The big data phenomenon expands the purpose and changes the role of data modeling. The level of uncertainty about data modeling in today’s data ecosystems is high. Most practitioners have more questions than answers. Has data modeling become obsolete? Does unstructured data make modeling impractical? Does NoSQL imply no data modeling? What are the implications of schema-on-read vs. schema-on-write for data modelers? Do entity-relationship and star-schema data models still matter?
Data modeling is still an important process—perhaps more important than ever before. But data modeling purpose and processes must change to keep pace with the rapidly evolving world of data. This course examines the principles, practices, and techniques that are needed for effective modeling in the age of big data.
You Will Learn
- To distinguish between data store modeling (schema on write) and data access modeling (schema on read) and when each is useful
- The elemental characteristics of data that provide a common denominator for data modeling for all types of data
- How the common denominator is used to map various kinds of databases, including relational, dimensional, NoSQL, NewSQL, graph, and document
- When traditional logical-to-physical modeling works and when it makes sense to reverse the process as physical-to-logical
- Trade-offs between methodological rigor and discovery-driven exploration in data modeling
- Data architects; data modelers; database developers; data integrators; data analysts; report developers; anyone else challenged with the need to make structured enterprise data and non-traditional data sources work together.