Master Data Management Data Hub

Now that we have examined the data and started to make determinations with the data, we now need to correlate our data into a Cross-Reference Hub as part of our Master Data Management solution.  In order to correlate the data, we import the separate files into our Master Data Management software application’s central database.

As discussed in the previous post, some of the data may need to be pre-processed for better match results.  Based on the data determinations, in addition to running the import process, we also choose the MDM solution’s data transformation processes that are relevant to the client’s data and goals.  

Once the data has been pre-processed and imported, we can run an initial Match Project.  The Match Project generates a Cross-Reference Table by linking every record in the database based on their relationships as indicated by the Entity Level IDs and Contact IDs.

With a Cross-Reference Table, the client can: 

1.  View - See matched and linked records from multiple data systems.  For example, a company may want to give a sales rep a view of all the authorized records and information that are part of the same organization regardless of the data source in order to help up-sell and cross-sell.

2.  Report – Analyze data utilizing information across multiple data systems and properly counting the number of entities and contacts based on their Level and Contact IDs.  For example, a CRM system could be linked to the Cross-Reference Table in order to generate a report indicating how many unique entities that are customers or prospects within a certain market.

3.  Data Quality – Various data quality reports can be developed.  For example, a duplicate record report can be generated.  Then, as the system gets updated, subsequent reports can indicate the progress being made on cleaning-up duplicates as there should be less duplicates each subsequent day, week, or month that you run the report. 

4.  Operational System Enhancement – As part of the Master Data Management project, you may reformulate the data for matching purposes in order to get better match results.  The reformulation does not impact the operational system of origin.  For example, you may want to remove Department data that is in an Address field in the operational system of origin for the Match.  However, if desired, you can then use the Cross-Reference Table to help change the data back in the original operational system. 

In addition to creating a Cross-Reference Table, some organizations need to take the Master Data Management solution a step further and create a Master Record.  Sometimes the Master Records are also referred to as the Record of Record or the Golden Record. 

A Master Record is a record that uses information on the record from any of the data silos that have been linked by the Cross-Reference Table in order to create a record that is complete, accurate, and acts as the standard for that record based on the best available information across the data silos.  Many times a Master Record is created so that reports can be created with a uniform record representing that entity or contact within all reports.

When it comes to creating a Master Record, you can choose a record that already exists and deem it to be the Master Record or you can create one by taking the best information from the different data sources and then build the Master Record. 

Comments

[Kingsley Idehen]

Steven,

Are you looking into the synergies between the RDF Linked Data meme and MDM?

Simple value proposition of RDF Linked Data in the MDM realm comes down to this: the Identifiers of the MDM Entities incorporate HTTP, and by doing so it becomes possible to de-reference the description of an MDM Entity (which means exposing Entity Attribute and Relationship Data using a range of represenatations e.g. HTML, RDF/XML, RDFa etc.).

Some live examples:

1. http://demo.openlinksw.com/Northwind/Customer/ALFKI - View of Customer "ALFKI" note the value of foaf:primarytopic property

2. http://tinyurl.com/6a24ts - a Product Information Integration example (basically a product catalog using our Virtuoso Universal Server as an example).

Kingsley