Data Mining - Narrative
Links to the essential reading, resources, activities, databases, advanced topic forum and assessment upload referred to in this narrative are in the theme block in myCourse.
Data mining and Knowledge Discovery in Databases (KDD) is a rapidly growing field. It is increasingly important for an organisation to "mine" its database for knowledge that can be used for business decision support and competitive advantage. KDD is generally considered to encompass more than just data mining, and includes the selection, cleansing, transformation and display of the discovered information. Data mining depends on integrating many skills including databases (including SQL), Artificial Intelligence/machine learning, statistics, and business intuition. Consequently, this will be a guided, practical introduction only to one particular data mining application and technique - "market basket analysis". The advantage of focussing on this particular data mining application in sales and marketing, apart from its clear business importance, is that you are already familiar with the common "orders, items, products" data structure of the mySports and OE database.
You are going to work through a self-paced, online resourced "A Practical Introduction to Data Mining" This is based on the essential reading for this theme of Berry & Linoff - Market Basket Analysis and Association Rules - Ch 9. The activity, which is based on the mySports database, references a number of other online resources including CRISP - a CRoss Industry Standard Process for data mining. We will continue to use iSQLPlus to access the mySports database, and we will also use Excel and the Weka open source data mining tool.
The assessed activity is based on performing a market basket analysis on the larger, more realistic OE database. Other Weka data mining functionality is the advanced topic and forum for this theme.