Data Cubes
- Databases have a lot of dimensions.
- They function more effectively on a small number of key dimensions.
- Also, it is useful for data visualisation to have a small number
of key dimensions.
- Consequently, it is important to reduce the number of dimensions
in a database.
- Other features can be attached to these primary dimensions.
- It's quite like having primary keys.
- If a database is translated to one with a few primary dimensions,
the data base is now called a data cube or just a cube.
- This data cube can now easily be used with machine learning
algorithms.
- This relates to the term Online Analytical Processing (OLAP),
and drilling down into the data.
- I looked at a 1997
article on Data Cube by Gray et al
for this one.
