| Knowledge discovery process of extracting previously unknown, actionable information from very large databases. -- Meta Group |
| Data mining is the process of discovering meaningful new correlations, patterns, and trends by sifting through large amounts of data stored in repositories, using pattern recognition technologies as well as statistical and mathematical techniques. -- Gartner Group |
| Analysis of data for relationships that have not been previously discovered. -- Whatis |
Three main objectives define the activities in data mining:
1. Prediction for future use/trends
Provides value in foreseeing customers' behavioral trends in a highly competitive market. Most business applications of data mining focus on this objective to guide marketing and retail decisions.
2. Description of patterns within a set of data
May be applied to a range of applications when analysis over time is disired. One example of this use involves statistical compilation of weather patterns ofr a geographic region.
3. Aiding Management decisions
Companies decide how to manage services for customers by examining the resks involved in providing these services to individual clients.
Data preparation is done prior to mining to ensure use of accurate data appropriate to yield the best possible results. The process of preparing the data involves:
* cleaning the data
* dealing with missing values
* data derivation
* merging data to make it more accessible for
excavation.
Data mining should not be implemented until the data has been adequately tailored to meet
the goals of the mining process.
Prepared by:
Susan Bravenec
Jennifer Morley
August 13, 1998