Health Information System Safeguards

24 D A T A B A S E T R E N D S A N D A P P L I C A T I O N S

TRENDS

OCTOBER/NOVEMBER 2016

DATA GOVERNANCE is sometimes viewed as a roadblock that keeps data scientists and analysts from turning data into business insights quickly and efficiently. Yet, it’s often a lack of sound data governance that prevents organizations from realiz- ing the full value of their data.

Data governance deals with such ques- tions as the origins, or lineage, of data; who can access data and what they can do with it; how data is categorized; and the quality and completeness of data.

Here are five ways that a modern approach to data governance can make your data scientists and analysts more productive, enabling your business leaders to gain insights more quickly.

Good business metadata is good for the business. Effectively gov-

erned metadata—that’s data that labels or categorizes other data—facilitates the dis- covery process for data scientists, helping them find the data they need, when they need it. Tagging and cataloging data at the time of ingestion will help your organiza- tion keep its data lake clean while giving your data scientists a better understanding of what’s available to them.

Effective schema management saves time and money, especially in

a big data environment. Schemas define how data should be read. It’s essential that data consumers know which schema to use when looking at particular files. Yet managing schemas can be difficult, par- ticularly in a big data environment. Pro- grammatic technical and business schema discovery eases the problem.

When a new dataset is ingested into a data lake, an open source tool can help you determine the schema automatically, and,

in a mature environment, match the newly discovered data to existing business meta- data, providing you with both the business and technical metadata immediately. Pub- lishing, curating, and governing all known schemas will save your data scientists and analysts considerable time, freeing them to focus on their primary roles.

Good data quality and profiling can accelerate time to insight. Poor

data quality is among the key reasons that 40% of business initiatives fail to achieve targeted benefits, according to a report by Gartner Inc., which also notes that data quality affects overall labor productivity by as much as 20%.

Developing a sound architecture and effective data quality protocols will help you keep your data lake from becoming a data swamp. Establishing data-usage agreements between producers and con- sumers of data will also prove helpful, as these agreements give everyone a better idea of the level of data quality expected and how it will be documented. Profiling data and storing the profiles with meta- data is also a useful practice, giving your data scientists a better understanding of the types of data contained in the system and allowing them to formulate hypothe- ses more quickly.

Data lineage can help keep you from getting sued or fired. In an

era of data breaches, data governance can provide important protections to your business and its employees. Data gov- ernance won’t stop determined hackers from gaining access to secure data, but, in the event of a breach, it will help you understand what has and hasn’t been compromised.

Data governance affords particular protections to people who work in regu- lated industries such as financial services and healthcare. In an audit, data gover- nance enables you to show exactly where your data came from and how you made particular calculations.

Your models and analyses w ill run right in production. If your

data governance program includes the measures discussed up to this point, your data will be of sufficiently high quality that you’ll experience fewer problems with models and analyses in production.

If you go a step further and establish preventive and detective controls, you’ll gain additional benefits. Preventive con- trols help ensure that low-quality data isn’t used by the business. Detective con- trols help the production and operations team troubleshoot jobs that fail as a result of data quality issues.

Saving Time, Energy, and Money With a modern data governance pro-

gram in place, data scientists needn’t spend their working hours looking for data, trying to understand definitions, wondering whether datasets are com- plete and accurate, or trying to determine where data originated. That saves time, energy, and money while improving the quality of business decision making. A sound data governance program will also keep your organization safe and compli- ant, with full documentation of how data is used and by whom. �

Ben Harden leads the Data and Analytics Practice at CapTech.

The Five Ways Modern Data Governance Helps Business Productivity

By Ben Harden

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.