by Phuong H
In the article ” Logical Data Modeling: A Key to Successful Enterprise Data Warehouse Implementations” written by Sreedhar Srikant. The author mentions five challenges to designing and implementing an logical data model (LDM). The first challenge is lack of common vision of data warehouse which means the company have an unclear requirements; they just build the model without planning for the future use. Second, the organizational muddle, in other word, the company assign a job without guidance. Third, adapting to existing models and migrations in which they transfer data to a different model without planning. They just want to use it right away and not thinking about the future use. Fourth, source system-based design which they use the “default” system rather than a customized system that the company needs. Lastly, “trying to create an all-in-one model” in which they create a “hybrid models that combine logical, physical and semantic information in a single model can also be a tempting trap. Such models are difficult to use and implement” (Srikant 2006).
I didn’t realized these challenges would happen in LDM especially the first challenge the author mentions, lack of common vision of data warehouse. I thought these already planned in the conceptual model. And I think most of the time people are trying to create an all-in-one model because it would save time. For example, rather than building a conceptual model, they would jump right into a logical or physical model.
Srikant, S. (2006). Logical data modeling: A key to successful enterprise data warehouse implementations. DM Review, 16(9), 13-13. Retrieved from http://search.proquest.com/docview/214675479?accountid=10357