Database Reference
In-Depth Information
Master Data Management
Master data management (MDM) creates and maintains a single version of
the truth for key business entities such as customers, patients, products,
parts, suppliers, accounts, and assets, among others. MDM is an operational
system of record, and it plays an important role in a Big Data ecosystem.
IBM's InfoSphere Master Data Management is the most comprehensive
MDM solution in the market.
We think that MDM can provide a compelling starting point for Big Data
analysis, as MDM by definition focuses on the highest value entities within
an organization. Many organizations imagine that they would like to analyze
social media to determine customer sentiment, but do they know who their
customers are? Do they know their best customers? When embarking on Big
Data analysis of a precise subject, it often makes sense to leverage the knowl-
edge of an MDM system within a Big Data analytic application; for example,
understanding the sentiment and next best action for a particular group of
profitable customers instead of analyzing broad sentiment toward your
company.
Integration points between MDM and Big Data include ingesting and ana-
lyzing unstructured data, creating master data entities, loading new profile
information into the MDM system, sharing master data records or entities
with the Big Data platform as the basis for Big Data analysis, and reusing the
MDM matching capabilities in the Big Data platform (such as customer
matching). You'll find that IBM is a leader in this space, and although we
can't comment on some of the integration points coming to this problem
domain, IBM plans to continue and build integration points between MDM
and Big Data based on real-world customer use cases.
You should consider using MDM solutions with Big Data when the target
of your analysis is precise versus broad (aggregate)—an individual customer
or group of customers, a particular product or group of products, and so on.
In addition, consider integrating your Big Data into MDM systems when the
output of Big Data analytics should be “operationalized”; specifically, when
acquired insights are to be acted upon in operational systems. For
example, consider the case in which the next best action for customer reten-
tion should be consistently acted upon across multiple channels; the action
flag (or insight) should be stored in a MDM system.
 
Search WWH ::




Custom Search