Data Warehouse , Data Cube Computation

1 1
MSCIT 5210/MSCBD 5002:
Knowledge Discovery and Data
Mining
Acknowledgement: Slides modified by Dr. Lei Chen based on
the slides provided by Jiawei Han, Micheline Kamber, and
Jian Pei
©2012 Han, Kamber & Pei. All rights reserved.

2
Chapter 4: Data Warehousing, On-line Analytical
Processing and Data Cube
 Data Warehouse: Basic Concepts
 Data Warehouse Modeling: Data Cube and OLAP
 Data Cube Computation: Preliminary Concepts
 Data Cube Computation Methods
 Summary

COMP3311 Fall 2011 CSE, HKUST Slide 3
Aspects of SQL
 Most common Query Language – used in all
commercial systems
• Discussion is based on the SQL92 Standard. Commercial
products have different features of SQL, but the basic structure
is the same
 Data Manipulation Language
 Data Definition Language
 Constraint Specification
 Embedded SQL
 Transaction Management
 Security Management .....

Basic Structure
• SQL is based on set and relational operations with certain
modifications and enhancements
• A typical SQL query has the form:
select A1, A2, …, An
from R1, R2, …, Rm
where P
- Ai represent attributes
- Ri represent relations
- P is a predicate.
• This query is equivalent to the relational algebra expression:
A1, A2, …, An(P(R1  R2  …  Rm))
• The result of an SQL query is a relation (but may contain
duplicates). SQL statements can be nested.

Projection
• The select clause corresponds to the projection operation of the relational
algebra. It is used to list the attributes desired in the result of a query.
• Find the names of all branches in the loan relation
select branch-name
from loan
Equivalent to: branch-name(loan)
• An asterisk in the select clause denotes “all attributes”
select *
from loan
• Note: for our examples we use the tables:
– Branch (branch-name, branch-city, assets)
– Customer (customer-name, customer-street, customer-city)
– Loan (loan-number, amount, branch-name)
– Account (account-number, balance, branch-name)
– Borrower (customer-name, loan-number)
– Depositor (customer-name, account-number)

Duplicate Removal
• SQL allows duplicates in relations as well as in query
results. Use select distinct to force the elimination of
duplicates.
Find the names of all branches in the loan relation,
and remove duplicates
select distinct branch-name
from loan
• The keyword all specifies that duplicates are not
removed.
select all branch-name
from loan
force the DBMS to
remove duplicates
force the DBMS not
to remove duplicates

Arithmetic Operations on Retrieved Results
• The select clause can contain arithmetic expressions
involving the operators,,, and , and operating on
constants or attributes of tuples.
• The query:
select branch-name, loan-number, amount * 100
from loan
would return a relation which is the same as the loan
relations, except that the attribute amount is
multiplied by 100

The where Clause
• The where clause specifies conditions that tuples in the relations
in the from clause must satisfy.
• Find all loan numbers for loans made at the Perryridge branch
with loan amounts greater than $1200.
select loan-number
from loan
where branch-name=“Perryridge” and amount >1200
• SQL allows logical connectives and, or, and not. Arithmetic
expressions can be used in the comparison operators.
• Note: attributes used in a query (both select and where parts)
must be defined in the relations in the from clause.

The where Clause (Cont.)
• SQL includes the between operator for convenience.
• Find the loan number of those loans with loan amounts between
$90,000 and $100,000 (that is,  $90,000 and  $100,000)
select loan-number
from loan
where amount between 90000 and
100000

The from Clause
• The from clause corresponds to the Cartesian product
operation of the relational algebra.
• Find the Cartesian product borrower  loan
select *
from borrower, loan
It is rarely used without a where clause.
• Find the name and loan number of all customers
having a loan at the Perryridge branch.
select distinct customer-name, borrower.loan-number
from borrower, loan
where borrower.loan-number = loan.loan-number and
branch-name = “Perryridge”

Aggregate Functions
• Operate on a column of a relation, and return a value
avg: average value
min: minimum value
max: maximum value
sum: sum of values
count: number of values
• Note: for our examples we use the tables:
– Branch (branch-name, branch-city, assets)
– Customer (customer-name, customer-street, customer-city)
– Loan (loan-number, amount, branch-name)
– Account (account-number, balance, branch-name)
– Borrower (customer-name, loan-number)
– Depositor (customer-name, account-number)

Aggregate Function Computation
• Find the average account balance at the Perryridge branch.
select avg(balance)
from account
where branch-name=“Perryridge”
account select balance
from account
where branch-name
=“Perryridge”
Avg()
120,000
Balances of Perryridge accounts

Examples of Aggregate Functions
• Find the numbers of tuples in the customer relation.
select count(*)
from customer
– remember * stands for all attributes
– Same as:
select count(customer-city)
from customer
– Different from:
select count(distinct customer-city)
from customer
– Because customer-city is not a key

Group by
• Find the number of accounts for each branch.
select branch-name, count(account-number)
from account
group by branch-name
• For each group of tuples with the same branch-name, apply
aggregate function count and distinct to account-number
branch-name count-account-no
Perryridge
Brighton
Redwood
2
2
1
branch-name account-number balance
Perryridge
Brighton
Perryridge
Brighton
Redwood
a-102
a- 217
a-201
a-215
a-222
400
750
900
750
700
branch-name account-number balance
Perryridge
Perryridge
Brighton
Brighton
Redwood
a-102
a-201
a-217
a-215
a-222
400
900
750
750
700
account table

• Attributes in select clause outside of aggregate functions must
appear in group by list, why?
select branch-name, balance, count( distinct account-number)
from account
branch-
name
account-
number
balance
Perryridge
Perryridge
Brighton
Brighton
Redwood
a-102
a-201
a-217
a-215
a-222
400
900
750
750
700
select … from account
group by branch-name, balance
OR
select branch-name, sum(balance), count(…)
from account group by branch-name
Group by Attributes
correct

• Find the number of depositors for each branch.
select branch-name, count( distinct customer-name)
from depositor, account
where depositor.account-number = account.account-number
• Perform Join then group by then count ( distinct () )
depositor (customer-name, account-number)
account (branch-name, account-number, balance)
Join  (customer-name, account-number, branch-name, balance)
• Group by and aggregate functions apply to the Join result
Group by with Join

Group by Evaluation
select branch-name, customer-name
from depositor, account
where depositor.account-number
= account.account-number
branch-name cust-name
Perryridge John Wong
Perryridge Jacky Chan
Uptown John Wong
Downtown John Wong
Uptown Mary Kwan
Downtown Pat Lee
Downtown May Cheung
Uptown John Wong
Downtown John Wong
Uptown Mary Kwan
Downtown Pat Lee
Downtown May Cheung
group by
Uptown John Wong
Downtown John Wong
Uptown Mary Kwan
Downtown Pat Lee
Downtown May Cheung
distinct
count
branch-name
Perryridge
Uptown
Downtown
count
2
2
3

Having Clause
• Find the names of all branches where the average account
balance is more than $700
select branch-name, avg(balance)
from account
having avg (balance) >700
• predicates in the having clause are applied to each group after
the formation of groups
branch-
name
account-
number
balance
Perryridge
Perryridge
Brighton
Brighton
Redwood
a-102
a-201
a-217
a-215
a-222
400
900
750
750
700

Group-by
• Motivation: Group-by permits us to display aggregate results (e.g., max,
min, sum) for groups. For instance, if we have GROUP-BY X, we will get a
result for every different value of X.
• Recall that aggregate queries without group-by return just a single number.
• If we put an attribute in SELECT, the attribute must also appear in GROUP-
BY. The opposite is not true: there may be attributes in GROUP-BY that do
not appear in SELECT.
• Any condition that appears in WHERE, is applied before the formation of
groups – in other words, records that do not pass the WHERE condition are
eliminated before the formation of groups.
• Any condition that appears in HAVING refers to the groups and is applied
after the formation of the groups. The condition must involve aggregate
functions, or attributes that appear in the SELECT or GROUP-BY lines.

Query 1: Find the total number of copies in stock for each poet
SELECT poet, SUM (copies_in_stock) as sum
FROM writer
GROUP BY poet
poet sum
Douglas Livingstone 23
Mongane Wally 13
poet book copies_in_stock
Douglas Livingstone The Skull 21
Douglas Livingstone A Littoral Zone 2
Mongane Wally Tstetlo 3
Mongane Wally Must Weep 8
Mongane Wally A Tough Tale 2

Query 2: For each poet, find the max, min, avg and total number of copies in
stock
SELECT poet, MAX(copies_in_stock) AS max, MIN(copies_in_stock) AS
min, AVG(copies_in_stock) AS avg, SUM(copies_in_stock) AS sum
FROM writer
GROUP BY poet
poet max min avg sum
Douglas Livingstone 21 2 11.5 23
Mongane Wally 8 2 4.33 13

Query 3: For each poet, find the max, min, avg and total number of copies in stock – take into
account only books that have > 5 copies in stock
SELECT poet, MAX(copies_in_stock) AS max, MIN(copies_in_stock) AS
min, AVG(copies_in_stock) AS avg, SUM(copies_in_stock) AS sum
FROM writer
WHERE copies_in_stock > 5
GROUP BY poet
poet max min avg sum
Douglas Livingstone 21 21 21 21
Mongane Wally 8 8 8 8

Query 4: Find the total number of copies in stock for each poet who has a
total of more than 20 copies in stock
SELECT poet, SUM(copies_in_stock) AS sum
FROM writer
GROUP BY poet
HAVING sum>20
poet sum
Mongane Wally 13

Query 5: Find the total number of copies in stock for each poet who has a total of more than
20 copies in stock – take into account only books that have more than 5 copies in stock
SELECT poet, SUM(copies_in_stock) AS sum
FROM writer
WHERE copies_in_stock > 5
GROUP BY poet
HAVING sum>20
poet sum
Mongane Wally 8

Query 6: Find the total number of copies in stock for each poet whose name
starts with any letter after “E”
SELECT poet, SUM(copies_in_stock)
as sum
FROM writer
WHERE poet > “E”
GROUP BY poet
poet sum
Mongane Wally 13
as sum
FROM writer
GROUP BY poet
HAVING poet > “E”

Query 7: Find the total number of copies in stock for each poet who has more
than 2 books
poet sum
Mongane Wally 13
as sum
FROM writer
GROUP BY poet
HAVING count(*)>2

DATA WAREHOUSES and OLAP
 On-Line Transaction Processing (OLTP) Systems
manipulate operational data, necessary for day-to-
day operations. Most existing database systems
belong this category.
 On-Line Analytical Processing (OLAP) Systems
support specific types of queries (based on group-
bys and aggregation operators) useful for decision
making.
 Data Mining tools discover interesting patterns in
the data

Why OLTP is not sufficient for Decision Making
Lets say that Welcome supermarket uses a relational database to keep
track of sales in all of stores simultaneously
SALES table
product
id
store
id
quantity
sold
date/time of
sale
567 17 1
1997-10-22
09:35:14
219
16 4 1997-10-22
09:35:14
219
17 1 1997-10-22
09:35:17
...

COMP231 Spring 2009
CSE, HKUST Slide 29
Example (cont.)
PRODUCTS table
Prod.
id
product
name
product
category
Manufac
t. id
567 Colgate Gel
Pump 6.4 oz.
toothpast
e
68
219 Diet Coke 12
oz. can
soda 5
...
STORES table
stor
e id
cit
y id
store
location
phone
number
16 34
510 Main
Street
415-555-
1212
17 58
13 Maple
Avenue
914-555-
1212
CITIES table
id name state Popul.
34
San
Francisco
California
700,000
58 East Fishkill
New York 30,000

COMP231 Spring 2009
CSE, HKUST Slide 30
Example (cont.)
 An executive, asks "I noticed that there was a Colgate promotion recently,
directed at people who live in small towns. How much Colgate toothpaste did
we sell in those towns yesterday? And how much on the same day a month
ago?"
select sum(sales.quantity_sold)
from sales, products, stores, cities
where products.manufacturer_id = 68 -- restrict to Colgate-
and products.product_category = 'toothpaste‘
and cities.population < 40000
and sales.datetime_of_sale::date = 'yesterday'::date
and sales.product_id = products.product_id
and sales.store_id = stores.store_id
and stores.city_id = cities.city_id

COMP231 Spring 2009
CSE, HKUST Slide 31
Example (cont.)
 You have to do a 4-way JOIN of some large tables.
Moreover, these tables are being updated as the
query is executed.
 Need for a separate RDBMS installation (i.e., a
Data Warehouse) to support queries like the
previous one.
 The Warehouse can be tailor-made for specific
types of queries: if you know that the toothpaste
query will occur every day then you can
denormalize the data model.

COMP231 Spring 2009
CSE, HKUST Slide 32
Example (cont.)
 Suppose Welcome acquires ParknShop which is using a different set
of OLTP data models and a different brand of RDBMS to support
them. But you want to run the toothpaste queries for both divisions.
 Solution: Also copy data from the ParknShop Database into the
Welcome Data Warehouse (data integration problems).
 One of the more important functions of a data warehouse in a
company that has disparate computing systems is to provide a view
for management as though the company were in fact integrated.

COMP231 Spring 2009
CSE, HKUST Slide 33
Motivation
 In most organizations, data about specific parts
of business is there -- lots and lots of data,
somewhere, in some form.
 Data is available but not information -- and not
the right information at the right time.
 To bring together information from multiple
sources as to provide a consistent database
source for decision support queries.
 To off-load decision support applications from the
on-line transaction system.

34
What is a Data Warehouse?
 Defined in many different ways, but not rigorously.
 A decision support database that is maintained separately from
the organization’s 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 management’s
decision-making process.”—W. H. Inmon
 Data warehousing:
 The process of constructing and using data warehouses

35
Data Warehouse—Subject-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

36
Data Warehouse—Integrated
 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.

37
Data Warehouse—Time 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”

38
Data Warehouse—Nonvolatile
 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

39
OLTP vs. OLAP
OLTP OLAP
users clerk, IT professional knowledge worker
function day to day operations decision support
DB design application-oriented subject-oriented
data current, up-to-date
detailed, flat relational
isolated
historical,
summarized, multidimensional
integrated, consolidated
usage repetitive ad-hoc
access read/write
index/hash on prim. key
lots of scans
unit of work short, simple transaction complex query
# records accessed tens millions
#users thousands hundreds
DB size 100MB-GB 100GB-TB
metric transaction throughput query throughput, response

40
Why a Separate Data Warehouse?
 High performance for both systems
 DBMS— tuned for OLTP: access methods, indexing, concurrency
control, recovery
 Warehouse—tuned 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
 Note: There are more and more systems which perform OLAP
analysis directly on relational databases

41
Data Warehouse: A Multi-Tiered Architecture
Data
Warehouse
Extract
Transform
Load
Refresh
OLAP Engine
Analysis
Query
Reports
Data mining
Monitor
&
Integrator
Metadata
Data Sources Front-End Tools
Serve
Data Marts
Operational
DBs
Other
sources
Data Storage
OLAP Server

42
Chapter 4: Data Warehousing and On-line
Analytical Processing
 Summary

43
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
 Dimension tables, such as item (item_name, brand, type), or
time(day, week, month, quarter, year)
 Fact table contains measures (such as dollars_sold) and keys
to each of the related dimension tables
 In data warehousing literature, an n-D base cube is called a base
cuboid. The top most 0-D cuboid, which holds the highest-level of
summarization, is called the apex cuboid. The lattice of cuboids
forms a data cube.

Data cube
 A data cube, such as sales, allows data to be
modeled and viewed in multiple dimensions
 Suppose ALLELETRONICS create a sales data
warehouse with respect to dimensions
 Time
 Item
 Location

Data cube
 A data cube, such as sales, allows data to be modeled
and viewed in multiple dimensions
 Suppose ALLELETRONICS create a sales data warehouse
with respect to dimensions
 Time
 Item
 Location
 Supplier

48
Cube: A Lattice of Cuboids
time,item
time,item,location
time, item, location, supplier
all
time item location supplier
time,location
time,supplier
item,location
item,supplier
location,supplier
time,item,supplier
time,location,supplier
item,location,supplier
0-D (apex) cuboid
1-D cuboids
2-D cuboids
3-D cuboids
4-D (base) cuboid

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

50
Star Schema: An Example
time_key
day
day_of_the_week
month
quarter
year
time
location_key
street
city
state_or_province
country
location
Sales Fact Table
time_key
item_key
branch_key
location_key
units_sold
dollars_sold
avg_sales
Measures
item_key
item_name
brand
type
supplier_type
item
branch_key
branch_name
branch_type
branch

Snowflake Schema: An Example
time_key
day
day_of_the_week
month
quarter
year
time
location_key
street
city_key
location
Sales Fact Table
time_key
item_key
branch_key
location_key
units_sold
dollars_sold
avg_sales
Measures
item_key
item_name
brand
type
supplier_key
item
branch_key
branch_name
branch_type
branch
supplier_key
supplier_type
supplier
city_key
city
state_or_province
country
city

Fact Constellation: An
Example
time_key
day
day_of_the_week
month
quarter
year
time
location_key
street
city
province_or_state
country
location
Sales Fact Table
time_key
item_key
branch_key
location_key
units_sold
dollars_sold
avg_sales
Measures
item_key
item_name
brand
type
supplier_type
item
branch_key
branch_name
branch_type
branch
Shipping Fact Table
time_key
item_key
shipper_key
from_location
to_location
dollars_cost
units_shipped
shipper_key
shipper_name
location_key
shipper_type
shipper

Concept Hierarchies
 A Concept Hierarchy defines a sequence of
mappings from a set of low-level concepts to
high-level
 Consider a concept hierarchy for the dimension
“Location”

A Concept Hierarchy for a Dimension (location)
all
Europe North_America
Mexico
Canada
Spain
Germany
Vancouver
M. Wind
L. Chan
...
...
...
... ...
...
all
region
office
country
Toronto
Frankfurt
city

Concept Hierarchies
 Many concept hierarchies are implicit within the
database system

Concept Hierarchies
 Concept hierarchies may also be defined by grouping
values for a given dimension or attribute, resulting in a
set-grouping hierarchy

OLAP Operation
 So, how are concept hierarchies useful in OLAP?
 In the multidimensional model, data are
organized into multiple dimensions,
 And each dimension contains multiple levels of
abstraction defined by concept hierarchies

58
Data Cube Measures: Three Categories
 Distributive: 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
 E.g., count(), sum(), min(), max()
 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()
 Holistic: if there is no constant bound on the storage size
needed to describe a subaggregate.
 E.g., median(), mode(), rank()

59
Multidimensional Data
 Sales volume as a function of product, month,
and region
Product
Month
Dimensions: Product, Location, Time
Hierarchical summarization paths
Industry Region Year
Category Country Quarter
Product City Month Week
Office Day

60
A Sample Data Cube
Total annual sales
of TVs in U.S.A.
Date
Country
sum
sum
TV
VCR
PC
1Qtr 2Qtr 3Qtr 4Qtr
U.S.A
Canada
Mexico
sum

61
Cuboids Corresponding to the Cube
all
product date country
product,date product,country date, country
product, date, country
0-D (apex) cuboid
1-D cuboids
2-D cuboids
3-D (base) cuboid

62
Typical OLAP Operations
 Roll up (drill-up): summarize data
 by climbing up hierarchy or by dimension reduction
 Drill down (roll down): reverse of roll-up
 from higher level summary to lower level summary or
detailed data, or introducing new dimensions
 Slice and dice: project and select
 Pivot (rotate):
 reorient the cube, visualization, 3D to series of 2D planes
 Other 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)

63
Fig. 3.10 Typical OLAP
Operations

64
Cube Operators for Roll-up
day 2 s1 s2 s3
p1 44 4
p2 s1 s2 s3
p1 12 50
p2 11 8
day 1
s1 s2 s3
p1 56 4 50
p2 11 8
s1 s2 s3
sum 67 12 50
sum
p1 110
p2 19
129
. . .
sale(s1,*,*)
sale(*,*,*)
sale(s2,p2,*)

65
s1 s2 s3 *
p1 56 4 50 110
p2 11 8 19
* 67 12 50 129
Extended Cube
day 2 s1 s2 s3 *
p1 44 4 48
p2
* 44 4 48
s1 s2 s3 *
p1 12 50 62
p2 11 8 19
* 23 8 50 81
day 1
*
sale(*,p2,*)

CS 336 66
Aggregation Using Hierarchies
region A region B
p1 56 54
p2 11 8
store
region
country
(store s1 in Region A;
stores s2, s3 in Region B)
day 2 s1 s2 s3
p1 44 4
p2 s1 s2 s3
p1 12 50
p2 11 8
day 1

CS 336 67
Slicing
day 2 s1 s2 s3
p1 44 4
p2 s1 s2 s3
p1 12 50
p2 11 8
day 1
s1 s2 s3
p1 12 50
p2 11 8
TIME = day 1

68
Products
d1 d2
Store s1 Electronics $5.2
Toys $1.9
Clothing $2.3
Cosmetics $1.1
Store s2 Electronics $8.9
Toys $0.75
Clothing $4.6
Cosmetics $1.5
Products
Store s1 Store s2
Store s1 Electronics $5.2 $8.9
Toys $1.9 $0.75
Clothing $2.3 $4.6
Cosmetics $1.1 $1.5
Store s2 Electronics
Toys
Clothing
($ millions)
d1
Sales
($ millions)
Time
Sales
Slicing &
Pivoting

69
69
Chapter 5: Data Cube Technology

70
70
Data Cube: A Lattice of Cuboids
time,item
time,item,location
time, item, location, supplierc
all
time,location
time,supplier
item,location
item,supplier
location,supplier
time,item,supplier
0-D(apex) cuboid
1-D cuboids
2-D cuboids
3-D cuboids
4-D(base) cuboid

71
Data Cube: A Lattice of Cuboids
 Base vs. aggregate cells; ancestor vs. descendant cells; parent vs. child cells
1. (9/15, milk, Urbana, Dairy_land)
2. (9/15, milk, Urbana, *)
3. (*, milk, Urbana, *)
4. (*, milk, Urbana, *)
5. (*, milk, Chicago, *)
6. (*, milk, *, *)
all
time,item
time,item,location
time, item, location, supplier
time,location
time,supplier
item,location
item,supplier
location,supplier
time,item,supplier
0-D(apex) cuboid
1-D cuboids
2-D cuboids
3-D cuboids
4-D(base) cuboid

72
72
Cube Materialization:
Full Cube vs. Iceberg Cube
 Full cube vs. iceberg cube
compute cube sales iceberg as
select month, city, customer group, count(*)
from salesInfo
cube by month, city, customer group
having count(*) >= min support
 Compute only the cuboid cells whose measure satisfies the
iceberg condition
 Only a small portion of cells may be “above the water’’ in a
sparse cube
 Ex.: Show only those cells whose count is no less than 100
iceberg
condition

Why Iceberg Cube?
 Advantages of computing iceberg cubes
 No need to save nor show those cells whose value is below the threshold
(iceberg condition)
 Efficient methods may even avoid computing the un-needed,
intermediate cells
 Avoid explosive growth
 Example: A cube with 100 dimensions
 Suppose it contains only 2 base cells: {(a1, a2, a3, …., a100), (a1, a2, b3,
…, b100)}
 How many aggregate cells if “having count >= 1”?
 Answer: 2101 ─ 4
 What about the iceberg cells, (i,e., with condition: “having count >= 2”)?
 Answer: 4

Is Iceberg Cube Good Enough? Closed Cube &
Cube Shell
 Let cube P have only 2 base cells: {(a1, a2, a3 . . . , a100):10, (a1, a2, b3, . . . ,
b100):10}
 How many cells will the iceberg cube contain if “having count(*) ≥ 10”?
 Answer: 2101 ─ 4 (still too big!)
 Close cube:
 A cell c is closed if there exists no cell d, such that d is a descendant of c,
and d has the same measure value as c
 Ex. The same cube P has only 3 closed cells:
 {(a1, a2, *, …, *): 20, (a1, a2, a3 . . . , a100): 10, (a1, a2, b3, . . . , b100):
10}
 A closed cube is a cube consisting of only closed cells
 Cube Shell: The cuboids involving only a small # of dimensions, e.g., 2
 Idea: Only compute cube shells, other dimension combinations can be
computed on the fly
Q: For (A1, A2, … A100), how many combinations to compute?

75
75
General Heuristics (Agarwal et al. VLDB’96)
 Sorting, hashing, and grouping operations are applied to the dimension
attributes in order to reorder and cluster related tuples
 Aggregates may be computed from previously computed aggregates,
rather than from the base fact table
 Smallest-child: computing a cuboid from the smallest, previously
computed cuboid
 Cache-results: caching results of a cuboid from which other
cuboids are computed to reduce disk I/Os
 Amortize-scans: computing as many as possible cuboids at the
same time to amortize disk reads
 Share-sorts: sharing sorting costs cross multiple cuboids when
sort-based method is used
 Share-partitions: sharing the partitioning cost across multiple
cuboids when hash-based algorithms are used

76
76
Chapter 5: Data Cube Technology
 Processing Advanced Queries by Exploring Data
Cube Technology
 Multidimensional Data Analysis in Cube Space
 Summary

77
77
Data Cube Computation Methods
 Multi-Way Array Aggregation
 BUC

78
78
Multi-Way Array Aggregation
 Array-based “bottom-up” algorithm
 Using multi-dimensional chunks
 No direct tuple comparisons
 Simultaneous aggregation on multiple
dimensions
 Intermediate aggregate values are re-
used for computing ancestor cuboids
 Cannot do Apriori pruning: No iceberg
optimization

79
79
Multi-way Array Aggregation for Cube
Computation (MOLAP)
 Partition arrays into chunks (a small subcube which fits in memory).
 Compressed sparse array addressing: (chunk_id, offset)
 Compute aggregates in “multiway” by visiting cube cells in the order
which minimizes the # of times to visit each cell, and reduces
memory access and storage cost.
What is the best
traversing order
to do multi-way
aggregation?
A
B
29 30 31 32
1 2 3 4
5
9
13 14 15 16
64
63
62
61
48
47
46
45
a1
a0
c3
c2
c1
c 0
b3
b2
b1
b0
a2 a3
C
B
44
28 56
40
24 52
36
20
60

80
Computation (3-D to 2-D)
all
A B
AB
ABC
AC BC
C
 The best order is
the one that
minimizes the
memory
requirement and
reduced I/Os

81
Computation (2-D to 1-D)

82
82
Multi-Way Array Aggregation for Cube
Computation (Method Summary)
 Method: the planes should be sorted and computed
according to their size in ascending order
 Idea: keep the smallest plane in the main memory,
fetch and compute only one chunk at a time for the
largest plane
 Limitation of the method: computing well only for a small
number of dimensions
 If there are a large number of dimensions, “top-down”
computation and iceberg cube computation methods
can be explored

83
83
Data Cube Computation Methods
 Multi-Way Array Aggregation
 BUC

84
84
Bottom-Up Computation (BUC)
 BUC (Beyer & Ramakrishnan,
SIGMOD’99)
 Bottom-up cube computation
(Note: top-down in our view!)
 Divides dimensions into partitions
and facilitates iceberg pruning
 If a partition does not satisfy
min_sup, its descendants can
be pruned
 If minsup = 1  compute full
CUBE!
 No simultaneous aggregation
all
A B C
AC BC
ABC ABD ACD BCD
AD BD CD
D
ABCD
AB
1 all
2 A 10 B 14 C
7 AC 11 BC
4 ABC 6 ABD 8 ACD 12 BCD
9 AD 13 BD 15 CD
16 D
5 ABCD
3 AB

85
85
BUC: Partitioning
 Usually, entire data set
can’t fit in main memory
 Sort distinct values
 partition into blocks that fit
 Continue processing
 Optimizations
 Partitioning
 External Sorting, Hashing, Counting Sort
 Ordering dimensions to encourage pruning
 Cardinality, Skew, Correlation
 Collapsing duplicates
 Can’t do holistic aggregates anymore!

86
Attribute-Based Data
Warehouse
 Problem
 Data Warehouse
 NP-hardness
 Algorithm
 Performance Study

87
Data Warehouse
Parts are bought from suppliers and then sold to customers at a sale price SP
part supplier customer SP
p1 s1 c1 4
p3 s1 c2 3
p2 s3 c1 7
… … … …
Table T

88
Data Warehouse
p1 s1 c1 4
p3 s1 c2 3
p2 s3 c1 7
… … … …
Table T
part
p1 p2 p3 p4 p5 supplier
s1
s2
s3
s4
customer
c1
c2
c3
c4
4
3
Data cube

89
Data Warehouse
e.g.,
select part, customer, SUM(SP)
from table T
group by part, customer
part customer SUM(SP)
p1 c1 4
p3 c2 3
p2 c1 7
e.g.,
select customer, SUM(SP)
from table T
group by customer
customer SUM(SP)
c1 11
c2 3
pc 3 c 2
p1 s1 c1 4
p3 s1 c2 3
p2 s3 c1 7
… … … …
Table T
AVG(SP), MAX(SP), MIN(SP), …

90
Data Warehouse
psc 6M
pc 4M ps 0.8M sc 2M
p 0.2M s 0.01M c 0.1M
none 1
p1 s1 c1 4
p3 s1 c2 3
p2 s3 c1 7
… … … …
Table T

91
Data Warehouse
psc 6M
pc 4M ps 0.8M sc 2M
p 0.2M s 0.01M c 0.1M
none 1
Suppose we materialize all views. This wastes a lot of space.
Cost for accessing pc = 4M
Cost for accessing ps = 0.8M
Cost for accessing sc = 2M
Cost for accessing p = 0.2M
Cost for accessing c = 0.1M
Cost for accessing s = 0.01M

COMP5331 92
Data Warehouse
psc 6M
pc 4M ps 0.8M sc 2M
p 0.2M s 0.01M c 0.1M
none 1
Suppose we materialize the top view only.
(not 4M)
Cost for accessing ps = 6M
(not 0.8M)
(not 2M)
Cost for accessing p = 6M
(not 0.2M)
Cost for accessing c = 6M
(not 0.1M)
Cost for accessing s = 6M
(not 0.01M)

COMP5331 93
Data Warehouse
psc 6M
pc 4M ps 0.8M sc 2M
p 0.2M s 0.01M c 0.1M
none 1
Suppose we materialize the top view and the view for “ps” only.
(still 6M)
(still 6M)
(not 6M previously)
(not 6M previously)
(still 6M)
(not 6M previously)

COMP5331 94
Data Warehouse
psc 6M
pc 4M ps 0.8M sc 2M
p 0.2M s 0.01M c 0.1M
none 1
Suppose we materialize the top view and the view for “ps” only.
(still 6M)
(still 6M)
(not 6M previously)
(not 6M previously)
(still 6M)
(not 6M previously)
Gain = 0
Gain = 5.2M
Gain = 0
Gain = 5.2M
Gain = 5.2M Gain = 0
Gain({view for “ps”, top view}, {top view}) = 5.2*3 = 15.6
Selective Materialization Problem:
We can select a set V of k views such that
Gain(V U {top view}, {top view}) is maximized.

95
Attribute-Based Data
Warehouse
 Problem
 Date Warehouse
 Algorithm

96
Greedy Algorithm
 k = number of views to be materialized
 Given
 v is a view
 S is a set of views which are selected to be
materialized
 Define the benefit of selecting v for
materialization as
 B(v, S) = Gain(S U v, S)

97
Greedy Algorithm
 S  {top view};
 For i = 1 to k do
 Select that view v not in S such that B(v, S)
is maximized;
 S  S U {v}
 Resulting S is the greedy selection

98
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
1st Choice (M) 2nd Choice (M)
pc
ps
sc
p
s
c
0 x 3 = 0
Benefit from pc =
Benefit
6M-6M = 0
k = 2

99
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit from ps =
Benefit
6M-0.8M= 5.2M
k = 2

100
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit from sc =
Benefit
6M-6M= 0
0 x 3 = 0
k = 2

101
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit from p =
Benefit
6M-0.2M = 5.8M
0 x 3 = 0
5.8 x 1= 5.8
k = 2

102
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit from s =
Benefit
6M-0.01M = 5.99M
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
k = 2

103
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit from c =
Benefit
6M-0.1M = 5.9M
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
k = 2

104
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
Benefit from pc = 6M-6M = 0
0 x 2 = 0
k = 2

105
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
Benefit from sc = 6M-6M = 0
0 x 2 = 0
0 x 2 = 0
k = 2

106
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
Benefit from p = 0.8M-0.2M= 0.6M
0 x 2 = 0
0 x 2 = 0
0.6 x 1= 0.6
k = 2

107
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
Benefit from s = 0.8M-0.01M = 0.79M
0 x 2 = 0
0 x 2 = 0
0.6 x 1= 0.6
0.79 x 1= 0.79
k = 2

108
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
Benefit from c = 6M-0.1M = 5.9M
0 x 2 = 0
0 x 2 = 0
0.6 x 1= 0.6
0.79 x 1= 0.79
5.9 x 1 = 5.9
k = 2

109
1.1 Data Cube
psc 6M
pc 6M ps 0.8M sc 6M
p 0.2M s 0.01M c 0.1M
none 1
pc
ps
sc
p
s
c
0 x 3 = 0
5.2 x 3 = 15.6
Benefit
0 x 3 = 0
5.8 x 1= 5.8
5.99 x 1= 5.99
5.9 x 1 = 5.9
0 x 2 = 0
0 x 2 = 0
0.6 x 1= 0.6
0.79 x 1= 0.79
5.9 x 1 = 5.9
Two views to be materialized are
1. ps
2. c
V = {ps, c}
Gain(V U {top view}, {top view})
= 15.6 + 5.9 = 21.5
k = 2

COMP5331 110
Performance Study
 How bad does the Greedy Algorithm
perform?

COMP5331 111
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit from b =
Benefit
200-100= 100
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
k = 2

COMP5331 112
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit from c =
Benefit
200-99 = 101
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
41 x 101= 4141
k = 2

COMP5331 113
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
41 x 101= 4141
41 x 100= 4100
k = 2

COMP5331 114
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
41 x 101= 4141
41 x 100= 4100
Benefit from b = 200-100= 100
21 x 100= 2100
k = 2

COMP5331 115
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
41 x 101= 4141
41 x 100= 4100
21 x 100= 2100
21 x 100= 2100
Greedy:
V = {b, c}
= 4141 + 2100 = 6241
k = 2

COMP5331 116
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
b
c
d
… … …
41 x 100= 4100
Benefit
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
41 x 101= 4141
41 x 100= 4100 41 x 100= 4100
Greedy:
V = {b, c}
= 4141 + 2100 = 6241
21 x 101 + 20 x 1 = 2141
Optimal:
V = {b, d}
= 4100 + 4100 = 8200
k = 2

COMP5331 117
1.1 Data Cube
a 200
b 100 c 99 d 100
p1 97
none 1
20 nodes
…
p20 97
q1 97
…
q20 97
r1 97
…
r20 97
s1 97
…
s20 97
Greedy:
V = {b, c}
= 4141 + 2100 = 6241
Optimal:
V = {b, d}
= 4100 + 4100 = 8200
Greedy
Optimal
=
6241
8200
= 0.7611
If this ratio = 1, Greedy can give an optimal solution.
If this ratio  0, Greedy may give a “bad” solution.
Does this ratio has a “lower” bound?
It is proved that this ratio is at least 0.63.
k = 2

COMP5331 118
Performance Study
 This is just an example to show that
this greedy algorithm can perform badly.
 A complete proof of the lower bound
can be found in the paper.

119
Summary
 Data warehousing: A multi-dimensional model of a data warehouse
 A data cube consists of dimensions & measures
 Star schema, snowflake schema, fact constellations
 OLAP operations: drilling, rolling, slicing, dicing and pivoting
 Data Warehouse Architecture, Design, and Usage
 Multi-tiered architecture
 Business analysis design framework
 Information processing, analytical processing, data mining, OLAM (Online
Analytical Mining)
 Implementation: Efficient computation of data cubes
 Partial vs. full vs. no materialization
 Indexing OALP data: Bitmap index and join index
 OLAP query processing
 OLAP servers: ROLAP, MOLAP, HOLAP
 Data generalization: Attribute-oriented induction

120
References (I)
 S. Agarwal, R. Agrawal, P. M. Deshpande, A. Gupta, J. F. Naughton, R. Ramakrishnan,
and S. Sarawagi. On the computation of multidimensional aggregates. VLDB’96
 D. Agrawal, A. E. Abbadi, A. Singh, and T. Yurek. Efficient view maintenance in data
warehouses. SIGMOD’97
 R. Agrawal, A. Gupta, and S. Sarawagi. Modeling multidimensional databases. ICDE’97
 S. Chaudhuri and U. Dayal. An overview of data warehousing and OLAP technology.
ACM SIGMOD Record, 26:65-74, 1997
 E. F. Codd, S. B. Codd, and C. T. Salley. Beyond decision support. Computer World, 27,
July 1993.
 J. Gray, et al. Data cube: A relational aggregation operator generalizing group-by,
cross-tab and sub-totals. Data Mining and Knowledge Discovery, 1:29-54, 1997.
 A. Gupta and I. S. Mumick. Materialized Views: Techniques, Implementations, and
Applications. MIT Press, 1999.
 J. Han. Towards on-line analytical mining in large databases. ACM SIGMOD Record,
27:97-107, 1998.
 V. Harinarayan, A. Rajaraman, and J. D. Ullman. Implementing data cubes efficiently.
SIGMOD’96

121
References (II)
 C. Imhoff, N. Galemmo, and J. G. Geiger. Mastering Data Warehouse Design:
Relational and Dimensional Techniques. John Wiley, 2003
 W. H. Inmon. Building the Data Warehouse. John Wiley, 1996
 R. Kimball and M. Ross. The Data Warehouse Toolkit: The Complete Guide to
Dimensional Modeling. 2ed. John Wiley, 2002
 P. O'Neil and D. Quass. Improved query performance with variant indexes.
SIGMOD'97
 Microsoft. OLEDB for OLAP programmer's reference version 1.0. In
http://www.microsoft.com/data/oledb/olap, 1998
 A. Shoshani. OLAP and statistical databases: Similarities and differences.
PODS’00.
 S. Sarawagi and M. Stonebraker. Efficient organization of large
multidimensional arrays. ICDE'94
 P. Valduriez. Join indices. ACM Trans. Database Systems, 12:218-246, 1987.
 J. Widom. Research problems in data warehousing. CIKM’95.
 K. Wu, E. Otoo, and A. Shoshani, Optimal Bitmap Indices with Efficient
Compression, ACM Trans. on Database Systems (TODS), 31(1), 2006, pp. 1-38.

Data Warehouse , Data Cube Computation

More Related Content

What's hot (20)

Similar to Data Warehouse , Data Cube Computation (20)

Recently uploaded (20)

Data Warehouse , Data Cube Computation