Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

Data Mining and Knowledge

Discovery
Sally I. McClean
University of Ulster

I. Data Mining and Knowledge Discovery


II. The Technologies
III. Data Mining for Different Data Types
IV. Key Application Areas
V. Future Developments

GLOSSARY Knowledge discovery in databases (KDD) is the main


objective in Data Mining. The two terms are often used
Association rules link the values of a group of attributes, synonymously, although some authors define Knowl-
or variables, with the value of a particular attribute of edge Discovery as being carried out at a higher level
interest which is not included in the group. than Data Mining.
Data mining process takes place in four main stages:
Data Pre-processing, Exploratory Data Analysis, Data
Selection, and Knowledge Discovery. DATA MINING is the process by which computer pro-
Data mining tools are software products; a growing grams are used to repeatedly search huge amounts of data,
number of such products are becoming commercially usually stored in a Database, looking for useful new pat-
available. They may use just one approach (single terns. The main developments that have led to the emer-
paradigm), or they may employ a variety of different gence of Data Mining have been in the increased volume
methods (multi-paradigm). of data now being collected and stored electronically, and
Deviation detection is carried out in order to discover an accompanying maturing of Database Technology. Such
Interestingness in the data. Deviations may be detected developments have meant that traditional Statistical Meth-
either for categorical or numerical data. ods and Machine Learning Technologies have had to be
Interestingness is central to Data Mining where we are extended to incorporate increased demands for fast and
looking for new knowledge which is nontrivial. It al- scaleable algorithms.
lows the separation of novel and useful patterns from In recent years, Database Technology has developed in-
the mass of dull and trivial ones. creasingly more efficient methods for data processing and

229

You might also like