Normalisation in Database management System (DBMS)
1 like569 views
The document discusses the process of normalization in database design, detailing the steps from first normal form to third normal form and highlighting the importance of removing partial and transitive dependencies. It then shifts focus to physical database design, emphasizing the need for efficient data storage and execution while maintaining data correctness. Key considerations include data types, attribute grouping, and strategies for efficient access to ensure optimal performance.
Normalisation in Database management System (DBMS)
- 1. For more Https://www.ThesisScientist.com Unit 5 Normalization Normalization is a technique to organize the contents of the table for transactional database and data warehouse. First Normal Form : Seeing the data in the example in the book or assuming otherwise that all attributes contain the atomic value, we find out the table is in the 1NF. Second Normal Form : Seeing the FDs, we find out that the K for the table is a composite one comprising of empId, projName. We did not include the determinant of fourth FD, that is, the empDept, in the PK because empDept is dependent on empId and empID is included in our proposed PK. However, with this PK (empID, projName) we have got partial dependencies in the table through FDs 1 and 3 where we see that some attributes are being determined by subset of our K which is the violation of the requirement for the 2NF. So we split our table based on the FDs 1 and 3 as follows : PROJECT (DNELCepi, projMgr, startDate) EMPLOYEE (ipDwd, empName, salary, empMgr, empDept) WORK (DNELCepi.oipDwd, hours, rating) All the above three tables are in 2NF since they are in 1NF and there is no partial dependency in them. Third Normal Form Seeing the four FDs, we find out that the tables are in 2NF and there is no transitive dependency in POJECT and WORK tables, so these two tables are in 3NF. However, there is a transitive dependency in EMNPLOYEE table since FD 1 say empId empDept and FD 4 say
- 2. For more Https://www.ThesisScientist.com empDept P empMgr. To remove this transitive dependency we further split the EMPLOYEE table into following two: EMPLOYEE (ipDwd, empName, salary, empDept) DET (ipDriDl, empMgr) Hence finally we got four tables PROJECT (DNELCepi, projMgr, startDate) EMPLOYEE (ipDwd, empName, salary, empDept) WORK (DNELCepi.oipDwd, hours, rating) DET (ipDriDl, empMgr) These four tables are in 3NF based on the given FD, hence the database has been normalized up to 3NF. Physical Database Design After completing the logical database design and then normalizing it, we have to establish the physical database design. Throughout the processes of conceptual and logical database designs and the normalization, the primary objective has been the storage efficiency and the consistency of the database. So we have been following good design principles. In the physical database design, however, the focus shifts from storage efficiency to the efficiency in execution. So we deliberately violate some of the rules that we studied earlier, however, this shift in focus should never ever lead to incorrect state of the database. The correctness of the database we have to maintain in any case. When we do not follow the good design principles then it makes it difficult to maintain the consistency or correctness of the database. Since the violation is deliberate, that is, we are aware of the dangers due to violations and we know the reasons for these violations so we have to take care of the possible threats and adopt appropriate measures. Finally, there are different possibilities and we as designers have to adopt particular ones based on certain reasons or objectives. We have to be clear about our objectives. The physical DB design involves :- Transforms logical DB design into technical specifications for storing and retrieving data
- 3. For more Https://www.ThesisScientist.com Does not include practically implementing the design however tool specific decisions are involved It requires the following input: Normalized relations (the process performed just before) Definitions of each attribute (means the purpose or objective of the attributes. Normally stored in some form of data dictionary or a case tool or may be on paper) Descriptions of data usage (how and by whom data will be used) Requirements for response time, data security, backup etc. Tool to be used Decisions that are made during this process are: Choosing data types (precise data types depend on the tool to be used) Grouping attributes (although normalized) Deciding file organizations Selecting structures Preparing strategies for efficient access That is all about today’s lecture, the discussion continues in the next lecture. Summary In today’s lecture we summarized the normalization process and also saw an example to practically implement the process. We have introduced our next topic that is the physical DB design. We will discuss this topic in the lectures to be followed. The Physical Database Design Considerations and Implementation The physical design of the database is one of the most important phases in the computerization of any organization. There are a number of important steps involved in the physical design of the database. Steps are carried out in sequence and need to be performed precisely so that the result of the first step is properly used as input to the next step. Before moving onto the Physical database design the design of the database should have undergone the following steps, Normalization of relations Volume estimate Definition of each attribute Description of where and when data is used (with frequencies)
- 4. For more Https://www.ThesisScientist.com Expectation or requirements of response time and data security. Description of the technologies. For the physical database design we need to check the usage of the data in term of its size and the frequency. This critical decision is to be made to ensure that proper structures are used and the database is optimized for maximum performance and efficiency. The following steps are necessary once we have the prerequisite complete: Select the appropriate attribute and a corresponding data type for the attribute. The process of selecting the attribute to be placed in a specific relation in the physical design. Need considerable care as it is one of the most important and basic aspects for the creation of the database. Grouping of attributes in the logical order so that the relation is created in such a way that no information is missing from the relation and also no redundant or unnecessary information is placed in the relation. Looking at the logical design at the time of transformation into physical design there may be stages when the information combined logically in the logical design looks odd when transforming the design into a physical one. Arrangement of Similar records into the secondary memory (hard disk) The scheme of storage on hard disk is important as it leads to the efficiency and management of the data on disk. Different types of data access mechanism are available and are useful for rapid access, storage, and modification of data. Different types of database structures can be used for placement of data on disks, management of data in the forms of indexes and different database architecture is vital and leads to better retrieval and recovery of records. Preparing queries and handling strategies for the proper usage of the database, so that any type of input or output operation performed on the database is executed in an optimized and efficient way. DESIGNING FIELDS The Field is the smallest unit of application data recognized by system software, such as a programming language or any database management system. Designing fields in the databases’ physical design as discussed earlier is a major issue and needs to be dealt with great care and accuracy. Data types are the structure defined for placing data in the attributes. Each data type is appropriate for use with certain type of data.
- 5. For more Https://www.ThesisScientist.com 4 major objectives for using data types when specifying attributes in a database are given as under: *Minimized usage of storage space *Represent all possible values *Improve data integrity *Support all data manipulation The correct data type selection and decision for proper domain of the attribute is very necessary as it provides a number of benefits. Most common data types used in the available DBMS of the day have the following set of common attributes. CODING AND COMPRESSION TECHNIQUES : There a re some attributes which have some sparse set of values, these values when they are represented in any data type are hard to express, for this purpose some codes are used. As the codes defined by the database administrator or the programmer consume less space so they are better for use in situations where we have large number of records and wastage of small amount of space in each record can lead to loss of huge amount of data storage space. Thus causing lowered database efficiency. STID STNAME HOBBY S1020 Sonam Reading S1038 narendra Gardening S1015 Tarun Reading S1015 ajay Movie S1018 naveen Reading Coding techniques are also useful for compression of data values appearing the data, by replacing those data values with the smaller sized codes we can further reduce the space needed by the data for storage in the database. Following tables give the use of codes and their utilization in the database environment
- 6. For more Https://www.ThesisScientist.com Coding Example: Student STID STNAME HOBBY S1020 Sonam R S1038 narendra G S1015 Tarun R S1015 ajay M S1018 naveen R Hobby Table CODE HOBBY R Reading G Gardening M Movies In the above example we have seen the implementation of the codes as replacement to the data in the actual table, here we actually allocated codes to different hobbies and then replace the codes instead of writing the codes in the table. We get a number of benefits by the use of data types and the benefit can be in a number of dimensions. Default value Default values are the values which are associated with a specific attribute and can help us to reduce the chances of inserting incorrect values in the attribute space. And also it can help us preventing the attribute value be left empty. Range Control Range control implemented over the data can be very easily achieved by using any data type. As the data type enforces the entry of data in the field according to the limitations of the data type. Null Value Control As we already know that a null value is an empty value and is distinct from zero and spaces, Databases can implement the null value control by using the different data types or their build in mechanisms. Referential Integrity
- 7. For more Https://www.ThesisScientist.com Referential Integrity means to keep the input values for a specific attribute in specific limits in comparison to any other attribute of the same or any other relation.