Using
Discover in Studio, you analyze the data and
discover insights hidden in data.
Discover lets you identify and pick meaningful
visualizations and capture the results of the data exploration phase.
You can find the
Discover area of Studio here:
Here is an example of how
Discover might look for your project:
In this diagram, you can see the
Discover area, with the Map visualization added.
To the right, notice some of the other visualization components you can add to
your project.
Some of the
examples of the insights you can create with
Discover are:
- Use search and Guided
Navigation to identify subsets of data for deeper analysis.
- Use interactive
visualizations to let the data speak for itself, in raw or aggregated forms, in
text or in numbers. Statistical techniques (regression, correlation) and
machine learning tools (classification, clustering) are used in Big Data
Discovery to drive these visualizations.
- Join (or combine) multiple
data sets for new insights.
- Create projects that serve
as models for different outcomes. These models automatically understand the
weight of different factors in contributing to given outcomes without requiring
experts to hand-code rules or heuristics.
- Add new data to existing
projects or run your projects on existing data that has been altered to examine
new scenarios. (This is known as "what if" analysis.)
- Export your data sets from
Big Data Discovery back into Hadoop, so that you can run complex models built
in other tools (such as R, or Python) on your custom-created data sub-sets,
such as on different customer segments.