DBMS #01: Introduction to DBMS

Data: raw facts; becomes information when it becomes meaningful.

Database: it is a large collection of related data. It models some real-world organization (e.g., a school), with entities (e.g., students, courses) and relationships (e.g., student 'a' takes course 'y'). It represents some aspect of the real world. There are various data models that are used to represent data in a database. For instance, the school database can be modelled in a tabular fashion, rather than a graphical fashion.

DBMS is a software that manages some database. It is an interface between a user and a database. It also enables to abstract data representation and storage, thereby insulating the application code from such specifications. Data retrieval is also efficient. It also offers concurrent access and crash recovery.

A data model is a set of rules or conventions to represent, store and manage the data. Models can be useful to examine or analyze some aspect of the real world.

To design a database, we will first analyze the requirements in hand. We can then have a conceptual design of the database, using any of the data models (e.g., ER model). This is done in a user-understandable fashion. Once this step is finalized, we proceed to the logical design phase, which is implementation oriented (e.g., Relational Model). In the coming lectures, we will discuss how to convert the ER model to Relational Model. The latter is more restrictive since it is more concerned with physical implementation and design of the database). Finally, we approach the implementation phase.

Popular posts from this blog

DBMS #05: Data Independence

The three levels of schema provided by the DBMS are: 1) Internal/ physical schema, which describes how the data is actually stored. 2) Conceptual/ logical schema, which describes what data is present in the database (and its meaning), and 3) External schema, which describes the use of the data. This is derived from the conceptual schema and not physically stored in the database. Physical data independence is a measure of how much a physical schema can be changed without affecting the logical schema. Similarly, we can define logical data independence. Logical data independence is harder to achieve, and hence changes are more possible to the external schema when there are changes in conceptual schema.
Read more

DBMS #04: Joins, Decomposition and Normalization

Checklist: > To understand data redundancies and anomalies. > To check a relational decomposition for generation of spurious records and preservation properties. > To achieve a normalized decomposition of a relation. 4.1. Natural Joins: -         To combine two relations using common attributes(s); joins only those records where the common attribute(s) match 4.2. Data Redundancy and Anomalies: -         Functional dependencies hold in the entire database, not just single relations. -         A bad database design has redundant storage of data and may allow spurious records to be generated. Update anomaly: update of an attribute in one record needs to be reflected in all records with the same attribute value. Data inconsistency arises when any relevant record is not updated. Insert anomaly: to insert a new domain value for some attribute, some record must be crea...
Read more