first presentation hierarchy and a second one of the presentation views corresponds to second presentation hierarchy  and wherein one or more nodes of the first presentation hierarchy are mapped differently than one or more nodes of the second presentation hierarchy.

Dated this 02nd day of September  2011

Swati Srivastava
IP Officer
Infosys Limited
Patent Agent No 110