![]() For instance, if you would process Customers of a Telecom Network Operator and these Customers could be either Customer of an Internal Service Provider or an External Service Provider, then you would have to Enrich the Customer data with data of the related Service Provider in order to Select the Customers of the internal Service Provider.Ĭoupling Rules are necessary to relate the Source data to Target data. I had some difficulty with the difference between Enrichment Rules and Coupling Rules.Įnrichment Rules add extra related data from the source model, that are necessary for further processing. This model is a usefull structure for functional design documents of ETL functions. The underlying model shows a possible structure. ![]() His thesis is that by structuring the functional design of the ETL process, it should be possible to automatically generate about 80% of the actual ETL code. ![]() It interested me because I both know the author and the Data Warehouse environment he uses as an example. Last week I read an interesting article about “Model driven Design of ETL functions” by Mark Zwijsen in Database Magazine. Now working on Business Intelligence (BI) projects for the last one and a half years, it seems to me that these kind of standards do not yet apply to Data Warehouse environments and more specific to ETL-processes (Extract Transform Load). Analysis and functional design for regular application development has been standardised over the years.
0 Comments
Leave a Reply. |