文档介绍：Data Mining: Concepts and T ec hniques
Jia w ei Han and Mic heline Kam ber
Simon F raser Univ ersit y
Note: This man uscript is based on a ing b o ok b y Jia w ei Han
c
and Mic heline Kam b er,
2000 c Morgan Kaufmann Publishers. All
righ ts reserv ed.
Preface
Our capabilities of b oth generating and collecting data ha v e b een increasing rapidly in the last sev eral decades.
Con tributing factors include the widespread use of bar co des for mercial pro ducts, puterization
of man y business, scien ti c and go v ernmen t transactions and managemen ts, and adv ances in data collection to ols
ranging from scanned texture and image platforms, to on-line instrumen tation in man ufacturing and shopping, and to
satellite remote sensing systems. In addition, p opular use of the W orld Wide W eb as a global information system has
o o ded us with a tremendous amoun t of data and information. This explosiv e gro wth in stored data has generated
an urgen t need for new tec hniques and automated to ols that can in telligen tly assist us in transforming the v ast
amoun ts of data in to useful information and kno wledge.
This b o ok explores the concepts and tec hniques of data mining , a promising and
ourishing fron tier in database
systems and new database applications. Data mining, also p opularly referred to as know le dge disc overy in datab ases
KDD , is the automated or con v enien t extraction of patterns represen ting kno wledge implicitly stored in large
databases, data w arehouses, and other massiv e information rep ositories.
Data mining is a m ultidisciplinary eld, dra wing w ork from areas including database tec hnology , arti cial in-
telligence, mac hine learning, w orks, statistics, pattern recognition, kno wledge based systems, kno wledge
acquisition, information retriev al, high p puting, and data visualization. W e presen t the material in
this b o ok from a datab ase p ersp e ctive . That is, w e fo cus on issues relatin