Table of Contents
Data Mining Techniques: Clustering (cont.); Memory-Based Reasoning
Today
What is Clustering in Data Mining?
Distance or Similarity Measures
Distance or Similarity Measures
Distance or Similarity Measures
Distance or Similarity Measures
Domain Specific Distance Functions
Distance (Similarity) Matrix
Example: Term Similarities in Documents
Similarity (Distance) Thresholds
Graph Representation
Simple Clustering Algorithms
Simple Clustering Algorithms
Simple Clustering Algorithms
Clustering with Existing Clusters
K-Means Algorithm
Example: K-Means
K-Means Algorithm
Hierarchical Algorithms
Hierarchical Agglomerative Clustering
Hierarchical Agglomerative Clustering
Clustering with CViz
Clustering with Cviz: The Data
Clustering with CViz
Clustering with CViz
Clustering with CViz
What is Memory-Based Reasoning?
What is Memory-Based Reasoning?
Basic Issues in Applying MBR
Combination Functions
Voting Approach - Example
Combination Functions
Dealing with Numerical Values
MBR in Collaborative Filtering
Collaborative Filtering: Pros & Cons
Collaborative Filtering On the Web
Back to Web Usage Mining Process
What’s in a Typical Server Log?
Usage Data Preprocessing
Example Page View
Log Entries for Page View
Page View Representation
Problems in Identifying User Sessions and Transactions
Heuristics for Identifying User Sessions
Session Inference Example
Inferring User Transactions from Sessions
Typical Usage Mining Techniques
Filtering Patterns Based on “Interestingness”
Domain versus Mined Knowledge
E-Commerce Events
Product-Oriented Events
E-Commerce vs. Usage Data
|