下載/瀏覽

Download Report

Transcript 下載/瀏覽

RESEARCH ON DATA MINING IN EBUSINESS WEBSITES
報告人:謝侃呈
ABSTRACT

Data mining is thus used to analyze customers’
buying behaviors and consuming habits to
better understand their needs.

present an architecture that consists of three
main components:
 Business
Data Definition,
 Customer Interaction
 Data Analysis
ABSTRACT

data mining in Ebusiness websites, including :

Characterization andDiscrimination

Association Analysis,

Cluster Analysis
INTRODUCTION

Finding and retaining customers is becoming
the major critical success factor.

mining a mountain for a vein of valuable ore
and searching for valuable business
information in a large database.
DATA SOURCE OF DATA MINING IN E-BUSINESS
WEBSITES
Server Data
 Server logs
 Cookie logs
 Error logs
 Query data
 Customer Register Information

ARCHITECTURE OF DATA MINING IN E-BUSINESS
WEBSITES
MAIN METHODS OF DATA MINING IN E-BUSINESS
WEBSITES

Characterizations and Discrimination

Association Analysis

Cluster Analysis
CHARACTERIZATIONS AND DISCRIMINATION

data mining system in amazon.com should be
able to produce a description summarizing the
characteristics of customers

The result could be a general profile of the
customers, such as they are 35-50 years old,
employed, and have excellent credit ratings.
ASSOCIATION ANALYSIS

X⇒Y, that is, “X1∧X2∧……∧Xm⇒Y1∧Y2∧……Yn”,
where Xi (for i∈ {1, …… ,m}) and Yj (for
j∈{1, ..….. , n})are attribute-value pairs.

The association rule X⇒Y is interpreted, as
“database tuples that satisfy the conditions in
X are also likely to satisfy the conditions in Y.”
ASSOCIATION ANALYSIS

age(X, “25...40”)∧degree(X, “bachelor, master,
doctor”) ∧income(X, “25000...35000”)
⇒buys(X, “laptop”) [support = 5%, confidence =
65%]

amazon.com customers, 5% are 25 to 40 years
of age with an educational degree of bachelor,
master or doctor and an income of 25000 to
35000
ASSOCIATION ANALYSIS

contains(T, “DVD player”)⇒ contains(T, “DVD
music”)[support=8%, confidence = 80%]

It means that if a transaction T contains “DVD
player”,there is an 80% probability that it
contains "DVD music" as well, and 8% of all of
the transactions contain both.
CLUSTER ANALYSIS
CLUSTER ANALYSIS
CONCLUSION

Used to analyze customer data and market
information,data mining is playing a more and
more important role in Ebusiness.

In this paper, the authors clarify the data
source,architecture and the main methods of
data mining in E-business websites.