Enriching the MDM Repository
Reading Anshuman's post on transactional data elements in the MDM repository for increasing business value, I was reminded of an interesting solution we defined for a Japanese publishing company. I am a fan of Japanese culture, and it was exciting to work with people who produce manga, anime, and lots more cool stuff.
- Basic/ Static information – demographic master data
That which refers to the information that do not change during the lifecycle of association of the consumer with the company during the normal course e.g. Name, User_id, Date of birth - Enriched Dynamic information or meta-data
That which refers to the information that changes during the lifecycle of association of the consumer with the company e.g. Education, Income, Usage of Vehicle, weight - Transactional Information
That which refers to the information that changes depending on the activity of the consumer on his purchase behaviour. This information is different from the other two in the way of ingestion as this is the information that is captured automatically.
The consolidated view of consumer information, aggregated from demographic static master data and enriched with dynamically generated meta-data was sent to an analytical search engine for recommendations generation, where the static data (basic data E.g. name) would be augmented based on the transactional details (publication details search query) and the dynamic data (interest area or previous buying pattern).
In order to implement this, whenever the user queries his/her static data is used to trigger web crawlers. The web crawlers, monitors web log activity and tracks his/her buying propensity and stickiness on the website (dynamic data). Based on the meta data so received, the transactional data, sent back to the user is enriched with additional details. The enrichment data, is customized based on offering set up by the publishing product promotion manager (in this case magazines or book ranges).
It is solutions like this that presage the convergence of data repositories and operational systems, mediated through intelligent middleware. We are also seeing convergence on the other side, with data marts consuming real time data from the repository. The information ecosystem of the enterprise will emerge as a seamless continuum of producers and consumers, facilitated through closed-loop data quality.
