Preparing for business-driven quality management
Data quality is vital for driving business value. IBM Information Server unified governance enables you to link business-driven quality standards to your information assets, so that you're not simply defining quality expectations; you are enforcing them.
When your glossary and your information aren't talking...
Let's take a moment to break down the relationship between the different components of Information Server's approach to data quality management.
First, you determine the way that you want your information to behave. Terms are the building blocks of your business vocabulary. Terms are associated to governance policies and rules, and all of these glossary assets combined make up your business glossary. The glossary is a statement of your governance strategy.
Next, you have your information assets: basically, all of the enterprise information including databases, tables, servers, etc. that make up the information landscape for your enterprise.
The goal is to take your mountain of information and make sure all of that information complies to the policies and rules set forth in your glossary. So how do you get the mountain to talk to the glossary and deliver trusted, meaningful information to the organization?
Information Server does a lot of this work for you automatically. Enterprise Search keeps track of your assets and the relationships between them, giving you a view into all of your enterprise information, cataloged or not, so you can see what's being governed and what needs attention. Automated discovery can import and analyze all of the data from a given data connection. Column analysis identifies properties in your data like data class, format, type, frequency distribution, etc. Based on analysis results, the system will assign terms from your glossary automatically. Associating terms to your information assets creates a link between the two components and makes it possible to enforce your quality standards.
To customize this feature and make it more powerful, you can associate terms to data classes. This way, any time a data class is selected for an asset during analysis, the term will also be associated.
New automation rules take this approach one step further, allowing you to associate conditional statements that enforce aspects of data quality to terms in your glossary. When those terms are associated to information assets, rules or quality dimensions defined by the automation rules get bound or applied automatically.