Intro

What Kinds of Patterns? 

Association rules 
Detect sets of attributes or items that frequently cooccur in many database recordsand rulesamong them

Classification

Build a model of classes on training dataset, and then,assign a new record to one of several predefined classes

Clustering 
Partition the dataset into groups such that elements in a group have lower inter-group similarity and higher intra-group similarity

Transactional Databases

I={x1
, …, xn
} is the set of items
 An itemset is a subset of I
 A transaction is a tuple (tid, X)
 Transaction ID tid
 Itemset X
 A transactional database is a set of transactions

Web Usage：Logs and IP package header streams
 Mine Weblog records to discover user accessing patterns of 
Web pages