Topic 6: Database Systems

Page 1

Database theory

Explain what is meant by:

Data normalisation

It is a technique used for analysing data and is a staged process which improves the database design at each stage which...

Minimises duplication of data - you do not need to enter data nay more times than necessary.

Eliminates the redundancy of data - avoids the unnecessary repetition of the same data.

Ensures that data integrity is maintained - it makes sure that the data is error free and that there are not different versions of the data, depending whether it has been updated or not.

Allows the flexible extraction of information from the database.

Relational database

A relational database contains many tables with each table containing related data. The tables are linked together which means that data in any of the tables can be combined to give the user meaningful information in the form of reports.

ADVANTAGES:

Data consistency - information which is stored on multiple departments will all get updated. E.g. if someone changes address it will change on the payroll department and human resources as well.

Data redundancy - the data can be shared between a number of applications. Data doesn't need to be duplicated.

Data independence - data is independent than the programs that access it.

Data integrity

Data warehousing

Because of the huge amount of data involved, a very large database, called a data warehouse. A data warehouse is used to store all an organisations historic data and it is used by management information systems to extract information that will help managers make decisions.

Example 1 - finding out the day of the week on which a particular store sold the greatest number of a certain product in 2006.

Example 2 - how employee absence due to sickness varied over the last year between the Newcastle branch and Manchester branch.

Data mining

Once the data has been stored in a data warehouse it then needs to be 'mined' to discover:

Patterns in the data.

Associations in the data (e.g. people who read the Times newspaper are more likely to drink red wine).

Trends over time (e.g. a person is buying more healthy food and is drinking less alcohol).

ADVANTAGES

Data mining allows users to understand the data more by discovering meaningful patterns in the data. Looking for meaningful patterns in a huge mass of data is possible with the use of data mining software which presents the results in forms of graphs and tables.

Data mining can produce information such as:

Lists of customers likely to buy a certain product (based on what they have bought before.

Comparison with competitors.

Useful 'what if' results from modelling exercises.

Predictions for future sales.

Analysis of best sites for shops.

Sales patterns.

Customer buying patterns.

Who is most likely to change their credit card.

Those customers who are most loyal to the company/product

Distributed database

A distributed database is when the data in the network are distributed across a number of sites sites.

ADVANTAGES:

Does not rely on the storage of the data in a single location, so there is less risk of all the data being inaccessible.

There is improved performance of the database - queries run faster and complex data mining is completed in less time.

Applications can continue to function if the local server experiences a failure, but other servers with replicated data remain accessible.

Applications may normally access a local database rather than a remote server to minimise network traffic and achieve maximum performance.

DISADVANTAGES

More complex and costs more to install and maintain.

Increased security risk because files are transferred across networks.

If one server fails, then it can have an effect on the database and staff may not be able to access some of the data.

The system relies on data communications, so if a communication line fails then the data may not be able to be accessed.

As large numbers of staff access the database, there is a chance that you could get inconsistencies of data.