Category: CSE Seminar Topics

High Level Conceptual Data Modeling for Database Design

For designing a victorious database application this conceptual modeling is playing a vital phase. This conceptual data modeling is playing higher level this will present the idea about the Entity-Relationship Model.

                                  Conceptual Design Workflow

               From the above figure we note that the first stride is requirements collection and analysis.  At some stage to recognize and the requirements are certified by the users that’s why the database designers are interrogate. At last the outcome of this step is the set of customer requirements are printed in a few words. When we are identifying the data requirements in parallel it is necessary to identify the functional requirements which we are acknowledged.  These user-defined operations include both retrievals and update.

The data flow diagrams, sequence diagrams are frequently used in the software design. And some of the functional requirements are identified. All of these requirements are examined together one time. And then by using a high-level conceptual data model then generate a conceptual schema in next step. The users’ database requirements are illustrated briefly in the conceptual schema the complete descriptions of the entity types, relationships and constraints are also contained.  By using a commercial DBMS the database is implemented really in the next stage of the database design.

                From a high level data model, the conceptual schema is changed into the implementation data model when the implementation of the data model is used by the current commercial DBMS in many ways.  This is known as logical design or data model mapping. The physical design phase is the very last step.  Individually the internal storage structures, access paths, and file organizations. 

DATA MODELING

           In the process of data modeling, depiction and observation of the developers regarding the data is created and this data modeling is the main vital task in the growth of the effective database request. When the developers find that the view of data is signified improperly by the data model, then the developers find it very complicated to extend, incomplete and it is very provoking of applications.

For the improvement of the databases and their applications for all successive work the data modeling is the starting point.

 For the data modeling we have two types of approaches. They are

• Entity relationship model
• Semantic object model

Entity relationship model

      We have four fundamental elements in the entity relationship model.

• Attributes
• Identifiers
• Relationships

 For these entities the attributes are behaving like assets. For an example the person is entity and the age of a person or name of the person is the attribute.

Individually for the particular entities we can identify these identifiers.

Between the two entities we are using the se relationships.

                    

                                             The Works in Relationship Set

For the entities we are using the attribute known as relationship. This relationship is in different types.

•  One-to-One relationship
•  One-to-Many relationship
•  Many-to-One relationship
•  Many-to-Many relationship

Semantic Object Model

       In the place of entities we are using the semantic objects in the semantic object model. In the user environment with a quantity of the identifiable thing the semantic object is characterized. For a semantic object that the object must have sufficient attributes that is the reason the word ‘sufficiently’ is working as a major role.

           For the E-R model and in the semantic object model there is one of the main difference is in the E-R model the entities are the basic focus where as in the semantic object model these semantic objects are focus in the other model. 

                                   

                                        A Semantic Model

          To signify the different types of data we are having variety of classes. Some of the necessary objects are Simple Objects, Composite Objects, Compound Objects, and Hybrid Objects.

A programming language which is the latest method began to use in the year 1980’s known as Object Oriented Programming (OOP). As the established programming language OOP has exposed the important benefits. As a different way we can understand the database models when there is a change in programming it is called as Object Oriented Database Management System (OODBMS).

           With the object oriented capabilities the DBMS combine with the database capabilities essentially the DBMS is an OODBMS. As a programming language objects the database objects come out which are created by an OODBMS. For the real world, article objects are the software demonstration. Both attributes and operational characteristics are consisted in an object because to detain the features and ability of the real world.

For objects templates are classes. Encapsulation is one of the main features for the object oriented model. For an object oriented database the main ability is to add new objects. There is a flexible thing in the database level.  Base class means there is an existing object that can be distinct as a condition of the new objects in the object oriented model (OOM). To subsist in the subclasses the polymorphism allocates a number of versions who have the similar performance. 

Relational Database Model (RDM)

In the year 1960’s, to hold the huge quantity of data Dr. E. F. Codd, a mathematician   appears a latest condition. Codd had some of the plan to explain various problems to relate disciplines and structures of mathematics to data management. But when it is planned for other database models it comes across unnecessary data and weak data integrity. In the year 1970, June the relational model of data is established by Codd.  On the two branches of mathematic the Codd support his model.

       The Codd does not get the concentration on the new papers but in the relational database concept the IBM became extra importance. For a System/R and a relational language to form a sequence of design documents the IBM is going to use their thoughts.

          The relational language is also called as structured query language. By using the SQL we can get back the data.SQL is one of the ordinary language by using this we can create, modify, and inquiry relational databases.

In the relational database model we can observe three essential components.

1. Relational data structure
2. The data structures association would be managed by the rules
3. And in the data structures the operations are presented.

There are more benefits in the Relational Database Model.

• Built in multi level integrity   

            To make certain that the records are not copied and to identify the values of primary key that is misplaced, we built data integrity into the model that is at table level.

• Logical and physical data independence from database applications

When we put the applications together it will not badly affect for the logical and physical design of the RDM when some of the modifications made by the user.

• Guaranteed data consistency and redundancy

Within the database the data can be forced when the data is exact and constant due to the different stage of integrity.

• Easy data retrieval

                In the database level we can recover the data from a particular table or any correlated table which is controlled by users.

            To create, maintain and manipulate a relational database a software program known as Relational Database Management System is used. To produce the request the RDBMS is used and the data which is stored in the database should be work together with the users. For creating an extra and better mainframe RDBMS programs the various companies are using relational database model which are moved from the hierarchical database model and network database model. Oracle and IBM’s DB2 are some of the mainframe computers in RDBMS this improvement is seen in 1980.

In the data base management systems to control and to continue the data generally we are using two types of data models. They are

• Hierarchical Database Model (HDM)
• Network Database Model (NDM)

Hierarchical Database Model

               The data is hierarchically prepared in the hierarchical database model. Without any difficulties we can imagine an inverted tree in this hierarchical database model. In this model the relationships are signified by a parent and child. In this HDM model as a root only one table is acting and the remaining other tables are acting as branches which are extending from this root.  Means the parent is a root and the child are branches. Means that the child table will have only one parent table and the parent table will have many child tables. These tables are connected clearly. In this model if we want to contact any documentation then the user first want to start from the root table from there onwards move through the tree to reach the object.

         For this type of database there is one of the benefits that is referential integrity which is built in and it is imposed repeatedly. For the record which is presented in the parent table the record of a child table is related. Suppose, in the parent table the records are deleted then the records which are connected in the child tables are also deleted.

         When we want to accumulate a record in a child table but it has the no relationship with the parent table then the problem occurs in the Hierarchical database model. In a parent table when the fake record is included that rules could be bend lacking of break.

           There is unnecessary data in HDM; this is one of the problems. In HDM we see many-to-many interactions but it is not sufficiently elastic that is the reason unnecessary data occurs in HDM. Mutually the data is repeating. In the year 1970’s by itself the hierarchical systems provide tape storage method which are using in main frames. And the companies used this systems turn into very popular.

            In many conditions the HDM was very helpful model because it is quick and we can access directly. Along with the data we are having unnecessary data and the complex relationships to address this we require a new database model.

                      

                            A Hierarchical Database Model Diagram

Network Database Model

              In the hierarchical model we are having some of the problems to avoid these problems Network Database Model were extended. In the network database model the branches can be shared in many imagine trees. By sharing these branches the many-to-many connection problem is solved which is not executed in a HDM. The model of data definition and the data definition language are established in NDM.