1.4 KiB
What shape is your data?
Sorts of structure
- Structures in programming environments
- Structures in data models
- Structures in serializations (data formats)
- Structures in exchange protocols
- Structures in user interfaces
Shapes available
- Tables
- Trees
- Graphs
- Media (raw data)
- Documents/objects
Tables
What does a table look like?
A table can be seen as a collection of cells where each row represents a thing and each column represents a type of information about the thing.
A strength of a table is that it's easy to understand, search and communicate meaning.
However, conveying hierarchy cannot be done.
Trees
They have only one parent node instead of two in family trees. The root node is represented at the top in diagrams.
Other
Tree nodes can only have one parent at the most. However, when there start to be more parents for a node, we're now talking about a graph.
Tables: General-purpose Trees: Heterogeneous and hierarchical, structured data Graphs: Heterogeneous, non-hierarchical, structured data Blobs: Inaccessible data for storage Features: Searchable information derived from blobs Documents: Rich, but not interrelated data
Summary
- Where does your data come from?
- How well does it fit your needs?
- How is it structured?
- How well can it be integrated?
- How should your data be structured?
- How will you use yoru data?
- Will others use your data?