Step 1: Understand the kinds of data needed for delivering the results
To construct a
Data Preparation View model, start by acquiring an understanding of existing data models and by selecting portions of data files and schema that are relevant or needed for the data analytics solution.
The following items are important things to consider during this step:
- The Question Goals (and their topics) and Insight elements (including their learning period and update frequency attributes) in the Business View model should be understood and referred to during this step
- Entity Relationship Diagrams or data warehouse schema models and other documentation can be used in this step.
To get started, you can ask the following question to guide you: “what kind(s) of data are actually needed for delivering the results and answering the Question Goals at hand?”
Visualization, and initial descriptive/statistical analyses may be performed at this step by data
scientists to understand the shape, size, and type of the data at hand and to verify the quality and
meaning of data attributes.
Step 1
Step 2
Step 3
‹
›