Data Mining Report

A data mining report for AS Computing.

HideShow resource information
  • Created by: Anonymous
  • Created on: 08-10-07 14:07
Preview of Data Mining Report

First 545 words of the document:

H.Hales Database management systems Data mining report
In this report I will be describing the functions and the purposes of data mining. Data
mining is the process of extracting and investigating large amounts of information
from the data warehouse.
The data warehouse is a system that stores huge amounts of information / data,
normally includes databases, tables etc. A database warehouse would commonly
feature in a legacy system which is a term which refers to existing computer systems
and applications with which new systems or applications must exchange information.
Data warehouses also include data about data which is commonly called Meta data
this basically is just information like who collected it, the time etc. about other
Data mining is just looking back and investigating all of the information that is held in
the data warehouse. Data mining looks for certain information in the data warehouse,
it looks for trends in the information, for example data mining would recognise trends
like "people buy more barbeques in the summer" or "people over the age of 55 don't
buy this particular brand of beer" and so on. Data mining also recognises patterns
that humans wouldn't easily be able to recognise.
Data mining works by using modelling, it then simply builds a model, and the models
are built from a set of different examples or from Algorithms (mathematical
instructions. Based on the data from the examples where the answer is known, it will
then apply this to various other different examples where it doesn't have the answer.
Data mining analyses different relationships and patterns that are stored in data
based user friendly queries. There are four main types of ways that relationships are
sought, they include:
1. Classes: this is where the data is used to locate data in different groups. An
example could be one of which I used earlier in this report and that is that
people would buy more barbeques in the summer, and as a result in the
summer they could put forward certain deals on barbeques to increase sales
and customer popularity.
2. Clusters: Data would be grouped by logical relationships or to customer
preferences. An example of this would be for Tesco's who own a chain of
shops all over the UK, and from clusters they would work out for example that
on the south coast they sell more sun tan lotion than in the Midlands, or in
London they sell more champagne than anywhere else in the country etc.
3. Associations: Here data would be mined to identify different associations. For
example where two products are bought at the same time like crisps and a
drink, this is not saying they are always bought at the same time but it is a
popular trend.
4. Sequential patterns ­ This is where data would be out to find different
behaviour patterns and trends. For example when a person buys a games
console they could predict the likelihood that you would buy games or
accessories with it.

Other pages in this set

Page 2

Preview of page 2

Page 3

Preview of page 3

Here's a taster:

H.Hales Database management systems Data mining report
There are five main functions of data mining, they include:
1. Extract, transform and loading transactions onto the data warehouse
2. Store and manage the data in a multidimensional database system
3. Providing data access to business analysts and IT professionals
4. Analyzing data by the application software
5. Presenting data in different appropriate formats, for example graphs, tables,
charts etc.…read more

Page 4

Preview of page 4

Here's a taster:


No comments have yet been made

Similar Computing resources:

See all Computing resources »See all resources »