
Data Dictionaries and Data Context

#175 - How to Solve Real-World Data Analysis Problems - David Asboth
Topic 5 of 12
Published May 20, 2024 · Clip 23:48–27:28 (3:40)
David explains that coherent data dictionaries are rare in the real world. Beyond recording columns and data types, a good data dictionary should capture the data generating process and business context. He gives the example of 'sold date' vs 'date sold' columns in used car sales data that looked redundant but actually measured different things due to dispute processes. The host connects this to data lineage and domain-driven design.