Data Modeling Fundamentals

Data Modeling Basics

• Data modeling involves the creation of a specific data model for a determined problem domain
• A Data model is a simple representation of complex, real-world data structures and useful for supporting a specific problem domain
• Models are abstractions of more complex real-world objects or events

Importance of Data Models

• Data modeling facilitates communication between different stakeholders
• Provides different perspectives of the database of data to various users
• Data modeling organizes data for various users
• Data modeling provides an abstraction for creating a database

Data Model Building Blocks

• An entity is a person, place, thing, or event about which data is collected and stored
• An attribute is a characteristic of an entity
• Relationships are associations among entities
  • One-to-many relationships include notations of (1:M OR 1..*)
  • Many-to-many relationships include notations of (M:N or ..)
  • One-to-one relationships include notations of (1:1 OR 1..1)
• A constraint is a restriction placed on data to ensure data integrity

Business Rules

• Business rules are brief, precise, and unambiguous descriptions of a policy, procedure, or principle
• They create and enforce actions within an organization's environment
• They establish entities, relationships, and constraints

Discovering Business Rules

• Sources for business rules include company managers, policy makers, department managers, written documentation, and interviews with end-users
• Recognizing and documenting business rules is essential for standardizing a company's view of data
• Documenting business rules facilitates communication between users and designers
• Documenting business rules assists designers in data modeling by understanding the nature, role, and scope of data and business processes, developing relationship rules and constraints, and creating an accurate model

Translating Business Rules

• Business rules are the basis for identifying entities, attributes, relationships, and constraints
• Nouns translate into entities
• Verbs translate into relationships among entities
• Relationships are bidirectional
• Identifying a relationship type involves questioning how many instances of B are related to one instance of A and vice versa

Naming Conventions

• Entity names requirements should be descriptive of the objects is the business environment
• Terminology used in entity names should be familiar to users
• Attribute names require a descriptive representation of the data
• Proper naming facilitates communication and promotes self-documentation

Hierarchical and Network Models

• Hierarchical models were developed to manage large amounts of data for complex manufacturing projects
• Represented as an upside-down tree containing segments equivalent to file system record types
• One-to-many relationships are depicted
• Network models were created to represent complex data relationships effectively
• Improved database performance and imposed a database standard
• Allows a record to have more than one parent
• Standard database concepts that emerged with network models are still used in modern data models
  • Concepts include: schema, subschema, data manipulation language (DML), and data definition language (DDL)

Relational Model

• The relational model led to an automatic transmission database that replaced standard transmission databases
• Based on relations (tables): matrices with intersecting tuples (rows) and attributes (columns)
• A precise set of data manipulation constructs is described
• A Relational Database Management System (RDBMS) performs basic functions like hierarchal and network DBMS system
• RDBMS makes relational data model easier to understand and implement
• Hides the complexities of the relational model from the user
• SQL-based applications have an end-user interface that allow interaction with data
• Tables are stored within the database, each independent, with rows related via common attribute values
• The SQL engine executes all queries

Entity Relationship Model

• The Entity Relationship Model has a graphical representation of entities within a database structure
• Entity relationship diagrams (ERD) use graphic representations to model database components
• Rows in a relational database are based on entities
• Attributes describe particular characteristics an object
• Connectivity is a term used to label the relationship types in a diagram

Object-Oriented Data Model

• Data and its relationships are both within a single structure known as an object
• Object-oriented Database Management Systems (OODBMS) are based on OODM
• Objects contains data and their relationships with operations
  • Basic building block for autonomous structures
  • Abstraction of real-world entity
• Attribute describes the properties of an object
• A class is a collection of similar objects with shared structure and behavior in a hierarchy
• A class hierarchy resembles an upside-down tree with each class only having one parent
• Inheritance occurs when objects inherits methods and attributes from classes above it
• Unified Modeling Language (UML) describes sets of diagrams and symbols to graphically model a system

Object/Relational and XML

• The Extended Relational Data Model (ERDM) supports OO features, extensible data types based on classes, and inheritance
  • The object relational database management system (O/R DBMS) is based on ERDM
• Extensible Markup Language (XML) manages unstructured data for efficient and effective exchange of structured, semistructured, and unstructured data

Big Data and NoSQL

• Big Data aims to find new and better ways to manage large amounts of web and sensor-generated data
• Big Data provides high performance at a reasonable cost
• Characteristics of Big Data: volume, velocity, and variety
• Challenges occur with Big Data because volume doesn't allow usage of conventional structures and is expensive
  • OLAP tools proved inconsistent dealing with unstructured data
• New technologies of Big Data: Hadoop, Hadoop Distributed File System (HDFS), MapReduce, NoSQL
• NoSQL databases are not based in relational models and support distributed database architectures
• NoSQL provide high scalability, high availability, and fault tolerance
• They support large amounts of sparse data, geared for performance rather than transaction consistency
• Provides a broad umbrella for data storage and manipulation

Data Model Summaries

• Data modeling is an abstraction of a complex real-world data environment
• Several types of data models exit: hierarchal, network, relational, object-oriented, extended relational, XML
• Data-modeling requirements are a function of different data views and level of data abstraction

Hierarchical Model

• Advantages: data sharing, parent/child relationship simplicity and integrity
• Limitations: Requires knowledge of storage and hierarchical processes, structural changes require changed application programs, implementation limitations due to no data definition or lack of standards

Network Model

• Data Integrity is promoted via data owner/member relationships
• Includes data definition language (DDL) and Data Manipulation Language (DML)
• Disadvantages: System complexity limits efficiency, complex implementation, and structural changes require changes in all application programs

Relational Model

• Advantages: Structural independence is promoted using independent tables where query capability is based on SQL
• Limitations: Requires hardware and system software overhead, conceptual simplicity may lead to information problems

Entity Relationship Model

• Advantages: Conceptual simplicity through visual modelling
• Limitations: limited representation of constraint and relationships, no data manipulation language, loss of data through attribute removal

Object-Oriented Model

• Advantages: Semantic content is added, along with inheritance to promote data integrity
• Disadvantages: a complex navigational system can cause slow development of standards

NoSQL

• Advantages: high scalably and fault tolerance are provided through low cost hardware that supports big data
• Disadvantages: there is complex programming with no relationship or transaction support, with data eventually consistent

Degrees of Data Abstraction

• End-User View, Designer’s View, DBMS View
• Data Abstraction includes: The External Model, the Conceptual Model, Internal Model, Physical Model

External Model

• The external model is the end users’ view of data environment
• People use application programs to manipulate data
• Diagrams are used to represent views, with external schema being a specific representation

The Conceptual Mode

• Global view for entire database by organizing
  • Conceptual schema is the basis for identifying and high-level description of the main data objects
• Macros of data environment
  • advantages are software and hardware independent

The Internal Model

• Representing database as seen by the DBMS mapping conceptual model to the DBMS
• Internal schema: specific representation of an internal model, using database constructs supported by the chosen database
• Logical Independence: changing internal model without affecting the conceptual one
• Hardware independent: unaffected by the type of computer on which it is installed

Physical Model

• Operates at lowest level of abstraction
• Describes how data are saved on magnetic, solid state, or optical media with definitions of storage and access methods that are software and hardware dependent
• Physical dependence means changes don't impact the internal model

Explore the fundamentals of data modeling, including the creation of data models for specific problem domains. Understand the importance of data models in facilitating communication, organizing data, and providing database abstraction. Learn about key building blocks such as entities, attributes, and relationships (one-to-many, many-to-many, and one-to-one).