Data Modeling [Comparison of data modeling techniques ]

1
Data ModelingComparison of data modeling
techniques

• By
  
• Renjini
  
• 
  Sindhuri

2
Contents

• Introduction
• E-R modeling
• Peter Chen
• Information Engineering
• Barkers Notation
• IDEFIX
• UML modeling
• XML modeling
• X- Entity modeling
• XUML
• Conclusion

3
Introduction

• Data modeling is the act of exploring data
  oriented structures.
• Examines and compares different data modeling
  techniques
• In the data modeling techniques we have
  traditional modeling and object oriented modeling
  of data

4
E-R modeling

• It is a conceptual data model that views the real
  world as consisting of entities and relationships
  
• It is used to transform relational tables that
  are easy to understand that enables easy
  communication with the end user
• Peter Chen developed E-R model

5
Peter Chen notation

• Entities are represented in the squared cornered
  and circles as attributes
• Many Many relationships can be represented
  without associative entity
• Relationship itself has attributes and are
  considered as objects
• It failed to represent unique identifier

6
Peter Chens Model
7
Information Engineering model

• Developed by Clive Finkelstein
• Entities are represented in the squared cornered
  and attributes are not shown at all they are
  shown in a separate list called entity list
• Relationships like mandatory 1 and many can be
  represented
• Unique identifiers are not represented

8
Information Engineering model diagram
9
Barkers Notation

• Adopted by Oracle corporation for its CASE method
  
• Entities can be represented by round cornered
  rectangle
• Same entity can be represented for role an
  interaction or another kind of association
• Relationship names are prepositions and not verbs
  
• Unique identifiers can be represented by hash
  marks next to the attribute

10
Barkers Notation diagram
11
IDEFIX Notation

• It is a modeling technique that is used by many
  branches of the United States Federal government
• A relationship name is a verb
• IDEFIX shows subtypes as separate entity boxes
• IDEFIX permits multiple inheritance and multiple
  type hierarchies

12
IDEFIX diagram
13
UML

• UML is an object modeling technique
• It models object classes instead of entities
• In the object oriented world the relationships
  are called as associations
• Cardinality and optionality in UML is conveyed by
  characters or numbers
• Express in the form of more complex upper and
  lower limits
• UML introduces a small flag that includes text
  describing any business rules

14
UML diagram
15
XML Notation

• Describing data and interchanging structured and
  unstructured data on the Internet
• It is a universal language of data on web
• XML tags are used to create data structures
• XML documents have been widely used for
  interchanging data between heterogeneous systems.

16
XML notation

• An example of XML notation
• http//www.essentialstrategies.com/publications/mo
  deling/xml.htm

17
X-Entity model

• Conceptual model of XML uses X entity model in
  order to represent additional features
• The entity can be denoted by E
• (A1,.An,R1,Rm,D1,.Dk)
• Each attribute A is associated with a domain
  Dom(Ai)
• Which specifies its value set
• Cardinality is denoted by Card(Ai)(min,max)

18
X entity model diagram
19
XUML

• XUML comprises the characteristics of XML and
  UML2.
• It is used to express the containment semantics
  more explicitly
• Supporting the concept of Business Components
• Specifying the data dependencies in multiple
  context

20
XUML diagram

• UML and XUML model of a book store

21
Comparison of data modeling techniques
S.No Modeling Technique Peter Chen Information Engineering IDEFIX Richard Barkers notation UML
1. Entities squared cornered and circles as attributes Squared cornered, attributes are not shown at all. Round or square cornered rectangle Round cornered rectangle Models object classes
2. Relationship Nouns. So the relationships can be represent as objects and has attributes Verbs Verb or verb phrase Preposition not verb Associations
3. Constraints between relationships Failed to represent the constraints directly exclusive or) Can represent Constraints exclusive or ,inclusive Cannot represent Can represent Constraints exclusive or) Can represent Constraints exclusive or )
22
Comparison of Data modeling techniques
S.No Modeling Technique Peter Chen Information Engineering IDEFIX Richard Barkers notation UML
4. Cardinality Many to Many relationships can be represented between the entities without the associative entity Can represent Can represent in different ways Can represent zero or more ,atleast at least one up to many, up to one relationships express more complex upper limits, zero, 3, 6-7, or 9
5. Sub types/ Super Types Cannot represent the sub types and super type sub-types can be represented inside their super-type box Sub types can be represented as separate entity boxes separate from its super type. sub-types can be represented inside their super-type box Can represent
6. Unique Identifier Cannot represent Cannot represent Represented in the form of primary key Represented in the form of hash next to the attribute Can represent
23
Comparison of Data Modeling techniques
S.No Modeling Technique Peter Chen Information Engineering IDEFIX Richard Barkers notation UML
7. Aggregation Cannot represent Cannot represent Cannot represent Cannot represent Can represent only binary aggregations
8. Business Rules / Components Cannot Represent Cannot Represent Cannot Represent Cannot Represent Can Represent
24
Comparison of Data Modeling techniques
S.No Modeling Technique Peter Chen Information Engineering IDEFIX Richard Barkers notation UML
9. Aesthetic Simplicity Score High Medium Low High High
10. Completeness Score Low medium medium medium medium
11. Language Notation Score medium medium low medium High
25
Advantages of XUML

• XUML can express the containment semantics more
  accurately.
• Support the concept of Business Component.
• Can specify the data dependencies in
• multiple context.

26
Contd..

• XUML is more expressive, precise and
• understandable.
• More rigorous and accurate.

27
Conclusion

• By comparing the aesthetic simplicity,
  completeness, language notation (relationship)
  Mr. Barker's notation is favorable for
  requirement analysis model
• XML is used in recent trends it follows a
  standard format for representing structured and
  semi structured data on web
• X-Entity model has the advantages of both XML
  schemas and extends the ER model so that it can
  explicitly represent important features of XML
  schemas
• The distinctive features of XUML made this
  technique of data modeling the latest trend for
  conceptual modeling of data.

28
References

• 1. Conceptual Modeling of XML schemas, Bernadette
  Farias Losio,Ana Carolina Salgado , Year
  2003,Publisher ACM
• 2. XML conceptual modeling with XUML, HongXing
  Liu HuaZhong University of Science and
  Technology, P. R. China, YanSheng Lu HuaZhong
  University of Science and Technology, P. R.
  China,Qing Yang Wuhan Uni Pages 973 976, Year
  of Publication 2006, Publisher ACM Press
• 3. PETER PIN-SHAN CHEN, The Entity Relationship
  Model-Toward a Unified View of Data ,
  Massachusetts Institute of Technology, ACM
  Transactions on Data base System Volume1, Issue
  1,Publisher-ACM
• 4. Data modeling in the understanding database
  course adding UML and XML modeling to the
  traditional content. Journal of Computing
  Sciences in Colleges, Volume 17, Issue 5 (April
  2002)

29
References

• 5. Data Modeling101.
• http//www.agiledata.org/essays/dataModeling101.ht
  ml
• 6.A comparison of Data Modeling ,David C
  Hay,Essential Strategies
• Inc,October 1999.