Chapter7: Session 4 - Data Modeling

Data model

- A formal way of representing the data that are used and created by a business system. - Shows the people, places and things about which data is captured and the relationships among them.

Entity

- A person, place, event, or thing about which data is collected. - Must be multiple occurrences to be an entity. Example: If a firm has only one warehouse, the warehouse is not an entity. However, if the firm has several warehouses, the warehouse could be an entity if the firm wants to store data about each warehouse instance.

What is an ERD?

- A picture showing the information created, stored, and used by a business system. - Entities generally represent similar kinds of information. - Lines drawn between entities show relationships among the data. - High level business rules are also shown.

Balancing ERDs with DFDs

- All analysis activities are interrelated - Process models contain two data components + Data flows and data stores. - The DFD data components need to balance the ERD's data stores (entities) and data elements (attributes) - Many CASE tools provide features to check for imbalance - Check that all data stores and elements correspond between models + Data that is not used is unnecessary + Data that has been omitted results in an incomplete system - Do not follow thoughtlessly -- check that the models make sense!

Relationships

- Associations between entities. - The first entity in the relationship is the parent entity; the second entity in the relationship is the child entity. - Relationships should have active verb names. - Relationships go in both directions.

Using the ERD to Show Business Rules

- Business rules are constraints that are followed when the system is in operation. - ERD symbols can show when one instance of an entity must exist for an instance of another to exist. Ex. A doctor must exist before appointments for the doctor can be made. - ERD symbols can show when one instance of an entity can be related to only one or many instances of another entity Ex. One doctor can have many patients; each patient may have only one primary doctor. - ERD symbols show when the existence of an entity instance is optional for a related entity instance. Ex. A patient may or may not have insurance coverage.

Independent Entity

- Can exist without the help of another entity - Identifiers created from the entity's own attributes - Attributes from other entities are not needed to uniquely identify instances of these entities

ERD Building Tips

- Data stores of the DFD should correspond to entities - Only include entities with more than one instance of information - Don't include entities associated with implementation of the system (they will be added later)

ERD Basics

- Drawing the ERD is an iterative process of trial and revision. - ERDs can become quite complex.

Add Attributes and Assign Identifiers

- Identify attributes of the entity that are relevant to the system under development + Check the process model repository entries for details on data flows and data stores + Check the data requirements of the requirements definition + Interview knowledgeable users + Perform document analysis on existing forms and reports - Select the entity's identifier

Identify the Entities

- Identify major categories of information + If available, check the process models for data stores, external entities, and data flows + Check the major inputs and outputs from the use cases - Verify that there is more than one instance of the entity that occurs in the system

Steps in Building ERDs

- Identify the entities - Add attributes and assign identifiers - Identify relationships

Advanced Syntax

- Independent Entity - Dependent Entity - Intersection Entity

Attributes

- Information captured about an entity. - Only those used by the organization should be included in the model. - Attribute names are nouns. - Sometimes entity name is added at the beginning of the attribute name for clarity.

The Data Dictionary and Metadata

- Metadata is information stored about components of the data model. - Metadata is stored in the data dictionary so it can be shared by developers and users throughout the SDLC. - A complete, shareable data dictionary helps improve the quality of the system under development.

Identifiers

- One or more attributes can serve as the entity identifier, uniquely identifying each entity instance. - Concatenated identifier consists of several attributes. - An identifier may be 'artificial,' such as creating an ID number. - Identifiers may not be developed until the Design Phase.

Modality

- Refers to whether or not an instance of a child entity can exist without a related instance in the parent entity. + Not Null means that an instance in the related entity must exist for an instance in another entity to be valid. + Null means that no instance in the related entity is necessary for an instance in another entity to be valid.

Identify Relationships

- Start with an entity and identify all entities with which it shares relationships - Describe the relationship with the appropriate verb phrase - Determine the cardinality and modality by discussing the business rules with knowledgeable users

Normalization

- Technique used to validate data models - Series of rules applied to logical data model to improve its organization - Three normalization rules are common

Summary

- The ERD is the most common technique for drawing data models. The building blocks of the ERD are: + Entities describe people, places, or things + Attributes capture information about the entity + Relationships associate data across entities - Intersection, dependent, and independent entities must be recognized. - The ERD must be balanced with the DFD.

Cardinality

- refers to the number of times instances in one entity can be related to instances in another entity. + One instance in an entity refers to one and only one instance in the related entity (1:1) + One instance in an entity refers to one or more instances in the related entity (1:N) + One or more instances in an entity refer to one or more instances in the related entity (M:N)

Unnormalized Entity

Begin with an entity from the logical data model

Intersection Entity

Exists in order to capture some information about the relationship that exists between two other entities. Typically, intersection entities are added to a data model to store information about two entities sharing an M : N relationship.

Dependent Entity

Relationships when a child entity does require attributes from the parent entity to uniquely identify an instance

ERD full meaning is _____

The entity-relationship diagram

Normalization

is the process analysts use to validate data models.

Data models____

should balance with process models

Physical data model

shows how the data will actually be stored in databases or files.

Logical data model

shows the organization of data without indicating how it is stored, created, or manipulated.

Second Normal Form (2NF)

If an entity has a concatenated identifier, look for attributes that depend only on part of the identifier. If found, remove to new entity.

Third Normal Form (3NF)

Look for attributes that depend only on another non-identifying attribute. If found, remove to new entity. Also remove any calculated attributes.

First Normal Form (1NF)

Look for repeating groups of attributes and remove them into separate entities