Master Data Management PDF

Master Data Management Master Data Management Introduction........................................................................................... 1 Overview............................................................................................... 2 Enterprise data....................................................................................... 4 Transactional Data.............................................................................. 4 Operational MDM.......................................................................... 5 Analytical Data................................................................................... 5 Analytical MDM.............................................................................. 5 Master Data........................................................................................ 5 Enterprise MDM............................................................................ 6 Information Architecture....................................................................... 6 Operational Applications................................................................... 6 Enterprise Application Integration (EAI)........................................ 7 Service Oriented Architecture (SOA).............................................. 7 The Data Quality Problem.............................................................. 7 Analytical Systems............................................................................. 8 Enterprise Data Warehousing (EDW) and Data Marts.................... 8 Extraction, Transformation, and Loading (ETL)............................. 9 Business Intelligence (BI)................................................................ 9 The Data Quality Problem.............................................................. 9 Ideal Information Architecture......................................................... 10 Oracle Information Architecture...................................................... 11 Master Data Management Processes..................................................... 13 Profile.............................................................................................. 14 Consolidate...................................................................................... 15 Govern............................................................................................. 15 Share................................................................................................ 15 Leverage.......................................................................................... 16 Oracle MDM High Level Architecture................................................. 16 MDM Platform Layer...................................................................... 17 Application Integration Services.................................................... 17 Enterprise Service Bus............................................................. 17 Business Process Orchestration Services.................................. 17 Business Rules......................................................................... 18 Event-Driven Services............................................................. 19 Identity Management............................................................... 19 Web Services Management....................................................... 19 Analytic Services.......................................................................... 20 Enterprise Performance Management...................................... 20 Data Warehousing................................................................... 20 Business Intelligence................................................................ 21 Master Data Management ii Publishing Services.................................................................. 21 Data Migration Services........................................................... 21 High Availability and Scalability.................................................... 22 Real Application Clusters.......................................................... 22 Mixed Workloads..................................................................... 22 Exadata.................................................................................... 22 Application Integration Architecture............................................. 23 AIA Layers.............................................................................. 23 Common Object Methodology................................................ 24 MDM Foundation Packs.......................................................... 24 MDM Process Integration Packs.............................................. 24 MDM Aware Applications............................................................. 25 Composite Application Development............................................ 25 Oracle Data Quality Services........................................................ 25 Oracle Customer Data Quality Servers..................................... 27 Product Data Quality............................................................... 29 End-To-End Data Quality....................................................... 31 Application Development Environment........................................ 32 MDM Applications Layer..................................................................... 32 MDM Pillars.................................................................................... 32 Oracle Customer Hub....................................................................... 33 Customer Data Model................................................................... 34 Consolidate................................................................................... 35 Cleanse......................................................................................... 36 Govern.......................................................................................... 36 Share............................................................................................ 38 Business Benefits.......................................................................... 39 Product Hub..................................................................................... 39 Import Workbench....................................................................... 40 Catalog Administration................................................................. 41 New Product Introduction............................................................ 41 Product Data Synchronization...................................................... 41 Oracle Site Hub................................................................................ 42 Golden Record for Site Data......................................................... 43 Application Integration................................................................. 43 Effective Site Analysis and Google Integration.............................. 43 Oracle Supplier Hub......................................................................... 45 Consolidate................................................................................... 45 Cleanse......................................................................................... 46 Govern.......................................................................................... 46 Share............................................................................................ 46 Supplier Lifecycle Management....................................................... 46 Business Benefits.......................................................................... 46 Oracle Hyperion Data Relationship Management............................. 47 Automated Attribute Management................................................ 48 Best-of-Breed Hierarchy Management........................................... 49 Integration with Operational and Workflow Systems.................... 49 Import, Blend, and Export to Synchronize Master Data................ 49 Versioning and Modeling Capabilities to Improve Analysis.......... 50 Master Data Management iii MDM Data Governance and Industries Layer...................................... 50 MDM Industry Verticalization..................................................... 50 Higher Education Constituent Hub.......................................... 50 Product Hub for Retail............................................................ 51 Product Hub for Communications........................................... 51 Data Governance............................................................................. 52 Data Watch and Repair for MDM................................................. 53 MDM Implementation Best Practices............................................... 54 Build vs Buy..................................................................................... 55 Conclusion........................................................................................... 56 Master Data Management iv Master Data Management INTRODUCTION Fragmented inconsistent Product data slows time-to-market, creates supply chain inefficiencies, results in weaker than expected market penetration, and drives up the cost of compliance. Fragmented inconsistent Customer data hides revenue recognition, introduces risk, creates sales inefficiencies, and results in misguided marketing campaigns and lost customer loyalty1. Fragmented and inconsistent Supplier data reduces supply chain efficiencies, negatively impacts spend control initiatives, and increases the risk of supplier exceptions. “Product”, “Customer”, and “Supplier” are only three of a large number of key business entities we refer to as Master Data. Master Data is the critical business information supporting the transactional and analytical operations of the enterprise. Master Data Management (MDM) is a combination of applications and technologies that consolidates, cleans, and augments this corporate master data, and synchronizes it with all applications, business processes, and analytical tools. This results in significant improvements in operational efficiency, reporting, and fact based decision-making. Over the last several decades, IT landscapes have grown into complex arrays of different systems, applications, and technologies. This fragmented environment has created significant data problems. These data problems are breaking business processes; impeding Customer Relationship Management (CRM), Enterprise Resource Planning (ERP), and Supply Chain Management (SCM) initiatives; corrupting analytics; and costing corporations billions of dollars a year. MDM attacks the enterprise data quality problem at its source on the operational side of the business. This is done in a coordinated fashion with the data warehousing / analytical side of the business. The combined approach is proving itself to be very successful in leading companies around the world. This paper will discuss what it means to ‘manage’ master data and outlines Oracle’s MDM solution2. Oracle’s technology components are ideal for building master data management systems, and Oracle’s pre-built MDM solutions for key master data objects such as Product, Customer, Supplier, Site, and Financial data can bring real business value in a fraction of the time it takes to build from scratch. Oracle’s MDM portfolio also includes tools that directly support data governance within the master data stores. What’s more, Oracle MDM utilizes Oracle’s Application Integration Architecture to create MDM Aware Applications3 and integrate the high quality authoritative master data into the IT landscape. This fusion of applications and technology creates a solution superior to other MDM offerings on the market. OVERVIEW How do you get from a thousand points of data entry to a single view of the business? This is the challenge that has faced companies for many years. Service Oriented Architecture (SOA) is helping to automate business processes across disparate applications, but the data fragmentation remains. Modern business analytics on top of terabyte sized data warehouses are producing ever more relevant and actionable information for decision makers, but the data sources remain fragmented and inconsistent. These data quality problems continue to impact operational efficiency and reporting accuracy. Master Data Management is the key. It fixes the data quality problem on the operational side of the business and augments and operationalizes the data warehouse on the analytical side of the business. In this paper, we will explore the central role of MDM as part of a complete information management solution. Master Data Management has two architectural components: The technology to profile, consolidate and synchronize the master data across the enterprise The applications to manage, cleanse, and enrich the structured and unstructured master data MDM must seamlessly integrate with modern Service Oriented Architectures in order to manage the master data across the many systems that are responsible for data entry, and bring the clean corporate master data to the applications and processes that run the business. MDM becomes the central source for accurate fully cross-referenced real time master data. It must seamlessly integrate with data warehouses, Enterprise Performance Management (EPM) applications, and all Business Intelligence (BI) systems, designed to bring the right information in the right form to the right person at the right time. In addition to supporting and augmenting SOA and BI systems, the MDM applications must support data governance. Data Governance is a business process for defining the data definitions, standards, access rights, quality rules. MDM executes these rules. MDM enables strong data controls across the enterprise. In order to successfully manage the master data, support corporate governance, and augment SOA and BI systems, the MDM applications must have the following characteristics: A flexible, extensible and open data model to hold the master data and all needed attributes (both structured and unstructured). In addition, the data model must be application neutral, yet support OLTP workloads and directly connected applications. A metadata management capability for items such as business entity matrixed relationships and hierarchies. Master Data Management 2 A source system management capability to fully cross-reference business objects and to satisfy seemingly conflicting data ownership requirements. A data quality function that can find and eliminate duplicate data while insuring correct data attribute survivorship. hA data quality interface to assist with preventing new errors from entering the system even when data entry is outside the MDM application itself. A continuing data cleansing function to keep the data up to date. An internal triggering mechanism to create and deploy change information to all connected systems. A comprehensive data security system to control and monitor data access, update rights, and maintain change history. A user interface to support casual users and data stewards. A data migration management capability to insure consistency as data moves across the real time enterprise. A business intelligence structure to support profiling, compliance, and business performance indicators. A single platform to manage all master data objects in order to prevent the proliferation of new silos of information on top of the existing fragmentation problem. An analytical foundation for directly analyzing master data. A highly available and scalable platform for mission critical data access under heavy mixed workloads. Master Data Management 3 This paper examines: the nature of master data; MDM’s central role in SOA and BI systems; the Oracle MDM Architecture; key MDM processes of profiling, consolidating, managing, synchronizing, and leveraging master data and how the Oracle MDM solution supports these processes; and Oracle’s portfolio of pre-built master data management solutions. Finally, this paper discusses build vs. buy tradeoffs given the power and flexibility in the Oracle MDM architecture and out- of-the-box capabilities of the pre-built and pre-connected MDM Hubs. ENTERPRISE DATA An enterprise has three kinds of actual business data: Transactional, Analytical, and Master. Transactional data supports the applications. Analytical data supports decision-making. Master data represents the business objects upon which transactions are done and the dimensions around which analysis is accomplished. Types of Data in the Enterprise Describes an Enterprise’s Operational State Describes an Enterprise’s Business Entities Describes an Enterprise’s Performance As data is moved and manipulated, information about where it came from, what changes it went through, etc. represents a fourth kind of enterprise data called metadata (data about the data). Though not a prime focus of this paper, the key role metadata plays in the broader information management space and how it relates directly to MDM is described. Transactional Data A company’s operations are supported by applications that automate key business processes. These include areas such as sales, service, order management, manufacturing, purchasing, billing, accounts receivable and accounts payable. These applications require significant amounts of data to function correctly. This includes data about the objects that are involved in transactions, as well as the transaction data itself. For example, when a customer buys a product, the transaction is managed by a sales application. The objects of the transaction are the Customer and the Product. The transactional data is the time, place, price, discount, payment methods, etc. used at the point of sale. The transactional data is stored in OnLine Transaction Processing (OLTP) tables that are designed to support high volume low latency access and update. Master Data Management 4 Operational MDM Solutions that focus on managing transactional data under operational applications are called Operational MDM. They rely heavily on integration technologies. They bring real value to the enterprise, but lack the ability to influence reporting and analytics. Analytical Data Analytical data is used to support a company’s decision making. Customer buying patterns are analyzed to identify churn, profitability, and marketing segmentation. Suppliers are categorized, based on performance characteristics over time, for better supply chain decisions. Product behavior is scrutinized over long periods to identify failure patterns. This data is stored in large Data Warehouses and possibly smaller data marts with table structures designed to support heavy aggregation, ad hoc queries, and data mining. Typically the data is stored in large fact tables surrounded by key dimensions such as customer, product, supplier, account, and location. Analytical MDM Solutions that focus on managing analytical master data are called Analytical MDM. They focus on providing high quality dimensions with their multiple simultaneous hierarchies to data warehousing and BI technologies. They also bring real value to the enterprise, but lack the ability to influence operational systems. Any data cleansing done inside an Analytical MDM solution is invisible to the transactional applications and transactional application knowledge is not available to the cleansing process. Because Analytical MDM systems can do nothing to improve the quality of the data under the heterogeneous application landscape, poor quality inconsistent domain data finds its way into the BI systems and drives less than optimum results for reporting and decision making. Master Data Master Data represents the business objects that are shared across more than one transactional application. This data represents the business objects around which the transactions are executed. This data also represents the key dimensions around which analytics are done. Master data creates a single version of the truth about these objects across the operational IT landscape. An MDM solution should to be able to manage all master data objects. These usually include Customer, Supplier, Site, Account, Asset, and Product. But other objects such as Invoices, Campaigns, or Service Requests can also cross applications and need consolidation, standardization, cleansing, and distribution. Different industries will have additional objects that are critical to the smooth functioning of the business. It is also important to note that since MDM supports transactional applications, it must support high volume transaction rates. Therefore, Master Data must reside in data models designed for OLTP environments. Operational Data Stores (ODS) do not fulfill this key architectural requirement. Master Data Management 5 Enterprise MDM Maximum business value comes from managing both transactional and analytical master data. These solutions are called Enterprise MDM. Operational data cleansing improves the operational efficiencies of the applications themselves and the business process that use these applications. The resultant dimensions for analytical analysis are true representations of how the business is actually running. What’s more, the insights realized through analytical processes are made available to the operational side of the business. Oracle provides the most comprehensive Enterprise MDM solution on the market today. The following sections will illustrate how this combination of operations and analytics is achieved. INFORMATION ARCHITECTURE The best way to understand the role that MDM plays in an enterprise is to understand the typical IT landscape. The best place to start is with the applications. Almost all companies have a heterogeneous set of applications. Some are home grown, others are bought from vendors, and still others are inherited during corporate mergers and acquisitions. Operational Applications The figure on the right illustrates the Heterogeneous Application Landscape typical heterogeneous operational application situation. Transactional Applications data exists in the applications local Sales data store. The data is designed Marketing specifically to support the features Service and transaction rates needed by the Inventory application. This is as it should be. Financials Partners But, in order to support business Supply Chain processes that cross these application Order Management boundaries, the data needs to be synchronized. The n2 Integration Problem Integration is an n2 problem in that Applications complexity grows geometrically with Sales the number of applications. Some Marketing companies have been known to call Service their data center connection diagram Inventory a “hair ball”. When synchronization Financials is accomplished with code, IT Partners projects can grind to a halt and the Supply Chain costs quickly become prohibitive. Order Management This problem literally drove the creation of Enterprise Application Integration (EAI) technology Master Data Management 6 Enterprise Application Integration (EAI) EAI uses a metadata driven approach to synchronizing the data across the operational applications. All information about what data needs to move, when it needs to move, what transformations to execute as it moves, what error recovery processes to use, etc. is stored in the metadata repository of the EAI tool. This information along with Enterprise Application Integration application connecters is used at run Applications time for needed synchronization. Hub and Spoke, Publish and Sales Subscribe, high volume, low latency Marketing content based routing are key Service features of an EAI solution. The Inventory figure on the right illustrates a set of Financials applications integrated via an EAI Partners technology. Depending on the Supply Chain topology deployed, these Order Management configurations are sometimes called EAI an enterprise service bus or integration hub. Service Oriented Architecture (SOA) In an SOA environment, the features and functions of the applications are exposed as shared services using standardized interfaces. These services can then be combined in end-to-end business processes by Service Oriented Architecture a technique called Business Process Orchestration Applications Orchestration. Sales In the figure on the left, we have Marketing added a business process Service orchestration layer to the Inventory architecture. This layer represents the tools used to design and deploy Financials business processes across Partners applications. Supply Chain OM The implication for MDM is that it EAI must not only support application to application (A2A) integration, it must also expose the master data to the business process orchestration layer as well. This is discussed in more depth in a later section. The Data Quality Problem EAI and SOA dramatically reduce the cost of integration but leave the data silos untouched. They are not designed to know what data ought to populate the various connected systems. They are designed to deal with the fragmentation, but Master Data Management 7 cannot eliminate it. All the data quality problems that existed in the pre-EAI/SOA environment still remain. Those problems continue to negatively impact business processes that cross these application boundaries. For example, an Order to Cash process may involve sales, inventory, order management and accounts receivable applications. While the data in each of the applications may be of sufficient quality to support the application, it may not be good enough to support the cross application business process. Product Ids and customer names may not be the same in each system. Which name or which Id is correct (if any). EAI and SOA are not designed to deal with these issues. A single view of the business remains elusive. MDM is the solution to this problem. In fact, Forrester4 considers MDM the most “strategic entry point for SOA” bringing the most strategic value to the business. But the anticipated improvements are never realized when data quality issues abound in the underlying applications. In fact, Gartner5 has pointed out that, without quality data, “…SOA will become a veritable ‘Pandora’s box’ of information chaos within the enterprise.” Oracle MDM insures that the potential value of SOA deployments is achieved. Analytical Systems Many companies turned to data warehousing to create a single view of the truth. Since the 1980s, data models have been deployed to relational databases with business intelligence features holding large amounts of historical data in schema designed for complex queries, heavy aggregation, and multiple table joins. This analytical space has three key components: 1. The Data Warehouse and subsidiary Data Marts 2. Tools to Extract data from the operational systems, Transform it for the data warehouse, and Load it into the data warehouse (ETL) 3. Business Intelligence tools to analyze the data in the data warehouse The following sections discuss these areas, and identify their MDM implications. Enterprise Data Warehousing (EDW) and Data Marts The Enterprise Data Warehouse (EDW) carries transaction history from operational applications including key dimensions such as Customer, Product, Asset, Supplier, and Location. Data Marts can be independent of the EDW, or connected to it and share common data definitions. Increasingly, the hybrid data warehouse (DW) has become common and consists of EDW-style third normal form schema and data mart star schema and OLAP cubes in the same database. Master Data Management 8 Data Warehousing Data Marts Orchestration Applications Data Warehouse Reporting Business Intelligence EAI ETL Extraction, Transformation, and Loading (ETL) ETL is a powerful metadata-based process that extracts data from source systems and loads data into a data warehouse. In the process, it performs transformations designed to improve overall data quality and reportability. The metadata maintains a history of the transforms and provides this information to business users through data lineage and impact analysis diagrams. Business Intelligence (BI) Business benefits gained by deploying a data warehouse solution are obtained through BI tools leveraged by the business user. The most widely accessed information is delivered in the form of reports. More sophisticated users who need to formulate their own questions and produce their own reports or spreadsheets use ad-hoc query tools. On-line analytical processing (OLAP) tools provide the ability to rapidly manipulate data containing a large number of table look-ups or dimensions and are particularly useful for performing trend analyses and forecasting. Where a very large number of variables are present and the goal is to determine an appropriate mathematical algorithm to determine likely outcomes, data mining tools can be leveraged. All of these BI tools can produce results that are viewed through dashboards or portals. The Data Quality Problem It is important to note that although data warehousing and business intelligence are an absolutely essential part of modern information technology and have brought great value to business decision-making and operational efficiencies, these solutions often did not create the desired ‘single view of the business’. Twenty five years after data warehousing’s inception, a recent survey found that a common top ten CIO request is to get a ‘single view of the customer’. One analyst reports “75% of leading companies are incapable of creating a unified view of their customer. 6” This is because, as we saw in earlier sections, the problem originates on the operational side of the business. An analytical solution cannot get to the root cause Master Data Management 9 of the problem. Fragmented dirty data being fed into the DW produces faulty reporting and misleading analytics. Garbage-in producing garbage-out still holds. What’s more, any data cleansing that is accomplished through ETL to the DW is invisible to the operational applications that also need the clean data for efficient business operations. In fact, all the segmentation and data mining results remain on the analytical side of the business. Key results such as customer profitability are not available via web services for real time high volume business processes. Ideal Information Architecture The ideal information architecture introduces the Master Data Management component between the operational and analytical sides of the business. The following figure illustrates this architecture. In this architecture, the master data is connected to all the transactional systems via the EAI technology. This insures that the clean master data is synchronized with the applications. A full cross-reference for every managed business object is created in the MDM system. This cross-reference is made available to the business process orchestration tool to insure the correct data objects are used as business processes cross applications boundaries. Ideal Information Architecture Data Marts Analytics Orchestration Applications DataMaster Warehouse Data Reporting Business Intelligence Master Data DW DW DW EAI ETL EDW Clean and accurate attribution for each master data business object is also maintained in the MDM system. The MDM system can supply these attributes back to connected systems and/or business processes. Ideally, appropriate master data attributes would be transferred to the applications to support real-time efficient business operations. But if (as is often the case with legacy applications) the quality attributes must remain federated in the MDM data store, then the business processes must retrieve the needed information from the MDM system at the lowest possible overhead. ETL is used to connect the Master Data to the Data Warehouse. The full cross- reference maintained in the MDM system is made available to the DW via the ETL tools. This enables accurate aggregation across the key business objects. In addition, the master data represents many of the major dimensions supported in Master Data Management 10 the data warehouse. Better quality dimension data improves the reporting out of the data warehouse. What’s more, the ETL tool can keep the MDM dimension tables in synch with the DW fact tables and support joins across these two domains. ETL is also used to populate master data attributes derived by the analytical processing in the data warehouse and by the assorted analytical tools. This information becomes immediately available to the connected applications and business processes. This is sometimes referred to as ‘Operationalizing’ the DW. This is the ideal information architecture. It unites the operational and analytical sides of the business. A true single view is possible. The data driving the reporting and decision making is actually the same data that is driving the operational applications. Derived information on the analytical side of the business is made available to the real time processes using the operational applications and business process orchestration tools that run the business. This is true information architecture. Oracle Information Architecture Oracle has state-of-the-art products and capabilities in each of the key information architecture domains. Oracle’s Multi-entity MDM Suite includes the applications. that consolidate, cleans, govern and share the master operational data: the Customer Hub, the Product Hub, the Supplier Hub, and the Site Hub. Each of these MDM hubs comes with a Data Steward component that provides a UI and Workbench for data professionals. The MDM Suite also includes state of the art Data Quality capabilities with: Oracle Data Quality servers for all party based structured data such as Name, Address, Customer, Supplier, Partner, Distributor, Regulator, Client, Doctor, Contact, Organization, Family, Constituent, etc.; and Product Data Quality for mastering unstructured item data such as Product, Catalog, Medical Procedure, Calling Plan, Asset, Parts, SKU, Service, etc. The MDM suite also includes the Data Relationship Management (DRM) application that consolidates, rationalizes, governs and shares master reference data such as Chart of Accounts and Cost Centers. DRM also manages enterprise hierarchies for Hyperion Enterprise Performance Management and other BI tools. All the Oracle MDM products offer Data Governance capabilities that provide business uses with the controls they need to enforce corporate standards. Oracle Data Watch and Repair is also included for overall data profiling and data governance. With the Oracle 11g Database, Oracle Data Warehousing is number one in the industry. Oracle Data Integrator Enterprise Edition (ODI EE) offers the best in class, high performance ETL architecture. These products understand MDM dimension, hierarchy and cross-reference files and can use them to correctly connect the detailed data elements flowing into the warehouse from transactional systems. Master Data Management 11 Oracle Information Architecture Orchestration Applications Master Data Analytics MDM Applications BI EBE,usiEnPesMs, InDtealtlaigMenincieng,. = DRM = BI P ETL = ODI EE EAI = FMW Exadata w/ RAC A tremendous amount of valuable information can accumulate in the master data store. Directly leveraging this data can provide critical business insights. Oracle provides the most complete portfolio of BI applications and technologies in the industry. Oracle MDM customers can utilize Oracle Business Intelligence Enterprise Edition Plus (OBI EE) with its ad-hoc query, highly interactive dashboards, business activity monitoring, and BI Publisher (BI P) to query, monitor and report on the master data. Oracle Data Mining is extremely valuable and includes a feature called Anomaly Detection. The goal of anomaly detection is to identify cases that are unusual within the data. This is an important tool for detecting fraud and other rare events that may have great significance but are hard to find. When used against central stores of master data in an Oracle MDM Data Hub, unusual activity can be identified and remediated if necessary. All these BI applications have direct access to the master data tables within the MDM Data Hubs. Metadata is managed by Oracle’s OWB Metadata Manager. Unstructured data is included via Oracle Universal Content Manager and the Content DB. Secure searching across MDM, DW, Metadata, and Universal Content Manager data volumes is provided with Oracle Secure Enterprise Search. With all the master data supporting business processes in one place, high availability becomes a must. Single points of failure are not acceptable. High performance from a transactional point of view is also essential. A system that is too slow to provide the needed data in real time is as bad as a system that is down. Real Application Clusters provide high availability, scalability, and mixed workload support for the MDM OLTP and Decision Support workloads on one Single Global Instance with all its associated cost savings. With the Sun Oracle Database Machine with Exadata, Oracle provides the highest performance lowest Total Cost of Ownership (TCO) hardware platform perfect for running both the entire MDM platform and the Data Warehouse – two databases on one clustered instance. Oracle’s Fusion Middleware Suite (FMW) provides the EAI and SOA capabilities. FMW includes Oracle Service Bus a full function high performance enterprise service bus. FMW also includes the Oracle Business Process Execution Language Master Data Management 12 Process Manager (BPEL PM). This is the best business process orchestration product on the market. Both Enterprise Service Bus and BPEL PM understand the Oracle MDM data structures and access methods. Additional components include: Oracle Business Rules engine that can be coordinated with the data quality rules engines inside the MDM applications; an Event Driven Architecture for real time complex event processes so essential for triggering appropriate MDM related business processes when key data items are changed; Web Services Manager to manage and secure the SOA web services, including the web services exposed by the MDM applications; Web Center and Oracle B2B for individual and partner participation; XSLT & Xquery Translation for open standards based data transformations as master data flows between applications and the MDM data store; and Oracle Identity Management (OIM) for full user identification, authorization, and provisioning. OIM helps insure full Role Based Access Controls (RBAC) for the MDM applications7 and their primary data governance functions. The following figure illustrates how all these components are connected at the Enterprise Service Bus level. Each and every one of these plays a role in Oracle’s MDM architecture. Once we outline the key MDM processes that any master data solution must support, we will overview this architecture and discuss its key components. We will then focus on the supporting capabilities in the Oracle MDM Hub applications themselves. MASTER DATA MANAGEMENT PROCESSES Now that we have identified the nature of master data and its place in an information architecture, we need to identify the key processes that MDM solutions must support. Master Data Management 13 These are the key processes for any MDM system. Profile the master data. Understand all possible sources and the current state of data quality in each source. Consolidate the master data into a central repository and link it to all participating applications. Govern the master data. Clean it up, deduplicate it, and enrich it with information from 3rd party systems. Manage it according to business rules. Share it. Synchronize the central master data with enterprise business processes and the connected applications. Insure that data stays in sync across the IT landscape. Leverage the fact that a single version of the truth exists for all master data objects by supporting business intelligence systems and reporting. Profile The first step in any MDM implementation is to profile the data. This means that for each master data business entity to be managed centrally in a master data repository, all existing systems that create or update the master data must be assessed as to their data quality. Deviations from a desired data quality goal must be analyzed. Examples include: the completeness of the data; the distribution of occurrence of values; the acceptable rang of values; etc. Once implemented, the MDM solution will provide the ongoing data quality assurance, however, a thorough understanding of overall data quality in each contributing source system before deploying MDM will focus resources and efforts on the highest value data quality issues in the subsequent steps of the MDM implementation. The Data Profiling and Correction option in Oracle Data Integration Suite (ODI) provides a systematic analysis of data sources chosen by the user for the purpose of gaining an understanding of and confidence in the data. It is a critical first step in the data integration process to ensure that the best possible set of baseline data quality rules are included in the initial MDM Hub. Master Data Management 14 Consolidate Consolidation is the key to managing master data. Without consolidating all the master data attributes, key management capabilities such as the creation of blended records from multiple trusted sources is not possible. This is the #1 fundamental prerequisite to true master data consolidation8. Oracle MDM Hubs utilize state-of-the-art extensible data models. They are operational data models designed for OnLine Transaction Processing (OLTP). They are application neutral and capable of housings all corporate master data from all systems in all heterogeneous IT environments. This includes business objects such as Customer, Supplier, Distributor, Partner, Site, Product, Assets, Installed Base and more. The models support all the master data that drives a business, no matter what systems source the master data fragments. This includes (but is not limited to) SAP, Siebel, JD Edwards, PeopleSoft, Oracle E-Business Suite, Microsoft, Acxiom, Dun & Bradstreet (D&B), billing systems, homegrown systems, and legacy systems. Tools to load the data are also provided. Scalable batch load tools manage the history and mappings from source systems. Oracle provides powerful Data Quality Servers to standardize, cleanse, and match master data attributes during the MDM Data Hub load process. This insures that the loaded data is clean and duplicates are eliminated on the way in. Govern Master data is consolidated so that it can be cleansed and governed. Specific business objects require specific management tools. Managing unstructured product data is very different than managing structured customer data. This is why Oracle provides Data Quality servers for customer/party data and product/item data that are easily extended to suppliers, materials and assets. Data Governance refers to the operating discipline for managing data and information as a key enterprise asset. This operating discipline includes organization, processes and tools for establishing and exercising decision rights regarding valuation and management of data. Data governance is essential to ensuring that data is accurate, appropriately shared, and protected. Oracle MDM applications provide significant data quality and data governance capabilities. Share Clean augmented quality master data in its own silo does not bring the potential advantages to the organization. For MDM to be most effective, a modern SOA layer is needed to propagate the master data to the applications and expose the master data to the business processes. SOA and MDM need each other if the full potential of their respective capabilities are to be realized. Oracle MDM maintains an integration repository to facilitate web service discovery, utilization, and management. What’s more, Oracle MDM Hubs leverage Oracle Application Integration Architecture (AIA) to provide pre-built comprehensive application integration with the MDM data and MDM data quality services. AIA delivers a composite application framework utilizing Foundation Packs and Process Master Data Management 15 Integration Packs (PIPs) to support real time synchronous and asynchronous events that are leveraged to maintain quality master data across the enterprise. We will cover this significant differentiating capability for Oracle’s MDM in more depth in later sections. Leverage MDM creates a single version of the truth about every master data entity. This data feeds all operational and analytical systems across the enterprise. But more than this, key insights can be gleamed from the master data store itself. 360o views can be made available for the first time since operational and analytical systems split in the 1980s. Alternate hierarchies and what-if analysis can be performed directly on the master data. Oracle MDM Hubs leverage Oracle BI tools such as OBI EE and BI Publisher to produce 360o views and cross-reference data to the Data Warehouse and maintain master dimensions in the master data store. Data quality and segmentation can be viewed directly from the master data repository. Out-of-the-box reports are provided and BI Publisher has full access to all master data attributes within constrains set up by the data security rules. Oracle’s Analytical MDM can create an enterprise view of analytical dimensions, reporting structures, performance measures and their related attributes and hierarchies using Hyperion DRM's10 data model-agnostic foundation. ORACLE MDM HIGH LEVEL ARCHITECTURE The following figure identifies the major layers in the Oracle MDM architecture. Oracle Fusion Middleware provides supporting infrastructure. The MDM Applications layer contains all the base pre-built MDM Hubs and shared services. The top layer includes MDM based solutions for Data Governance Master Data Management 16 MDM Platform Layer There are a number of Fusion Middleware (FMW) technologies used to support MDM Applications. These include application integration services, business process orchestration services, data quality & standardization services, data integration and metadata management services, business rules engine, event-driven architecture, web services management, user identity management, and analytic services. The following sections identify the FMW components that support these services and how they are leveraged by MDM Hubs. Application Integration Services Application integration services are provided by Oracle’s award winning Fusion Middleware. The following sections cover the key components used by the Oracle MDM suite of applications. MDM Hubs. What’s more, MDM Hubs, configured as a ‘System of Record’, dramatically reduce the complexity associated with A2A integration. When all applications are in sync with the MDM Hub, they are in sync with each other. This effectively turns the n2 harmonization of fragmented data problem into a linier management of consolidated data problem. Architecturally speaking, this is a very significant improvement and leads to many cost of ownership advantages. Business Process Orchestration Services The Business Process Execution Language Process Manager (BPEL PM) provides business process orchestration services. BPEL is the standard for assembling sets of discrete services into an end-to-end process flow, radically reducing the cost and complexity of process integration initiatives. Master Data Management 17 Built-in integration services enable developers to easily leverage advanced workflow, connectivity, and transformation capabilities from standard BPEL processes. These capabilities include support for XSLT and Xquery transformation as well as bindings to hundreds of legacy systems through JCA adapters and native protocols. The extensible WSDL binding framework enables connectivity to protocols and message formats other than SOAP. Bindings are available for JMS, email, JCA, HTTP GET, POST, and many other protocols enabling simple connectivity to hundreds of back-end systems13. BPEL PM has deep knowledge of Oracle MDM Data Hub structures and access methods. It comes with hot pluggable components for bringing the quality information in the MDM Hubs to all business processes and applications across the enterprise and beyond with business-to-business (B2B) integration support for EDI, HL7, Rosetta Net, UCCNet, GDSN, 1Sync and more. Business Rules The Oracle Business Rules is a business rules engine for making decisions about aggregating federated information in real time; resolving sourcing system conflicts; and coordinating data visibility with the data quality rules incorporated into MDM Applications. Privacy policies, regulatory policies, corporate policies, and data policies can be enforced. Policies can be as simple as: Every patient can read their own records, or as complex as: A “gold” customer has total assets under management > $100k and average monthly transaction volume > $10k and is in good standing. Oracle Business Rules allow business analysts to manage rules by defining and maintaining those rules in a separate repository, with an intuitive web-based interface. It can be executed from within an application via Java code, the Oracle Business Rules API, or a web services interface. This integration is especially attractive for MDM Hubs. Flexible System of Entry Master Datta Hub Mastterr Datta Hub Entry into the System of Record Entry into a Connected System Coordinating the data quality rules configured in the data quality engine inside MDM Hubs with the rules configured in the integration layer allows the Oracle solution to manage master data updates from spoke systems and the MDM application. This is a key feature. Many master data management systems require that the MDM application be the only System of Entry. While having one System of Entry may simplify maters, most companies cannot shut down data entry in major Master Data Management 18 systems. The Oracle MDM solution, when deployed with OSB and BPEL PM with its Business Rules based policy engine can support multiple Systems of Entry. What’s more, the integration layer in conjunction with the MDM Hubs can enforce central data quality rules. Event-Driven Services Oracle Event-Driven Architecture (EDA) is comprised of best-in-class technologies for event-enabling a business. It allows applications to register events, associate payload packages and trigger designated business processes and workflows. All changes to master data in the Oracle MDM Hubs can be exposed to the EDA engine. This enables the real time enterprise and keeps master data in sync with all constituents across the IT landscape. Human workflow services such as notification management are provided as built-in BPEL and OSB services that enable the integration of people and manual tasks into business processes. Payload packages identified by EDA accompany the workflow messages. Oracle MDM can automatically trigger EDA events and deliver predefined XML payload packages appropriate for the event. These packages are easily modified. Identity Management Oracle Identity Management (OIM) is an integrated system of business processes, policies and technologies that enable organizations to facilitate and control their users' access to critical online applications and resources — while protecting confidential personal and business information from unauthorized users15. It supports user authentication, authorization and provisioning. OIM uses MDM as source of truth for external identities (e.g. non-staff); insures that accounts for external Create Person identities are properly enabled/disabled according to organization business rules; MDM Start event- and provides attestation reports to driven OIM user account provisioning identify and disable rogue accounts. Oracle MDM Hubs utilize Oracle’s full function Identify Management solution. OIM enables the Role Based Access Controls (RBAC) so vital for controlling LDAP Directory EMAIL Server Application Create, Read, Update, and Delete (CRUD) access to the master data. Web Services Management Oracle Web Services Manager (WSM), a component of Oracle’s SOA Suite, is a comprehensive solution for managing service oriented architectures. It allows IT Master Data Management 19 managers to centrally define policies that govern web services operations such as access policy, logging policy, and content validation, and then wrap these policies around services with no modification to existing web services required. MDM Hub services are exposed as web services that are controlled and secured via WSM. Analytic Services The Analytics Server is key to leveraging the master data to gain insights into the business. This includes improved real time decision making and quality reporting. Corporate governance is enhanced and financial risk is reduced. Key components of the Analytics Server include Enterprise Performance Management, Data Warehousing, ETL, and Business Intelligence tools. The following sections identify the Oracle Analytic Server products used by and/or with MDM Applications. Enterprise Performance Management Oracle Hyperion Enterprise Performance Management (EPM) system brings management processes under a single umbrella, connecting financial and operational decisions and activities with transactional systems to form a comprehensive management picture. The MDM Data Hubs provide data to Oracle’s EPM applications. Oracle’s Enterprise Planning and Budgeting application, can leverage the reliable, accurate master data residing in an MDM Hub-fed data warehouse to perform its complex aggregations and produce actionable ‘what if’ plans around customers, products, accounts, etc. insuring reporting accuracy. What’s more, Oracle’s MDM solution for reference data and enterprise hierarchy management, Data Relationship Management (DRM), is fully integrated into Oracle Hyperion EPM. More detailed information on this Analytical MDM component of the Oracle MDM suite is included in the MDM Applications Layer chapter later in this document. Data Warehousing MDM supplies critically important dimensions, hierarchies and cross reference information to the Oracle Data Warehouse. With all the master data in the MDM Hubs, only one pipe is needed into the data warehouse for key dimensions such as Customer, Supplier, Account, Site, and Product. The “Star Schema” illustrated on the left from the Oracle whitepaper on Data Warehousing Best Practices16 shows all dimensions except Time are managed by Oracle MDM. Not only that, Oracle MDM maintains master cross-references for every master business object and every attached system. This cross-reference is available regardless of data warehousing deployment style. For example, an enterprise data warehouses or data mart can seamlessly combine historic data retrieved from those operational systems and placed into a Fact table, with real-time dimensions form the MDM Hub. This way the business intelligence derived from the warehouse is based on the same data that is used to run the business on the operational side. Master Data Management 20 Business Intelligence Oracle Business Intelligence Enterprise Edition (OBI EE) provides ad-hoc query and real time dashboard capabilities. Oracle BI EE is designed to support heterogeneous data sources. With this capability and the data consistency an MDM solution provides, it is possible to view the master data, the data residing in transaction tables, and the data warehouse. The following figure illustrates the role MDM plays in creating accurate data from pre-built ETL Mappings. BI EE Dashboarding CCoommmmoonnSScchheemmaa Pre-built ETL Mappings DDaattaaHHuubbss Transaction Master Data Operational Summaries Dimensional Summaries Tables Application knowledge is critical. These types of high value analytical applications require a tight linkage to the transactional applications in order to provide the out- of-the-box mappings. Oracle MDM Hubs play a central role in rationalizing master data across the heterogeneous application landscape and therefore play a central role in these types of analytics. In fact, Oracle delivers thousands of Oracle Data Integrator attribute maps for dozens of dimensions out-of-the-box. Publishing Services Oracle BI Publisher provides the ability to rapidly create high fidelity reports and forms using common desktop authoring tools. BI Publisher is included in the Oracle BI EE Suite and can be used to author reports around data produced in queries submitted by Answers. For example, one might use BI Publisher to build reports that include master data, transaction data, and warehouse data. Oracle MDM uses BI Publisher with out-of-the-box report generation capabilities. Data Migration Services Data Migration Services include Data Integration and Metadata Management. Data Integration is essential for MDM in that it loads, updates and helps automate the creation of MDM data hubs. MDM also utilizes data integration for integration quality processes and metadata management processes to help with data lineage and relationship management. The Oracle Data Integration Suite (ODI) provides a fully unified solution for building, deploying, and managing complex data warehouses. In addition, it combines all the elements of data integration—data movement, data synchronization, data quality, data management, and data services—to ensure that information is timely, accurate, and consistent across complex systems. What’s more, Oracle Data Integrator Enterprise Edition delivers unique next-generation, Extract Load and Transform (E-LT) technology that improves performance, reduces data integration costs, even across heterogeneous systems17. This makes Master Data Management 21 ODI the perfect choice for 1) loading MDM Data Hubs, 2) populating data warehouses with MDM generated dimensions, cross-reference tables, and hierarchy information, and 3) moving key analytical results from the data warehouse to the MDM Data Hubs to ‘operationalize’ the information. High Availability and Scalability With all the master data in one place, High Availability becomes a requirement. Just as importantly, an organization cannot deploy an MDM solution that won’t keep up with growing data volumes, users, application suites and business processes. Real Application Clusters Real Application Clusters (RAC) provides the needed high availability and scalability. RAC is an option to Oracle Database 11g Enterprise Edition. It supports the deployment of a single database across a cluster of servers—providing unbeatable fault tolerance, performance and scalability with no application changes necessary. As your system grows, nodes can be added without downtime. Oracle MDM applications run seamlessly on RAC clusters. Mixed Workloads Oracle’s ability to run mixed workloads across RAC nodes is unique in the industry and adds another significant dimension to Oracle’s MDM solution. OLTP workloads are routed to a subset of clustered servers accessing the MDM tables while Data Warehousing workloads are routed to other nodes in the same cluster. With world class transaction processing and record data warehousing performance, on a single instance becomes possible. The data warehousing tables and MDM tables reside in an environment all managed through a single Oracle Enterprise Manager console. This lower cost, more easily managed environment provides dramatic savings and puts a company on the path to a more rational IT infrastructure. Exadata With the acquisition of Sun Microsystems, Oracle has become a Software- Hardware-Complete vendor. No software-hardware application illustrates the benefits of this combination better than Master Data Management running on the Exadata database computer20. The OLTP MDM tables can coexist with the Data Warehouse star schemas moving data around a disk array, creating summarized tables and utilizing materialized views without offloading terabytes of data through costly networks to multiple hardware platforms. Availability and scalability are maximized even as the total cost of ownership is reduced. Only Oracle offers a full suite of business applications, state of the art middleware, a world class clustered database, and a low cost high performance hardware platform. MDM takes full advantage of this combination to the significant benefit of our customers. Master Data Management 22 Application Integration Architecture Application Integration Architecture (AIA) utilizes Oracle’s premier SOA suite to build out-of-the-box Oracle Application integrations in the context of enterprise business processes. These integrations directly support MDM. There are multiple levels of integration between MDM and SOA. Connectors and transformations Mutually understood data structures and access methods Pre-Built Application and Master Data synchronization Pre-Built SOA/MDM Enterprise Business Processes The business value goes up the more levels a vendor provides. All MDM vendors provide some level of connectors and templates for transformations. Only vendors that provide both MDM and SOA can integrate the two. Most of these vendors provide their SOA with knowledge of their MDM data structures and access methods. But only vendors who actually have applications can provide pre-built master data synchronizations and pre-built enterprise business processes. Oracle has integrated its market leading MDM suite of applications with its powerful Fusion Middleware driven SOA suite. The following section describes how Oracle brings these two together at all three integration levels listed above22. AIA Layers AIA delivers the following components deployed at the various levels of the SOA stack. 1) Industry reference models cover the best practice business processes for key industries. 2) This layer is supported by Enterprise Business Objects (EBOs). 3) These objects are orchestrated into process & task flows with data transformations. 4) Flows utilize Enterprise Web Services. 5) These services are provided by the underlying Oracle Applications and MDM Hubs. Oracle is utilizing its MDM and SOA capabilities, along with the AIA EBOs and associated structures, to pre-build application integrations in the context of key industry specific business processes. The EBOs are deployed as part of a common object methodology that uses the EBOs to define all transformation from target and source applications included in the business process. Master Data Management 23 Common Object Methodology The following figure illustrates how the common object methodology is used to integrate the Oracle Applications. As data flows from source systems, it is transformed via provided maps into a common object model. Once the appropriate business logic is executed for a particular business process, the data is again transformed via provided maps from the common object model to the format needed by target systems. Oracle has leveraged its ability to fuse applications and technology to incorporate MDM applications as a foundation component for its SOA Suite. Delivering more than just connectors and templates, Oracle’s MDM-SOA combination delivers out- of-the-box, fully tested and extensible, pre-built SOA enterprise business processes that synchronize MDM data stores with applications. Deploying the MDM Hubs and connecting them to the common object model in the AIA integrations, provides the consolidated, cleansed, deduplicated, augmented authoritative ‘Golden Record’ to every application and business process in the delivered pre-built SOA integration. MDM Foundation Packs AIA delivers its base integration capabilities in Foundation Packs. Master Data Management processes such as ‘Publishing’ master data to multiple subscribers is designed in. For example, customer information can be created, updated, and deleted in multiple participating applications and sent to the Oracle Customer Hub for updating, cleaning, and validation. Subsequently, the Customer Hub, as the single source of truth, publishes the customer information for multiple subscribing participating applications to consume23. Foundation Packs are available for all application integration scenarios. MDM Process Integration Packs AIA delivers pre-built integrations as Process Integrations Packs (PIPs). These packs are pre-built composite business processes across enterprise applications. They allow organizations to get up and running with core processes quickly. When delivered with MDM, these complete, out-of-the-box, integrations Master Data Management 24 include everything needed to gain immediate business value and increase business and IT efficiencies. Key MDM PIPs include: Process Integration Pack for Oracle Customer Hub is a collection of core processes to support out-of-the-box Customer MDM integration processes across Oracle Customer Hub, Siebel CRM and Oracle E- Business Suite, as well as a framework to enable MDM integrations with other Oracle and non-Oracle applications24. Process Integration Pack for Oracle Product Hub is a collection of core processes to support out-of-the-box Product MDM integration processes across Oracle Product Hub, Siebel CRM and Oracle E-Business Suite25. MDM Aware Applications When an application understands that the key data elements within its domain have business value beyond its borders, we say it is “MDM Aware”. An MDM Aware application is prepared to: Use outside data quality processes for data entry verification Pull key data elements and attributes from an outside master data source Push its own data to external master data management systems In short, MDM Aware applications are pre-disposed to participating in the MDM data quality process. This dramatically speeds the deployment of MDM solutions, reduces risk and insures quality data is distributed as needed across the IT landscape. Oracle Applications are MDM Aware. Recent combined MDM and AIA releases enable non-Oracle applications to become MDM Aware. A composite application user interface is available that enables non-Oracle applications such as legacy and web applications to also become MDM Aware26. Composite Application Development With AIA Foundation Packs, MDM PIPs and MDM Aware applications, organizations can more readily create new business processes by combining relevant elements of existing applications. This is called Composite Application Development. This was the SOA vision. That vision has been slow to realize due to the large number of complex integration components, lack of built in governance, and continuing data quality problems in the underlying applications. The power of the AIA out-of-the-box pre-built SOA with open, extensible and governable components together with the power of the pre-cabled Oracle MDM solution to ensure quality data across the applications and composite applications eliminates these SOA roadblocks. The SOA promise of IT flexibility is finely realized. Oracle Data Quality Services Data cleansing is at the heart of Oracle MDM’s ability to turn data into an enterprise asset. Only standardized, de-duplicated, accurate, timely, and complete data can effectively serve an organization’s applications, business processes, and analytical systems. From point-of-entry anywhere across a heterogeneous IT Master Data Management 25 landscape to end usage in a transactional application or a key business report, Oracle MDM’s Data Quality tools provide the fixes and controls that ensure maximum data quality. Oracle recognizes that there are two primary data categories: relatively structured party data and relatively unstructured item data. Party data includes people and company names such as customers, suppliers, partners, organizations, contacts, etc., as well as address, hierarchies and other attributes that describe who a party is. The following figure illustrates this kind of data and the problems most frequently encountered. This is relatively structured error-prone data that is subject to country specific rules. It must be cleansed in order to find matches. Pattern matching tools are best for cleansing this kind of data. Item data includes Products, Services, Materials, Assets, and the full range of attributes that describe what an item is. This is relatively unstructured data that is subject to category specific rules. It is common for widely different descriptions to actually be identifying the same item. Semantic matching tools are required for cleansing this kind of data. Any vendor that does not have this kind of semantic technology cannot satisfactorily master product data. Master Data Management 26 These differences between party and item data are fundamental to the data itself. This is why Oracle provides data quality tools specifically designed to handle these two kinds of data. One is our suite of Customer Data Quality servers and the other is our state-of-the-art Product Data Quality. The following sections discuss these two product areas in more depth. Master Data Management 27 duplicates and establish relationships in real time; and build relationship link tables and match external files and databases. Oracle Data Quality Profiling Server When data quality is measured, it can be effectively managed. Data profiling is the first step for data quality. Oracle Data Quality Profiling Server gives users with the ability to measure the level and nature of data quality problems across multiple data sources—to identify, quantify, and categorize current and potential data quality issues and adherence to business rules. The product also provides the metrics and reports that business information owners need to continuously measure, monitor, track, and improve data quality at multiple points across the organization. Master Data Management 28 international customer information, preventing against identity loss, and assisting in fraud detection and prevention, as well as directly reducing mail based marketing costs. The server performs validation of address data against published standards and directories by utilizing reference data. As postal codes change in different countries due to settlements, system changes or name changes, reference tables for each country can also be updated on a monthly, quarterly or biannual basis. Oracle leverages arrangements with leading postal reference data provider to enable customers to receive the updates on a periodical basis. These reference tables are provided in a separate, platform-independent database making it easily updateable at any time. Oracle Customer Data Quality servers enable business information owners and IT to work together to deploy lasting customer data quality programs. Business information owners use Oracle Data Quality to build data quality business rules and define data quality targets together with the IT team, which then manages deployment enterprise-wide. Product Data Quality Typical product data is unstructured, non-standard and often missing important information. These product data errors slow procurement, development, distribution and negatively impact service. Oracle Product Data Quality (PDQ) with patented Data Lens™ semantic-based technology is designed to solve this problem. PDQ is built from the ground up to tackle the unique challenges of assessing, and improving the ongoing management of product data. Oracle Product Data Quality has been proven in a wide range of customer scenarios involving Product, Item, Catalog, Medical Procedure, Calling Plan, Asset, Parts, SKU, Service, and other forms of product or product-like data across a range of industries. Oracle Product Data Quality is designed to handle data that is: Poorly structured—requires sophisticated semantic parsing capabilities Non-standard—required standards to be applied and data transformed to meet enterprise standards Highly variable—practically infinite combinations of acronyms, spelling and vocabulary variations and other cryptic information requires flexible recognition and transformation capabilities Category-specific—requires the ability to both categorize an item and apply different rules based on content and context Variable quality—requires integrated exception management and remediation capabilities Duplicated—requires sophisticated semantic matching capabilities Oracle Product Data Quality delivers: Semantic-based recognition—based on context to enable accurate parsing, standardization and matching along with auto-learning to handle the extreme variability and unpredictability of product data Master Data Management 29 user who best understand the rules and nuances of the data Enterprise-wide Applicability—a standard process that can 'plug-in' to existing systems and processes to enforce product data quality standards in any process or system PDQ represents a breakthrough in product data solutions with next generation semantic technology that automates what has traditionally been a costly, labor- intensive and largely unreliable process28. Product Data Quality provides an integrated capability to recognize, cleanse, match, govern, validate, correct and repurpose product data from any source. It can standardize product classifications, attributes, and descriptions and translate languages as data flows into the Product Hub tables. All input items are compared against the semantic model for contextual recognition & validation. Semantic recognition uses whole context to determine meaning and category. Semantic recognition identifies item Scalability—to manage category and key attributes even with millions of items across highly variable data. The system thousands of categories infers meaning from unrecognized Integrated items and asks for confirmation from Governanc the user. Category-specific semantic e—to allow models flag missing information for data potential remediation. If the user stewards to confirms, new rules can be created to monitor extend semantic model. Data is overall converted, transformed and re- process assembled into the Product Hub. effectivenes s as well as Product Data Quality can standardization any attributes in any form. It can make drive direct imperial/metric conversions and handle multiple classifications. Its auto-translation data capabilities can translate millions of rows in seconds with full double-byte support. remediation Exception management for validation and remediation is a key component of the solution and quality is governed via a supplied Data Steward dashboard. The Business User dashboard provides quality metrics by process, source, and product category and Interface— management metrics such as task management overviews that measure productivity designed using a and identify bottlenecks. code-free interface for use A Data Governance Studio dashboard provides visibility to enterprise data and by the business metrics to drive process improvements. The key to accomplishing these data quality Master Data Management 30 goals is the solution’s ability to learn through semantic recognition that gains contextual knowledge over time. End-To-End Data Quality Data quality tools are applied against data as it is held in databases, as it is moved, and as it is entered. The value of the data quality goes up as the number of places it can be brought to bear increases. The left hand side of the following picture illustrates data quality improvements being applied as the data as it flows from applications into the MDM Data Hub. The data integration technology is ETL. Most data quality tools can clean up the data in the applications and in databases such as a data warehouse or an MDM Hub. Oracle’s MDM DQ tools can do the same. This is exactly what Oracle has been able to do because of its fusion of MDM and SOA. The right hand side of the above picture illustrates how Oracle, using AIA PIPs, can trap data entry in real time and perform key DQ functions such as Fetch, match, Enhance, and Synchronize for MDM Aware Applications. This is the final Data Quality answer. It fixes the DQ problem at its source. But DQ technology that stops there is still letting poor quality data enter the IT landscape. Potentially thousands of people are entering customer data into a wide variety of operational applications. In fact, data errors are entering the system at a significant rate. Data quality tools that clean things up after the fact and try undo any damage caused by poor quality data are certainly performing a beneficial service. But a DQ technology that can prevent the data entry errors in the first place would be far superior. Master Data Management 31 Application Development Environment JDeveloper serves as the development environment for new application creation, existing application extensions and composite application development. Web Service Invocation Framework (WSIF), and Java Business Integration (JBI) are fully supported. Application functions are designed and built to be deployed as web services. This is the ideal tool for creating new and composite applications around the Oracle MDM Hubs. JDeveloper has full access to the Application Development MDM Hub APIs and Web Services. Directly on the Oracle MDM Data Hubs Oracle MDM Hubs can actually support applications directly. This Integ rate gives Oracle’s master data solution a Develop Orchestrate key differentiation over all other MMMaaassstteteerrr Secure approaches. When new applications DDDaaattataa Hub are written to replace older apps, or Oracle applications are deployed in Change Manage place of existing apps, physical silos Monitor are removed from the IT landscape. This is key to IT infrastructure simplification and is why we call Oracle MDM a first step on the ‘path to a suite’. MDM APPLICATIONS LAYER Oracle MDM includes a large portfolio of purpose built master data management applications. The MDM Applications include all MDM Hubs and their corresponding data quality servers. Data Governance is also included. The following sections will cover: Oracle Customer Hub o Oracle Higher Education Constituent Hub Oracle Product Hub o Oracle Product Hub for Retail o Oracle Product Hub for Comms Oracle Supplier Hub Oracle Site Hub Oracle Data Relationship Management No other vendor on the market has this breath of master data element coverage. MDM Pillars The MDM Applications are organized around five key pillars. The following figure illustrates these pillars of every MDM Application. Master Data Management 32 Trusted Master Data is held in a central MDM schema. Consolidation services manage the movement of master data into the central store. Cleansing services de- duplicate, standardize and augment the master data. Governance services control access, retrieval, privacy, auditing, and change management rules. Sharing services include integration, web services, ETL maps, event propagation, and global standards based synchronization. These pillars utilize generic services from the MDM Foundation layer and extend them with business entity specific services and vertical extensions as described in the MDM High Level Architecture covered in an earlier section of this paper. The following sections describe Oracle’s MDM Applications for customer, product, supplier, site, and financial master data as well as a section on the very important role played by data governance. Oracle Customer Hub Oracle Customer Hub29 (OCH) is Oracle’s lead Customer Data Integration (CDI) solution. OCH’s comprehensive functionality enables an enterprise to manage customer data over the full customer lifecycle: capturing customer data, standardization and correction of names and addresses; identification and merging of duplicate records; enrichment of the customer profile; enforcement of compliance and risk policies; and the distribution of the “single source of truth” best version customer profile to operational systems. Oracle Customer Hub is a source of clean customer data for the enterprise. The primary roles are: Consolidate & govern unique, complete and accurate master Customer information across the enterprise. Distributes this information as a single point of truth to all operational & analytical applications just in time. Master Data Management 33 To accomplish this, OCH is organized around the five MDM Pillars: Trusted Customer Data is held in a central MDM schema. Consolidation services manage the movement of master data into the central store. Cleansing services de-duplication, standardize and augment the master data. Governance services control access, retrieval, privacy, audit and change management rules. Sharing services include integration, web services, event propagation, and global standards based synchronization. These pillars utilize generic services from the MDM Foundation layer, and extend them with business entity specific services and vertical extensions. Customer Data Model The OCH data model is an extensible proven transaction processing schema that has evolved and developed over many years to the point where it is able to master not only the customer profile attributes required in the front office but also those required by all applications and systems in the enterprise. A few of its key characteristics include: Roles and Relationships – The customer model provides support for managing the roles and hierarchical relationships not only within a master entity but also across master entities. Related and Child Data Entities and their Configuration – The customer model is designed to store all the related and child level customer data entities such as Addresses, Related Organizations, Related Persons, Assets, Financial Accounts, Notes, Campaigns, Partners, Affiliations, Privacy and so on. Industry Variants - The prebuilt customer data model is designed to model and store the customer profile attributes for many major vertical markets and industries, including (but not limited to); Financial Services, Master Data Management 34 Workbench- The List import workbench empowers business users to load data into the hub through metadata and template driven approaches including support for flat files, XML files and excel files. The import workbench also includes user interfaces to resolve error conditions. In addition to providing a high performance, high volume batch import, the list import functionality is also available as a web service for real time integration. Identification and Cross-reference - To capture the best version customer profile, OCH provides a prebuilt and extensible customer lifecycle management process. This process manages the steps necessary to build the trusted “best Consolidate version” customer profile: identification; registration; cleansing; matching; Customer data is enrichment; and linking. The best version customer record also leverages Universal distributed across the Unique ID (UUID) to uniquely identify the record across the entire enterprise. enterprise. It is typically The customer record is also tracked across the enterprise by using one-to-many fragmented and cross referencing mechanism between the OCH customer Id and other duplicated across applications’ customer records. operational silos, resulting Source Data History - OCH also maintains a history of the changes that have in an inability to provide a been made to the customer profile over time. This history allows the Data Steward single, trusted customer to not only see the lineage associated with a customer record, but also provides the profile to business ability to optimize the source and attribute survivorship rules which contribute to consumers. It is often building the “best version” record. The history also enables the Data Steward to impossible to determine roll back the system to a prior point in time to undo events such as a customer which version of the merge. customer profile (in Survivorship - The data stewardship and survivorship capability consists of a set which system) is the most of features and processes to analyze the quality of incoming customer data to accurate and complete. determine the best version master record. OCH include rules based survivorship The “Consolidate” pillar that not only includes pre-seeded rules through its integration with a powerful resolves this issue by Business Rules Engine (Haley) but also allows users to define any new rule in delivering a rich set of addition to allowing users to integrate OCH with any other rules engine. OCH also interfaces, standards includes support for new rule types like Master and Slave that would determine compliant services and survivor and victim records. Finally OCH also supports configuring household processes necessary to survivorship rules. consolidate customer information from across the enterprise. This allows the deploying organization to implement a single consolidation point that spans multiple languages, data formats, integration modes, technologies and standards. Some of the key features in the “Consolidate” pillar include: List Import Master Data Management 35 Cleanse Centralizing the management of customer data quality has always been a goal of Customer Hub solutions. The “Cleanse” pillar in OCH provides an end-to-end integrated data quality management functionality to analyze/profile the data, standardize and cleanse the data, match and de-duplicate the data and finally enrich the data to create the best version master record. The powerful end-to-end data quality capabilities around profiling, matching, standardization, and address validation are available for all Oracle MDM Hubs, and are discussed later in this document. Additional Cleanse pillar capabilities include: Data Decay - Data decay refers to the way in which managed information becomes degraded or obsolete or stale over time. OCH provides data decay dashboards to monitor and fix the data decay of Account and Contact records. These dashboards are accessed through the OCH administration screens. OCH Data Decay management consists of the following key components: o Decay Detection: Captures updates on the monitored attributes / relationships of a record and sets the decay metrics at the attribute level granularity or relationship level granularity. o Decay Metrics Re-Calculation - Process to retrieve Decay Metrics for the monitored attributes/relationships of a record, use a set of predefined rules to calculate the new metrics value and update the metrics. o Decay Correctness - Identifies stale data based on certain criteria and triggers a pre-defined action. o Decay Report - generates Decay Metrics charts on a periodic basis Guided Merge & Un-Merge - Guided Merge allows end-user to review duplicate records and propose merge by presenting three versions (Victim, Survivor and Suggested) of the duplicate records and allows end users to decide how the record in the OCH should look like after the merge task is approved and committed. Similarly the Un-Merge feature rolls back a previously committed merge request. Enrichment - OCH provides an out of the box integration with enriched content from external providers such as Acxiom and Dun & Bradstreet. High Performance - Oracle data quality solution has a proven track record in terms of scalability and performance, handling large volume, highly-scalable, critical applications. It is a highly-scalable so

Master Data Management PDF

Document Details

Tags

Related

Summary

Full Transcript