In this chapter, we started by looking at the concept of guided analysis and we worked through a relatively sophisticated example of how a guided analysis can be constructed. We moved on to consider the importance of metadata as an essential communication protocol for data analysis, particularly where a self-service approach is sought. If you talk to BI professionals or read about the subject, you will find strong recognition for the importance of metadata dictionaries and semantic layers, but you will also find an acknowledgment that we are poor at implementing business-targeted metadata.

This chapter should give you a sense of how metadata can be a common language through which user requirements can be expressed and understood by BI consumers, developers, and applications. You should now see how easy it can be to incorporate configurable business rules into an analysis using metadata tables.

This chapter finished with a metadata-driven self-service analytics case study. This example...