cancel
Showing results for 
Search instead for 
Did you mean: 

Consistent Master Data

Former Member
0 Kudos

Hi All,

Harmonizing and ensured consistency of master data in the daily operations of an enterprise are only two of the functions of master data management.

But once harmonizing and consistency of master data is ensured, how can one

prevent the duplicates being created in client systems or Central MDM ?

Regards,

Priti

Accepted Solutions (1)

Accepted Solutions (1)

MichalKrawczyk
Active Contributor
0 Kudos

hi,

>>>>how can one

prevent the duplicates being created in client systems or Central MDM ?

they will be created in central MDM but you can remove them

using merging for example

if we talk about client systems we cannot delete any master data there

what we can do is we can mark them as "inactive"

as if they were used in most systems (like sap) it's quite difficult to

delete any master data (like materials, etc)

Regards,

michal

Former Member
0 Kudos

Hi,

As we know that CMDM can compare and merge records , this process is manual.

Is there any way for automatic merging of records? Or since the decision of merging records is with the business people , the merging of records has to be manual.

Please elaborate on this.

Regards,

Priti

MichalKrawczyk
Active Contributor
0 Kudos

Hi,

>>>>Or since the decision of merging records is with the business people , the merging of records has to be manual.

exactly because of that

would you like your customers to become inactive in your system

only because merging rules were to broad ?

Regards,

michal

michael_theis
Active Contributor
0 Kudos

Hi Priti,

> Is there any way for automatic merging of records?

The standard MDM functionality in Data Manager does not yet provide the automatic merge. Nevertheless you could implement an automatic merge using the APIs.

> Or since the decision of merging records is with the business people , the merging of records has to be manual.

In MDM 3.0 we had the possibility to merge automatically and it caused a lot of trouble. Merging records may be trivial for an business user, but there are lots of things you have to think about if you want to merge automatically. First of all, your matching strategy has to be very concrete, it must find records, that are real duplicates (e.g. the match score should be nearly 100%). Then you have to have a look at the records themselves: which data from which record should be kept? Imagine you've got two streets, the first is "21st Hemingway Str.", the second "21st Hemmingway Street". With your matching strategy you're able to find out, that you have to merge the records. Manually, you'd select the first one, as there's no typo in and it's more readable. Automatically, well, it's hard to implement a merge that selects the one without typos. An automatic merge would pick the the first one, based on its implementation. The more complex your data is, the more complex the merge itself will be. And master data is very important data for each company. So it should be really useful to take a look at the merge.

BR Michael

Former Member
0 Kudos

Hi Priti,

<i>the decision of merging records is with the business people , the merging of records has to be manual.</i>

In our first project of MDM SP 03, we had automatic merge and it caused a lot of trouble. Even when we were having rule of merging only records with above 92% match. This was because of clients different business processes. Thus its not system but Business to make decision on their data.

Moreover there is no standard way to unmerge the records.

Since Master data is backbone of transactional data and we are here for the need of consistent master data, it wont take us anywhere if MDM can not provide single truth of data.

Hope this helps,

+ An

Answers (1)

Answers (1)

Former Member
0 Kudos

Hi Priti Deshmukh ,

This website gives details regarding elimination of duplication:

How To identify identical master data records using SAP MDM 5.5 Java API's

https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/a06478f5-b28d-2910-e083-c28eb96b...

cheers!

gyanaraj

****Pls reward points if u find this helpful