Data Warehousing: and OLAP

1
Data Warehousing and OLAP

• Data Mining Concepts and Techniques
• J. Han, M. Kamber

2
Chapter 2 Data Warehousing and OLAP Technology
for Data Mining

• What is a data warehouse?
• A multi-dimensional data model
• Data warehouse architecture
• Data warehouse implementation
• Further development of data cube technology
• From data warehousing to data mining

3
What is Data Warehouse?

• Defined in many different ways, but not
  rigorously.
• A decision support database that is maintained
  separately from the organizations operational
  database
• Support information processing by providing a
  solid platform of consolidated, historical data
  for analysis.
• A data warehouse is a subject-oriented,
  integrated, time-variant, and nonvolatile
  collection of data in support of managements
  decision-making process.W. H. Inmon
• Data warehousing
• The process of constructing and using data
  warehouses

4
Data WarehouseSubject-Oriented

• Organized around major subjects, such as
  customer, product, sales.
• Focusing on the modeling and analysis of data for
  decision makers, not on daily operations or
  transaction processing.
• Provide a simple and concise view around
  particular subject issues by excluding data that
  are not useful in the decision support process.

5
Data WarehouseIntegrated

• Constructed by integrating multiple,
  heterogeneous data sources
• relational databases, flat files, on-line
  transaction records
• Data cleaning and data integration techniques are
  applied.
• Ensure consistency in naming conventions,
  encoding structures, attribute measures, etc.
  among different data sources
• E.g., Hotel price currency, tax, breakfast
  covered, etc.
• When data is moved to the warehouse, it is
  converted.

6
Data WarehouseTime Variant

• The time horizon for the data warehouse is
  significantly longer than that of operational
  systems.
• Operational database current value data.
• Data warehouse data provide information from a
  historical perspective (e.g., past 5-10 years)
• Every key structure in the data warehouse
• Contains an element of time, explicitly or
  implicitly
• But the key of operational data may or may not
  contain time element.

7
Data WarehouseNon-Volatile

• A physically separate store of data transformed
  from the operational environment.
• Operational update of data does not occur in the
  data warehouse environment.
• Does not require transaction processing,
  recovery, and concurrency control mechanisms
• Requires only two operations in data accessing
• initial loading of data and access of data.

8
Data Warehouse vs. Heterogeneous DBMS

• Traditional heterogeneous DB integration
• Build wrappers/mediators on top of heterogeneous
  databases
• Query driven approach
• When a query is posed to a client site, a
  meta-dictionary is used to translate the query
  into queries appropriate for individual
  heterogeneous sites involved, and the results are
  integrated into a global answer set
• Complex information filtering, compete for
  resources
• Data warehouse update-driven, high performance
• Information from heterogeneous sources is
  integrated in advance and stored in warehouses
  for direct query and analysis

9
Data Warehouse vs. Operational DBMS

• OLTP (on-line transaction processing)
• Major task of traditional relational DBMS
• Day-to-day operations purchasing, inventory,
  banking, manufacturing, payroll, registration,
  accounting, etc.
• OLAP (on-line analytical processing)
• Major task of data warehouse system
• Data analysis and decision making
• Distinct features (OLTP vs. OLAP)
• User and system orientation customer vs. market
• Data contents current, detailed vs. historical,
  consolidated
• Database design ER application vs. star
  subject
• View current, local vs. evolutionary, integrated
• Access patterns update vs. read-only but complex
  queries

10
OLTP vs. OLAP
11
Why Separate Data Warehouse?

• High performance for both systems
• DBMS tuned for OLTP access methods, indexing,
  concurrency control, recovery
• Warehousetuned for OLAP complex OLAP queries,
  multidimensional view, consolidation.
• Different functions and different data
• missing data Decision support requires
  historical data which operational DBs do not
  typically maintain
• data consolidation DS requires consolidation
  (aggregation, summarization) of data from
  heterogeneous sources
• data quality different sources typically use
  inconsistent data representations, codes and
  formats which have to be reconciled

12
Chapter 2 Data Warehousing and OLAP Technology
for Data Mining

• What is a data warehouse?
• A multi-dimensional data model
• Data warehouse architecture
• Data warehouse implementation
• Further development of data cube technology
• From data warehousing to data mining

13
From Tables and Spreadsheets to Data Cubes

• A data warehouse is based on a multidimensional
  data model which views data in the form of a data
  cube
• A data cube, such as sales, allows data to be
  modeled and viewed in multiple dimensions
• Dimensions
• A sale data warehouse with respect to dimension
• time, item, branch, location
• Dimension tables, such as
• item (item_name, brand, type), or
• time(day, week, month, quarter, year)
• Facts numerical measures
• dollars_sold sales amount in dollars
• units_sold number of units sold
• Fact table contains measures (such as
  dollars_sold) and keys to each of the related
  dimension tables

14
A 2-D data cube A table or spreadsheet for
sales from AllElectronics
AllElectronics sales data for items sold per
quarter in the city of Istanbul
Time dimension organized in quarters item
dimension by types of items sold The fact or
measure is dollar_sold
15
A three dimensional cube
Dimensions time, item, location for the cities
16
A Data Cube Representation of the same data

• Sales volume as a function of product, month, and
  region

Dimensions Product, Location, Time Hierarchical
summarization paths
izmir
locations
ankara
istabbul
items
phon
comp
Quarters
Q1
17
A 4-D Cube as a series of 3-D cubes
Dimensions item,time,location,supplier
From Supplier A
From Supplier B
From Supplier C
locations
items
Quarters
18
n-Dimensional Cube

• Any n-D data as a series of (n-1)-D cubes
• In data warehousing literature,
• A data cube is referred to as a cuboid
• The lattice of cuboids forms a data cube.
• The cuboid holding the lowest level of
  summarization is called a base cuboid.
• the 4-D cuboid is the base cuboid for the given
  four dimensions
• The top most 0-D cuboid, which holds the
  highest-level of summarization, is called the
  apex cuboid.
• Here this is the total sales, or dollars_sold
  summarized over all four dimensions
• typically denoted by all

19
Cube A Lattice of Cuboids
all
0-D(apex) cuboid
time
item
location
supplier
1-D cuboids
time,item
time,location
item,location
location,supplier
2-D cuboids
time,supplier
item,supplier
time,location,supplier
time,item,location
3-D cuboids
item,location,supplier
time,item,supplier
4-D(base) cuboid
time, item, location, supplier
20
Conceptual Modeling of Data Warehouses

• Modeling data warehouses dimensions measures
• Star schema A fact table in the middle connected
  to a set of dimension tables
• Snowflake schema A refinement of star schema
  where some dimensional hierarchy is normalized
  into a set of smaller dimension tables, forming a
  shape similar to snowflake
• Fact constellations Multiple fact tables share
  dimension tables, viewed as a collection of
  stars, therefore called galaxy schema or fact
  constellation

21
Example of Star Schema

Sales Fact Table
time_key
item_key
branch_key
location_key
units_sold
dollars_sold
avg_sales
Measures
22
(No Transcript)
23
Example of Snowflake Schema
Sales Fact Table
time_key
item_key
branch_key
location_key
units_sold
dollars_sold
avg_sales
Measures
24
A unnormalized City Table
City province and Country is repeated For every
steet in Istanbul
A Normalized City Table
Unnecessary repitations of province and country
are eliminated Memory gain but complex queries
25
Example of Fact Constellation
Shipping Fact Table
time_key
Sales Fact Table
item_key
time_key
shipper_key
item_key
from_location
branch_key
to_location
location_key
dollars_cost
units_sold
units_shipped
dollars_sold
avg_sales
Measures
26
Measures Three Categories Based on Aggreate
Functions Used

• A multidimensional point in the data cube space
• dimension-value pairs
• (timeQ1,locationIstanbul,itemcomputer)
• A data cube measure is a numerical function that
  can be evaluated at each point in the data cube
  space
• computed for a given point by aggregating the
  data corresponding to the respective
  dimension-value pairs defining the given point

27
Measures Three Categories

• distributive
• Suppose the data set D is partitioned into n sets
  Di i 1,..n.
• Then the computation of the function f on each
  partition derives one aggregate value
• Ai f(Di) i 1,..,n,
• If the result derived by applying the function to
  n aggregate values is the same as that derived by
  applying the function on all the data without
  partitioning, i.e., f(D)f(A1,A2,..,An)
• Then the measure is distributive and the function
  can be computed in a distributed manner
• E.g., count(), sum(), min(), max().

28
Example Sum()

• Data set D1,3,6,8,9
• Sum(D) 27
• Partition the set into D1 end D2 as
• D11,3,6), D28.9
• Sum(D1) 10, Sum(D2) 17
• Sum(sum(D1),sum(D2)) sum(10,17) 27
  sum(D)

29
Measures Three Categories

• algebraic if it can be computed by an algebraic
  function with M arguments (where M is a bounded
  integer), each of which is obtained by applying a
  distributive aggregate function.
• E.g., avg(), min_N(), standard_deviation().
• E.g.,avg() sum()/count()
• both sum() and count() are distributive agg.
  Functions
• Show that
• min_N(), standard_deviation().

30
Measures Three Categories

• holistic if there is no constant bound on the
  storage size needed to describe a subaggregate.
• There is not an algebric function with M
  arguments(M being bounded) that characterizes
  the computation
• E.g., median(), mode(), rank().

31
Example

• The relational database scheme for AllElectronics
  Co.
• time(time_key,day,day_of_week,month,quarter,year)
• item(item_key,item_name,brand,type,supplier_type)
• branch(branch_key,branch_name,branch_type)
• location(location_key,street,city,province_or_stat
  e,country)
• sales(time_key,item_key,branch_key,location_key,nu
  mber_of_units_sold,price)

32
Example cont.

• Select s.time_key,s.item_key,s.branch_key,s.locati
  on_key,
• sum(s.number_of_units_solds.price),
• sum(s.number_of_units_sold)
• from time t, item i,branch b,location l,sales s,
• where s.time_keyt.time_key and
  s.item_keyi.item_key and s.branch_keyb.branch_ke
  y and s.location_keyl.location_key
• group by s.time_key, s.item_key, s.branch_key,
  s.location_key

33
Example Cont.

• The cube created is the base cuboid of the
  sales_star datacube
• it contains all of the dimensions
• granularity of each is at the join key level
• by changing the group by clauses we may generate
  other cuboids for the sales-star data cube
• E.g.,
• group by t.month sum up the measures of each
  group by month
• removing group by s.branch_key generate a
  higher-level cuboid
• removing all group bys total sum of dollars sold
  and total count of units_sold
• zero-dimensional cuboid is apex cuboid

34
Time by month

• Select t.year,t.month,s.item_key,s.branch_key,s.lo
  cation_key,sum(s.number_of_units_solds.price),sum
  (s.number_of_units_sold)
• from time t, item i,branch b,location l,sales s,
• where s.time_keyt.time_key and
• s.item_keyi.item_key and
• s.branch_keyb.branch_key and
• s.location_keyl.location_key
• group by t.year, t.month, s.item_key,
  s.branch_key,s.location_key

35
A three dimensional cuboid

• Select s.time_key,s.item_key, s.location_key,
• sum(s.number_of_units_solds.price),
• sum(s.number_of_units_sold)
• from time t, item i,branch b,location l,sales s,
• where s.time_keyt.time_key and
• s.item_keyi.item_key and
• s.branch_keyb.branch_key and
• s.location_keyl.location_key
• group by s.time_key, s.item_key,s.location_key

36
Concept hierarchies

• Defines a sequence of mappings from a set of
  low-level concepts to high-level more general
  concepts
• E.g., dimension location is described by
• number,street,city,province_or_state,zipcode and
  country
• are related by a total order, forming a concept
  hierarchy
• streetltcityltprovince_or_stateltcountry
• The attributes of a dimension may be organized in
  a partial order, forming a lattice
• day,week,month,quarter, year
• dayltmonthltquarter,weekltyear

37
A Concept Hierarchy Dimension (location)
all
all
Europe
North_America
...
region
Mexico
Canada
Spain
Germany
...
...
country
Vancouver
...
...
Toronto
Frankfurt
city
M. Wind
L. Chan
...
office
38
Partially ordered co

• The attributes of a dimension may be organized in
  a partial order, forming a lattice
• day,week,month,quarter, year
• dayltmonthltquarter,weekltyear
• predefined in the data mining system
• time
• fiscal year starting on April 1
• academic year starting on September 1

39
Multidimensional Data

• Sales volume as a function of product, month, and
  region

Dimensions Product, Location, Time Hierarchical
summarization paths
Region
Industry Region Year Category
Country Quarter Product City Month
Week Office Day
Product
A partially ordered hierarchy
Month
40
Cuboids Corresponding to the Cube
all
0-D(apex) cuboid
location
item
time
1-D cuboids
item,time
item,location
time, location
2-D cuboids
3-D(base) cuboid
item, time, location
41
Set grouping hierarchy

• Set-grouping hierarchy
• discretizing or grouping values for a given
  dimension or attribute
• Ex price
• There may be more than one concept hierarchy for
  a given attribute or dimension based on
  different user viewpoints
• price by defining ranges for
• inexpensive, moderately_priced,expensive

42
How defined?

• provided by manually by
• system users
• domain experts
• knowledge engineers or
• automatically generated based on statistical
  analysis of the data distribution

43
Typical OLAP Operations

• In the multidimensional model data are organized
  into multiple dimensions
• each dimension contains multiple levels of
  abstraction defined by concept hierarchies
• This organization provides users with the
  flexibility to view data from different
  perspectives

44
Example

• Refer to figure 2.10 in Hans book
• data cube for AllElls sales
• dimensions location,time,item
• location -- city
• time -- quarters
• item -- item types
• measure displayed is dollars-sold

45
Cuboids Corresponding to the Cube
all
0-D(apex) cuboid
location
item
time
1-D cuboids
item,time
item,location
time, location
2-D cuboids
3-D(base) cuboid
item, time, location
46
Roll-up (drill-up)

• Climbing up a concept hierarchy for a dimension
  or
• by dimension reduction
• Exroll-up operation aggregates data by ascending
  the location hierarchy
• from the level of city
• to the level of country
• rather than grouping the data by city,the cubes
• groups the data by country

47
By a drill up opperation examine sales By country
rather than city level
roll up
48

• when performed by dimension reduction
• one or more dimensions are removed from the cube
• Ex a sales cube with location and time
• roll-up may remove the time dimension
• aggregation of total sales by location
• rather than by location and by time

Two dimensional cuboid
One dim. cuboid
49
Drill-down (roll-down)

• reverse of roll-up
• navigates from less detailed data to more
  detailed data
• from higher level summary to lower level summary
  or detailed data, or
• stepping down a concept hierarchy for a dimension
• introducing new dimensions
• Ex drill-down for time
• dayltmonthltquarterltyear
• form the level of quarter
• to the more detailed level of month
• Adding a new dimension to the data

50
Drill down
51
Slice and dice

• Slice a selection on one dimension of the cube
• resulting in subcube
• Ex sale data are selected for dimension time
  using time Q1
• dice defines a subcube by performing a selection
  on two or more dimensions
• Ex a dice opp. Based on
• locationtoronto or vencover and
• time Q1 or Q2 and
• item home entertainment or computer

52
slice
dice
53
Pivot (rotate)

• Visualization opp. Rotates the data axes in view
  to provide an alternative presentation of data
• Exitem and location axes in a 2-D slice are
  rotated
• or transforming a 3-D cube into a series of 2-D
  planes

54
(No Transcript)
55
(No Transcript)
56
Other OLAP operations

• drill across involving (across) more than one
  fact table
• drill through through the bottom level of the
  cube to its back-end relational tables (using
  SQL)
• ranking the top N or bottom N items in lists
• moving averages
• growth rates
• interests

57
Summary

• Data warehouse
• A subject-oriented, integrated, time-variant, and
  nonvolatile collection of data in support of
  managements decision-making process
• A multi-dimensional model of a data warehouse
• Star schema, snowflake schema, fact
  constellations
• A data cube consists of dimensions measures
• OLAP operations drilling, rolling, slicing,
  dicing and pivoting

58
Data Mining

• Organizes and employs information and knowledge
  from databases
• Statistical, mathematical, artificial
  intelligence, and machine-learning techniques
• Automatic and fast
• Tools look for patterns
• Simple models
• Intermediate models
• Complex Models

59
Data Mining

• Data mining application classes of problems
• Classification
• Clustering
• Association
• Sequencing
• Regression
• Forecasting
• Others
• Hypothesis or discovery driven
• Iterative
• Scalable

60
Tools and Techniques

• Data mining
• Statistical methods
• Decision trees
• Case based reasoning
• Neural computing
• Intelligent agents
• Genetic algorithms
• Text Mining
• Hidden content
• Group by themes
• Determine relationships