The attached schema provides a broad overview of a different dimension - raw, transformed and enriched - by source/cross channel. This design is specific for Estee Lauder Companies.
- Each data source is depicted by a table with a black header, showing the dimensions from the source and the Beckon Dimension HID it gets transformed into.
- Relevant maps used for the data source are to its right, in purple, with:
- The first row being the key to the map.
- The second (purple colored) row being the map's name (sometimes with a QA dimension marked as ., explained later)
- And the following rows being the dimensions that get enriched into the data.
- For some maps, there is a QA dimension linked to the maps name (refer legend on the top right for the exact QA dimension pertaining to that map).
- This QA Dimension can be used to find unmapped dimensions for an instance under a rich dimension table. The relevant QA dimension can be filtered to specified or unspecified to differentiate whether that instance has been unmapped or if the instance has been mapped as blank in the map.
- Some Cross Channel/Platform maps are in Red.
- Where parsing occurs, the dimensions brought in are depicted in a green table, along with the regex-ed instances.
- In some places, a cleansing or mapping functions are depicted as DATE (-> SN, FY) , meaning the date dimension is used to denote the season and fiscal years. Different processes may be used (rich dimensions, transformer function, parsing, etc).