Data Modeling

On Line Lecture Unit 1

Notes

Part A

History and Tradition
   IBM during the Apollo project created the Generalized Update Access Method (GUAM) due to the massive amounts of data sharing that was necessary for the project. (Contractors with subcontractors with subcontractors, etc.)
   IBM then looked for a way to market this technology. DL1 was the result of this program. It was a hierarchical model for working with the data.

   In the mid 60's GE changed the Integrated Data Storage (IDS), which was a network model. Though never fully standardized like its developers wanted it to be, this model, named after the conference responsible for trying to standardize it, CODASYL, became very widely used.

   In the 1990's the relational database model (which was originally proposed in the 1970's) became more widely used.

   Other models were used as well. They include:

• Inverted File Model - Similar to the concept of a book with extensive index, with the book being the data and the index the information pointers to where data is.
• Object Oriented Model (OODBMS)
• deductive Database Model
• Knowledge Database model
• Expert systems - sort of developed from last 2, a professional would input some key words that he needed information on and system would respond with the data based on rules set forth to give him information needed to make expert decisions.

---

Part B

What is a database? - Before reading book - A set of data stored someplace , usually a computer, that is organized in some fashion and the items are related to each other.

What is Database? - After reading - Organized data stored in a structure where the data may consist of multiple type of entities, the attributes of these entities are the relationships between the entities.

As we look at things in a model we can understand:

• Entities are tables
• Attributes are the columns in the tables
• Relationships are items in the tables that would be in more than one table in a database (columns that have same data in them).

Terminology
Entity - hard to describe. It could be defined as, for example, all books or just one specific book.
Entity has to do with relationship.
table - an abstract way for us to think of the data. We need this to be able to visualize how the data is stored.
File - a way for us to hold the information.

Ways of looking at the data - if we use the term in column one then the terms in column two and three should be used to define things. Too often the terms wind up getting exchanged.

Table	Row	Column
Relation	Tuple	Attribute
File	Record	Field
Entity	Entity occurrence	Attribute

---

Part C

Lecture on the use of MSAcess as our RDBMS
Examples on how to do things that would be to complicated to reproduce here.
Note: Domain is a jargon term to get to know.

---

Part D

Hierarchical Model & CODASYL Model

So far we have been looking at the relational model of databases mostly. In it a database consist of rows and columns. Columns have 'attribute names' and domains. We use them as abstractions to let us think about them in concepts that are easier to understand. Basically a particular column in a table has a relationship with at least one other table and we can use that to store the data without having to make duplicate information.

In a hierarchical model an entity is owned by another entity. Ownership is maintained by a chain. A schema, or way of describing how data is to be laid out, defines how this layout of data is.

Sponsor
|
|
Team

In this example we can see that an entity can only have one owner. It is possible for an entity to own more than one other entity.

CODASYL model, also know as the network model removes the restriction of only one owner of an entity. The example of this might be a class at a college. It is possible for both the department and the instructor to own the class. In the hierarchical model, maybe the department would own the instructor who would own the class. This model breaks down if instructor teaches for more than one department, hence the need for the CODASYL model.