DeE_Data_Architecture_QA

Copyright © 2014 Data-e-Education.com All Rights Reserved. Presented By: Prem Nath R.K. Data Architect Consultant
Data Architect - Interview Q and A

Introduction
1Copyright © 2014 Data-e-Education.com All Rights Reserved. Presented By: Prem Nath R.K. Data Architect Consultant
Data-e-Education is a self e-learning portal aimed to provide an in-depth knowledge with a subject focus on Data Architecture, System Analyse Method,
Programming in SQL Server and Database Administration. Data-e-Education Target Audiences such as Business Analyst, Data Analyst, Data Architect,
Developer, Database Administrator and Subject Master Expert.
This e-learning portal gives an elaborate e-course self learning material with 300 interrelated Technical Articles. These courses are prepared after an
extensive research on various resources like Trainings Materials, Research Papers and White Papers. All these valuable e-courses are absolutely FREE.
Based on your interest, click on any course below:
This e-book describes “Fundamental of Data Architecture’s Questions and Answers”.
These 68 Questions and Answers helps you to attend Telephonic Interview.
Please you will send your feedback and additional Questions to prem@data-e-education.com
* * * * *

Normalization - Q and A

3
Question 01
What is Normalization?
Answer 01
 Normalization is a mathematical
approach that helps to decompose a
complex Table Relation (Rx) into
multiple simplex Relations (R1, R2,
R3 ... Rn), based on Functional
Dependency Rules.
 E.F.Codd introduced Normal Form
1,2 and 3 to design Relational
Database Structure.
Reference URL: http://www.data-e-education.com/RDBMS/Database_Normalization.php

4
Question 02
What is First Normal Form(1 NF)?
Answer 02
 First Normal Form - 1NF is used to
identify and eliminate repeating
group of data or a repeating group
of attributes from one entity that
decompose into two different
entities to produce atomic data.
Reference URL: http://www.data-e-education.com/RDBMS/First_Normal_Form_1NF.php

5
Question 03
What is Second Normal Form(2 NF)?
Answer 03
 Second Normal Form - Each 1NF
entity must imply with 2NF that
helps to identify and eliminate
redundant information from 1NF
entity to avoid update and delete
anomalies.
Reference URL: http://www.data-e-education.com/RDBMS/Second_Normal_Form_2NF.php

6
Question 04
What is Third Normal Form 3 NF?
Answer 04
 Third Normal Form - 3NF is used to
eliminate transitive dependency
attributes from one entity that
moved into external entity to avoid
update, insert and delete anomalies.
Reference URL: http://www.data-e-education.com/RDBMS/Third_Normal_Form_3NF.php

7
Question 05
What is Boyce Codd Normal Form
BCNF?
Answer 05
 Boyce and Codd Normal Form -
BCNF is used to eliminate non-trivial
dependencies between candidate
key attributes from one entity that
moved into external entity to avoid
update, insert and delete anomalies.
Reference URL: http://www.data-e-education.com/RDBMS/Boyce_Codd_Normal_Form.php

8
Question 06
What is Fourth Normal Form 4 NF?
Answer 06
 Fourth Normal Form - 4NF is used
to decompose many non-trivial
multivalued dependencies into
individual non-trivial multivalued
dependencies that avoid redundant
data.
Reference URL: http://www.data-e-education.com/RDBMS/Fourth_Normal_Form_4NF.php

9
Question 07
What is Fifth Normal Form 5 NF?
Answer 07
 Fifth Normal Form - 5NF is used to
decompose semantically related
many non-trivial multivalued
dependencies into individual non-
trivial multivalued dependencies that
avoid redundant data.
Reference URL: http://www.data-e-education.com/RDBMS/Fifth_Normal_Form_5NF.php

De-normalization - Q and A

11
Question 08
What is De-normalization?
Answer 08
 De-normalization is used to optimize
data processing while update or
retrieval of records. De-
normalization can be applied in four
different approaches…
1. Maintain “Controlled Redundancy
Data” in Parent Table itself.
2. Maintain “Derived and Calculation
Attributes” in Separate Entity.
3. Decompose “Table Structure” from
one Entity into many Entities.
4. Decompose “Records SET” from
one Entity into many Entities.
Reference URL: http://www.dfpug.de/loseblattsammlung%5Conline/workshop/jim_booth/seite3.htm

12
Question 09
What is Horizontal De-normalization?
Answer 09
 Decompose Table Structure from
one Entity into two or many different
Entities based on Business Unit
Process like Commercial and
Finance transaction.

13
Question 10
What is Vertical De-normalization?
Answer 10
 Decompose Records SET from one
Entity into many sequence Entities
based on Range of Key, Group of
Business Unit, and Date of
Transaction.

Entity Relationship Model - Q and A

15
Question 11
What is Entity in E-R Model?
Answer 11
 Entity, Relationship and Attribute are
Components of Entity Relationship
Model.
 Entity is used to store Real (Person
/ Place / Thing) or Abstract (Concept
/ Event) Information in the
Database.
 Entity is equivalent to Table in the
Relational Database.
Reference URL: http://www.data-e-education.com/ERM/images/ERM_01_Entity.png

16
Question 12
What is Relationship in E-R Model?
Answer 12
 Relationship is a connection
between one or more entities in the
Data Model.
 The interrelated entities must have
relevant information.
Reference URL: http://www.data-e-education.com/ERM/Relationship.php

17
Question 13
What is Degree of Relationship?
Answer 13
 The number of participating entity
types in a relationship is called
Degree of Relationship.
Reference URL: http://www.data-e-education.com/ERM/Degree_Of_Relationship.php

18
Question 14
What is Binary Relationship?
Answer 14
 A relationship of degree two is
called binary relationship.
 For An instance, an I.T. Company is
formed with many Departments and
each Department is formed with
Employees.

19
Question 15
What is Ternary Relationships?
Answer 15
 A relationship of degree three is
called ternary relationship.
 The Ternary and N-ary
Relationships should be redesign
into to multi-level Binary
Relationship.
 For An instance, an I.T. Company is
developed and maintaining many
Projects with many Tasks. Each
Task is assigned to one or many
Employees.

20
Question 16
What is Unary Relationship?
Answer 16
 A relationship of degree one is
called recursive relationship.
 In Recursive Relationship, one
instance of an entity (Employee) is
associated with another instance of
same entity (Employee).

21
Question 17
What is Connectivity and Cardinality?
Answer 17
 The connectivity of a relationship
describes the mapping of
associated entity instances in the
relationship.
 The cardinality of a relationship is
the actual number of related
occurrences for each of the two
entities.
 In E-R Modelling, we can design
data model with three different types
of connectivity and Cardinality
Relationships.
Reference URL: http://www.data-e-education.com/ERM/Connectivity_And_Cardinality.php

22
Question 18
What is Associate Entity?
Answer 18
 Relational Database does not
support Many-To-Many relationships
that happen frequently in the real
time Business Process Workflow. To
avoid this RDBMS limitation, we
need to define a Common Entity
and map the attributes into two or
more different entities. This
Common Entity is called Associate
Entity.
Reference URL: http://www.data-e-education.com/ERM/Associate_Entity_I_M_M.php

23
Question 19
What is Redundant Relationships?
Answer 19
 Relationship between two entities
that is equivalent to another
Relationship between two entities
that transitive through an
intermediate entity is called
Redundant Relationships.
Reference URL: http://www.data-e-education.com/ERM/Eliminate_Redundant_Relationships.php

24
Question 20
What is Identify Relationship?
Answer 20
 An instance of a Child Entity is
identified through its association
with the Parent Entity is called
Identifying Relationship and one
instance of the Parent Entity is
related with one or more instances
of Child Entity. This kind of Child
Entity is called Weak Entity.
Reference URL: http://www.data-e-education.com/ERM/Types_Of_Relationships.php

25
Question 21
What is Non-identify Relationship?
Answer 21
 An instance of a Child Entity is
identified by itself and the Child
Entity is not identified through its
association with the Parent Entity is
called Non-Identifying Relationship
and one instance of the Parent
Entity is related with one or more
instances of Child Entity.
Reference URL: http://www.data-e-education.com/ERM/Types_Of_Relationships.php

26
Question 22
What is Attribute in E-R Model?
Answer 22
 Attribute is an element of an Entity.
 Entity can design with one or more
related Data Attributes.
 Attribute can categories into two
types that Identifier and Descriptor.
Reference URL: http://www.data-e-education.com/ERM/Attribute.php

27
Question 23
What is Identifier Attribute?
Answer 23
 Identifier Attribute is called Key
Attribute that specifies a unique
value in the Entity to identify an
instance of the transaction(s).
Reference URL: http://www.data-e-education.com/ERM/Identifier_Attribute.php

28
Question 24
What is Descriptor Attribute?
Answer 24
 Descriptor Attribute is called Non-
Key Attribute that specifies a non-
unique value in the Entity that
describes more information about
an instance of the transaction(s).
Reference URL: http://www.data-e-education.com/ERM/Descriptor_Attribute.php

29
Question 25
What is Conceptual Data Model?
Answer 25
 Conceptual Data Model is a System
Analytical Tool in the Entity
Relationship Model that is used to
define a Business Process Flow
based on Business Information.
 Conceptual Data Model helps to
identify and defines Business
Entities with Degree of Relationship,
Connectivity and Cardinality of
Relationships and Types of
Relationships.
Reference URL: http://www.data-e-education.com/ERM/Conceptual_Data_Model.php

30
Question 26
What is Logical Data Model?
Answer 26
 Logical Data Model is a System
Analytical Tool in the Entity
Relationship Model that is used to
define a Business Process Flow and
Logical Database Structure with
Entities, Relationships, Attributes
and Domain Rules.
 Logical Data Model is an extension
of Conceptual Data Model that helps
to design Physical Database
Structure.
Reference URL: http://www.data-e-education.com/ERM/Conceptual_Data_Model.php

31
Question 26
What is Physical Data Model?
Answer 26
 Physical Data Model is used to
define an Application's Database
Structure with Database Space,
Segment, Block, Partition,
Abbreviation, Glossary, Table,
Column, Domain Type, Domain
Size, Domain Constraint, Key
Constraint, Referential Integrity
Constraint, Stored Procedure,
Trigger, Function and Index Data
Objects.
Reference URL: http://www.data-e-education.com/ERM/Physical_Data_Model.php

32
Question 27
What is Generalization Hierarchy?
Answer 27
 A generalization hierarchy is an
Inheritance re-presentation method
that specifies common attribute of
multiple entities into generalized or
higher level entity type is called
Supertype Entity and the lower level
entity designed with unique attribute
of each entity is called Subtype
Entity.
Reference URL: http://www.data-e-education.com/ERM/Generalization%20Hierarchies.php

33
Question 28
What is Supertype Entity?
Answer 28
 The Supertype Entity is a generic
entity type that is related to one or
more Subtype Entities.
 Supertype Entity defines common
characteristics attributes.

34
Question 29
What is Subtype Entity?
Answer 29
 The Subtype Entity is a specific
entity type that is related to one
Supertype Entity.
 Subtype Entity defines unique
characteristics attributes.

35
Question 30
What is Overlapping Subtype Entity?
Answer 30
 Overlapping Subtype Entity defines
non-unique characteristics attributes
and related with more than one
layer Supertype Entity.

36
Question 31
What is Disjoint Subtype Entity?
Answer 31
 Disjoint Subtype Entity defines
unique characteristics attributes and
related with one Supertype Entity.

37
Question 32
Accountability of Data Architect?
Answer 32
 Data Architecture helps …
1. To Optimize Business Process by
Feasibility Study.
2. To Improve System Analytical
Process by Structured System
Analysis Method.
3. To Transform Knowledge between
Non-technical and Technical Team.
4. To Optimize Legacy Database
Structure.
5. To Re-engineering Legacy
Database Structure.

Dimensional Data Model - Q and A

39
Question 33
What is Data Warehouse?
Answer 33
 Data warehouse is a centralized
repository of Enterprise's Historical
and Operational Information that
helps to query and analyse
Enterprise Operation and to make
Management Decision.
Reference URL: http://www.data-e-education.com/DDM/index.php

40
Question 34
What is Data Mart?
Answer 34
 Data Mart is a subset of the Data
Warehouse that is based on
Subjective of Enterprise Business
Unit or Team.
 Each Data Marts may or may not be
dependent or related to other Data
Marts in the Data Warehouse.
Reference URL: http://www.data-e-education.com/DDM/Data_Marts.php

41
Question 35
What is W.H.Inmon Approach?
Answer 35
 Inmon vision is a Data warehouse of
centralized repository for “Corporate
Information Factory” (CIF), which
provides a Logical Framework for
delivering Business Intelligence (BI)
and Business Management
Capabilities.
 Inmon approach is called Top to
Bottom approach and it follows
Water Fall Software Development
Methodology.
Reference URL: http://www.data-e-education.com/DDM/images/DDM_01_Inmon_Methodology.png

42
Question 36
What is Ralph Kimball Approach?
Answer 36
 Ralph Kimball is a well-known
author of data warehouse system
concept and introduced
“Dimensional Data Architecture”
based on Fact and Dimension
Technique.
 Kimball approach is called Bottom to
Top approach and it follows Spiral
Software Development
Methodology.
Reference URL: http://www.data-e-education.com/DDM/images/DDM_01_Inmon_Methodology.png

43
Question 37
What is Dimensional Data Model?
Answer 37
 Dimensional Data Model is a
specialized adaptation of the
Relational Data Model that is used
to store and manage Enterprise
Historical and Transactional
Information in the Data Warehouse.
 In dimensional data model, every
transaction data are decomposed
and stored in “FACT” and
“Dimension” data format. FACT
table contains Composite Primary
Key and Measurement data and
Dimension table contains Foreign
Key and Context data that refer the
FACT table.
Reference URL: http://www.data-e-education.com/DDM/Dimensional_Data_Model.php

44
Question 38
What is Business Intelligence?
Answer 38
 BI - Business Intelligence is an
Enterprise-aided Model that uses a
compilation of Process Analytical
Methods, Data Architecture
Methodologies, Technologies, and
Applications to get the Right
Information, to the Right Decision
Makers, at the Right Time.
Reference URL: http://www.data-e-education.com/DDM/Business_Intelligence.php

45
Question 39
What is OLTP?
Answer 39
 OLTP - Online Transaction
Processing System is used to store
and maintains Business
Transactional Information in the
Database.
 OLTP Database helps to generate
MIS Transaction Report for
Management to know the status of
Business Transaction.
Reference URL: http://www.data-e-education.com/DDM/OLTP_Vs_OLAP.php

46
Question 40
What is OLAP?
Answer 40
 OLAP - Online Analytical Processing
System helps to generate MIS
Analysis Report from Data
warehouse by using CUBE. In
OLAP, a predefined
Multidimensional Data Structures is
called CUBE.
 CUBE is used to organize and
summarize Data warehouse data
and it helps to generate Typical
Explorative Analysis Questions that
can be answered with little or no
querying of the Relational Database.
Reference URL: http://www.data-e-education.com/DDM/OLTP_Vs_OLAP.php

47
Question 41
What is Data Mining?
Answer 41
 Data Mining is an artificial
intelligence process to identify
uncovering hidden patterns from
data.
 Data Mining is used in marketing,
surveillance, fraud detection and
scientific industries to discover
uncovering hidden patterns from
data.
Reference URL: http://www.data-e-education.com/DDM/Data_Mining.php

48
Question 42
What is Data Hierarchy?
Answer 42
 Data Structure represents a
Systematic Organization of Data
that can form through collection of
interrelated Columns, Records and
Tables in Relational Database
Management System.
 Data Hierarchy represents a
Logical Top-To-Bottom Ordered
Structure of Record(s).
Reference URL: http://www.data-e-education.com/DDM/Data_Hierarhy.php

49
Question 43
What is FACT Table?
Answer 43
 FACT Table is used to store
measurements, metrics or facts of a
business process in the Data
warehouse.
 Fact table design should start with
declaration of grain. Grain is the
base business definition of Fact
table that determine measurement
of business event.
 Granularity is an extent of
information that represents
particular transaction data in the
Fact table.
Reference URL: http://www.data-e-education.com/DDM/FACT_Table.php

50
Question 44
What is Additive Fact?
Answer 44
 Additive fact can summarize and
can be added across all dimensions
around fact table. Transaction level
fact table must be design with
Additive facts.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Facts.php

51
Question 45
What is Semi-additive Fact?
Answer 45
 Semi-additive fact can summarize
and can be added across few
dimensions around fact table.
Summarization level fact table must
be design with Semi-additive facts.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Facts.php

52
Question 46
What is Non-additive Fact?
Answer 46
 Unit Price, Profit Margin,
Percentage and Ratio are Non-
additive fact that cannot be
summarized and added across all
dimensions around fact table.
 Summarization level fact table
optionally designed with Non-
additive facts.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Facts_02.php

53
Question 47
What is Transaction Fact Table?
Answer 47
 Transaction fact table is used to
record one row per transaction with
measures of particular business
event. Transaction fact table must
be designed with Additive facts.
Additive fact can summarize and
added across all dimensions around
fact table.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Fact_Tables.php

54
Question 48
What is Periodic Fact Table?
Answer 48
 Periodic fact table is used to record
one row for a group of transactions
with aggregated measures of
business events that happened over
a period of time.
 Periodic fact table must be designed
with Semi-additive facts. Semi-
additive fact can be summarized
and added across few dimensions
around fact table.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Fact_Tables_01.php

55
Question 49
What is Accumulating Fact Table?
Answer 49
 Accumulating fact table is used to
record one row for the entire lifetime
of transaction with measures of
business events that happened over
a period of time.
 Accumulating fact table must be
designed with combination of
Additive facts with multiple date
keys that helps to track the lifetime
of business event.
 Non-additive facts are optional in
this type of fact table.
Reference URL: http://www.data-e-education.com/DDM/Types_Of_Fact_Tables_02.php

56
Question 50
What is Aggregate Fact Table?
Answer 50
 Fact table might design with
transactional level facts or
summarization level facts.
 Aggregate FACT table must be
design with summarization level
facts. Summarization level facts are
used to store aggregated or
grouped or summed up data that is
based on certain level of
hierarchical data of dimension table.
Reference URL: http://www.data-e-education.com/DDM/Aggregate_FACT.php

57
Question 51
What is Factless Fact Table?
Answer 51
 Relational Database is to support
Binary and Unary relationship and
that does not support Ternary and
N-ary relationships.
 Factless fact table is used to
capture many-to-many relationships
between fact and dimensions, but
contains no numeric or textual facts
that help to record events or
coverage information.
Reference URL: http://www.data-e-education.com/DDM/Factless_FACT.php

58
Question 52
What is Event Factless Fact Table?
Answer 52
 Event Factless fact table is used to
captures many-to-many
relationships between fact and
dimensions, but no numeric or
textual facts.
 The existence of the relationship
itself is the fact. Boolean data type
attribute should be added in this
factless table that helps to stores 1
or 0 data value.
Reference URL: http://www.data-e-education.com/DDM/Factless_FACT.php

59
Question 53
What is Coverage Factless Fact Table?
Answer 53
 Coverage Factless fact table is used
to capture many-to-many
relationships between fact and
dimensions, but no numeric or
textual facts.
 The existence of the relationship
itself is the fact that helps to find out
non-existence event records.
Reference URL: http://www.data-e-education.com/DDM/Factless_FACT_01.php

60
Question 54
What is Dimension Table?
Answer 54
 Dimension Table is used to maintain
Hierarchy Textual Information of
business process that is integrated
with Fact Table in the Data
warehouse.
 FACT Table is used to store
measurements, metrics or facts of a
business process in the Data
warehouse.
 Grain is the base business definition
of Fact table that determine
measurement of business event.
Reference URL: http://www.data-e-education.com/DDM/Dimension_Table.php

61
Question 55
What is Shared Dimension Table?
Answer 55
 Shared Dimensions describes the
core set of dimensions that is
shared across the various cube
perspectives in a Data Mart.
 These dimensions allows the
measures in the cube to be
summarized according to the
different analytical questions.
 Shared Dimensions are used to
maintain and provide consistent
information for similar queries.
Reference URL: http://www.data-e-education.com/DDM/Dimension_Table.php

62
Question 56
What is Confirmed Dimension Table?
Answer 56
 Conformed Dimension is described
as common structured dimension
that is shared across the various
FACT tables in the Data
Warehouse.
 Conformed Dimensions are used to
avoid redundant data in the Data
Warehouse.
Reference URL: http://www.data-e-education.com/DDM/Conformed_Dimension.php

63
Question 57
What is Role-Playing Dimension Table?
Answer 57
 Role-Playing Dimension is
described as a common structured
dimension that appears more than
one time in the same Fact table.
 Role-Playing Dimension is used to
avoid multiple SQL Query Joins that
helps to optimize SQL Query
Operation.
Reference URL: http://www.data-e-education.com/DDM/Role_Playing_Dimension.php

64
Question 58
What is Slowly Changing Dimension
Table?
Answer 58
 Slowly Changing Dimension is used
to maintain slowly changing data
rather than changing data on a time-
based, regular schedule.
 Slowly Changing Dimension
categories into SCD Type 01, SCD
Type 02, SCD Type 03 and Hybrid
SCD that based on data storage
technique.
Reference URL: http://www.data-e-education.com/DDM/Slowly_Changing_Dimension.php

65
Question 59
What is SCD Type 01?
Answer 59
 SCD Type 01 method is used to
overwrite old data with new data in
the Data warehouse.
correct mistake data in the Data
warehouse.
 SCD Type 01 method doesn't
maintain historical data that is not
good idea to maintain Slowly
Changing Data in the Data
warehouse.
Reference URL: http://www.data-e-education.com/DDM/Slowly_Changing_Dimension.php

66
Question 60
Answer 60
maintain Slowly Changing Unlimited
Historical Data through actual table
structure with new Version Number
or Effective Period like Start_Date
and End_Date attributes.
Reference URL: http://www.data-e-education.com/DDM/Slowly_Changing_Dimension_01.php

67
Question 61
Answer 61
maintain Slowly Changing Limited
Historical Data through actual table
structure with new Effect Date and
Previous Data attributes.

68
Question 62
What is Hybrid SCD?
Answer 62
 SCD Type 06 method is a
combination of SCD Type (1 + 2 +
3) approach that is used to maintain
Slowly Changing Unlimited and
Detailed Historical Data through
actual table structure with new
Effective Period of Start_Date,
End_Date and Current Flag
attributes.
 Ralph Kimball describes that Hybrid
SCD / SCD Type 06 is an
"Unpredictable Changes with
Single-Version Overlay". SCD Type
06 is advisable to maintain Historical
Detail Data in the Data Warehouse.

69
Question 63
What is Surrogate Key?
Answer 63
 Surrogate Key is a System Key that
does not have any intelligence
information, typically database
generate the System Key
information with internal
mechanism. Slowly Changing
Dimension Table must design with
Surrogate Key that helps to identify
unique record from multiple version
records.
 The current version record can
retrieve through combination of
Surrogate Key + Primary Key +
MAX(Version Number). The
particular period historical records
can retrieve through combination of
Surrogate Key + Primary Key +
Start_Date AND End_Date.
Reference URL: http://www.data-e-education.com/DDM/Surrogate_Key.php

70
Question 64
What is Junk Dimension Table?
Answer 64
 Junk Dimension is used to record a
collection of low-cardinality Flags
and Indicators data.
 Flag data may be non-generic
question's & answers like Yes/No or
True/False or Activate/Deactivate.
 Indicator data may be Height, Width,
Weight, Color, Status and other tiny
text data.
Reference URL: http://www.data-e-education.com/DDM/Junk_Degenerate_Dimension.php

71
Question 65
What is Degenerate Dimension Table?
Answer 65
 Transaction Number or Document
Number is not a Fact and it is not an
Entity of Event to maintain in the
Dimension table.
 Most of the time, Transaction
Number is a standalone attribute in
the Dimension table.
 Every Transaction Number has one
to one relationship between Product
Code or Service Code
measurement data in the Fact table.
Degenerate Dimension Concept
helps to eliminates standalone
Transaction Number attribute in
Dimension table and improve
database operation.Reference URL: http://www.data-e-education.com/DDM/Junk_Degenerate_Dimension_01.php

72
Question 66
What is STAR Schema?
Answer 66
 STAR Schema describes a logical
database structure of Data
Warehouse or Data Mart.
 STAR Schema can design with one
de-normalized FACT and one or
many shared Dimension table(s)
that looks like STAR.
Reference URL: http://www.data-e-education.com/DDM/STAR_Schema.php

73
Question 67
What is Snowflake Schema?
Answer 67
 Snowflake Schema describes a
logical database structure of Data
Warehouse or Data Mart.
 Snowflake Schema can design with
one de-normalized FACT and one or
many normalized Dimension
table(s).
 Snowflake Schema is an extended
and normalized STAR Schema.
Reference URL: http://www.data-e-education.com/DDM/Snowflake_Schema.php

74
Question 68
What is Fact Constellation Schema?
Answer 68
 FACT Constellation Schema is
describes as a logical database
structure of Data Warehouse or
Data Mart.
 FACT Constellation Schema can
design with collection of de-
normalized FACT, Shared and
Conformed Dimension tables.
 FACT Constellation Schema is an
extended and decomposed STAR
Schema.
Reference URL: http://www.data-e-education.com/DDM/images/DDM_20_Fact_Constellation_Schema.png

Thank You
Data Architecture Expert
https://www.facebook.com/groups/dataarchitectureexperts/

DeE_Data_Architecture_QA

More Related Content

Similar to DeE_Data_Architecture_QA (20)

DeE_Data_Architecture_QA