IBM InfoSphere Master Data Management, Version 11.3

Managing data with InfoSphere MDM Reference Data Management Hub

IBM® InfoSphere® Master Data Management Reference Data Management Hub is installed and configured on IBM WebSphere® Application Server so that you can use its browser-based graphical user interface to manage your reference data.

InfoSphere MDM Reference Data Management Hub provides centralized management, stewardship, and distribution of enterprise reference data. It supports defining and managing reference data as an enterprise standard. It also supports maintaining mappings between the different application-specific representations of reference data that are used within the enterprise. InfoSphere MDM Reference Data Management Hub supports formal governance of reference data, putting management of the reference data in the hands of the business users, reducing the burden on IT, and improving the overall quality of data that is used across the organization.

InfoSphere MDM Reference Data Management Hub is built on IBM InfoSphere Master Data Management. It is a new master data domain with its own data model, services, and stewardship user interface for managing enterprise reference data. For business users, the user interface and flexible data model ensure a quick implementation and minimize the need for IT involvement on an ongoing basis.

InfoSphere MDM Reference Data Management Hub include the following key functions:

Role-based user interface with security and access control
Management of reference data sets and values
Management of mappings between reference data sets
Import and export of reference data by using CSV and XML format
Versioning support for reference data sets and maps
Change process that is controlled through configurable lifecycle management
Hierarchy management within sets, or through multiple levels across sets, of reference data

InfoSphere MDM Reference Data Management Hub also integrates with and complements IBM InfoSphere Business Glossary and the broader portfolio of IBM Information Management products.

Some examples include the following drivers for managing reference data:

Banking: European Union countries adopted different versions of the European NACE industry classification code set as their national standard. Banks with operations in multiple countries must support the individual national NACE code sets of each country at the local level. Those banks also must reconcile the differences across countries at the corporate level. With InfoSphere MDM Reference Data Management Hub, you can manage the different versions of the NACE codes from a central point. It simplifies creation of mappings between the different versions and supports reconciling values across the different set versions.
Banking: Banks face the dual challenge of distributing reference data changes to downstream applications and assessing what applications are impacted by a reference data change. InfoSphere MDM Reference Data Management Hub provides a subscription model that makes it possible to assess what applications are impacted by a particular reference data set change.
Healthcare: The United States government mandated that healthcare payers and providers must use ICD-10 codes for diagnoses and procedures by October 2013. The challenge the healthcare industry faces is the lack of a one-to-one relationship between the ICD-9 and ICD-10 codes. InfoSphere MDM Reference Data Management Hubsupports central management of the ICD standards and mappings and the definition of company-specific mapping schemes.
Cross Industry: Loading data into a data warehouse or a master data management hub typically requires reconciling reference data from multiple sources. InfoSphere MDM Reference Data Management Hub supports publishing reference data map files for batch and load jobs to support warehouse or InfoSphere MDM hub data load.

What is reference data?

Reference data is any data that is used to categorize other data within the enterprise. Reference data is commonly stored in the form of code tables or lookup tables, such as country codes, state codes, and gender codes. Reference data is used within every enterprise application, from back-end systems through front-end commerce applications to the data warehouse. Business users recognize reference data as code choices within the pick-lists of their business application user interfaces.

Reference data code tables are often implemented in the database as relatively simple structures with a key column that contains a code value and a description column. Some code tables, such as NACE, have few values (in the tens or hundreds). Others, such as healthcare ICD-10 codes, have larger numbers of values (in the tens of thousands). They can be flat lists or have a hierarchical code structure. A hierarchy can be defined over the values within the code table, or a hierarchy can be defined where each level is a code table in its own right.

The structural simplicity and static nature of code tables belies the cost and difficulty of managing code tables at the enterprise level. The problems with code tables include the sheer number of code tables that are used within and across enterprise applications. Each application often has its own representation and set of values for code sets defining the same thing. When you integrate data across applications, you must translate between the different code table representations to categorize data in a consistent way. Mapping between the different representations and tracking changes across all the different code table variations on an ongoing basis can be a major challenge. Many enterprises struggle with this challenge by using spreadsheets and other error-prone manual processes to record and manage changes to reference data sets and their relationships to each other. The lack of change management, audit controls, and security is often a compliance risk. Since reference data is used to drive key business processes and application logic, errors in the reference data can have a major business impact. Mismatches in reference data can have the following effects:

Major impact on data quality
Loss of integrity of BI reports
Common source of system integration failure

Last updated: 27 June 2014