Data Bases
- Data comes in a range of formats from highly unstructure,
to highly structured.
- There are techniques to deal with all of these types of data,
but it is important to realise that different techniques are
appropriate in different circumstances.
- Traditional relational databases are generally quite
structured.
- The more structured something is, the easier it is for a
computer to deal with.
- Obviously, many large data bases can be used to find the useful
things that for which they are designed. A payroll database
can easily tell you how much total pay was sent out last month.
However, it might also be able to tell you how likely people are
to retire early.
- That is, data bases can be used for unexpected tasks.
- This is one domain of application for machine learning techniques,
finding new knowledge.
- Additionally, less structured data bases can be used.
- There are techniques for adding structure and regularity.
- You can even use collections of documents, like journal or
newspaper articles, to find information.
