Data quality is not magic

Data quality is not magic
IMS Team Sandro Oldenburg
Contributed by: Sandro Oldenburg

Sandro is one of the database specialists at IMS. If Sandro does not optimize customer databases, he drives his little Fiat 500 off track with just as many things - after all, Tempo is in many cases a good answer.

Databases are essential for work in the Facility and network management, Over the years they grow, and with them the unnecessary ballast that they drag along. This includes duplicates as well as stale data and changed parameters such as new device locations.

With IMSWARE However, it is easy to identify and correct such correspondence.

First: nomenclature!

Essential for good data quality is a good nomenclature. Only when labels are consistent and stringent can data be reliably managed, corrected and updated. The nomenclature should also consider two other issues: internal processes and compliance requirements. An important point in the definition of nomenclature is keeping an eye on the purpose of data usage. It should also be clear that enough time should be allowed for the development of nomenclature, because the more complex the content and task structure, the more complex the nomenclature must be.

Structure and monitor ...

For the data to be reliable and easy to control, a clear structure is essential. Conclusive logic is the best guide here, because then you can run queries safely. Once the concept has been defined, it must be implemented in the company in a binding and consistent manner. Interpretations by colleagues are to be stopped.

Incidentally, if the database is used in multiple languages, all queries must be defined in multiple languages and, of course, tested.

Search and find?

If the database is set up correctly, the work can begin. An example would be to find duplicates. In the case of tabular databases, sorting according to the relevant criterion is often sufficient here. With extensive datasets like in the FM, however, a specific query is necessary. If fuzzy duplicates are to be found, for example entries with typos, only special tools such as the DeduplicationWizard or the DataQualityTools will help. For example, for certain searches, IT devices might use unique criteria such as the MAC address. For rooms or their furnishings, the unique space markers help to narrow down the result set.

And now: keep an eye on things.

Databases are growing and changing constantly. It is therefore important to carry out regular checks on data consistency and to constantly update the databases. Then you also keep perspective in perspective.

In this sense…

Data quality is not magic - IMSWARE.de

This blog post is based on the more detailed article "Ensuring data quality through software", published in the LANline 03/2017.

Your feedback?

Be the first to rate :-)

Great. Thank you very much!

Maybe you want to follow us on ...

We are sorry that you did not like this post so much.

Blog
Zurück
The palace of the Maharajah, Mullewapp and CAFM
Weiter
Secured: SIGNAL IDUNA relies on IMSWARE