Marilyn A. Bosito                                    06/10/08

O0A

Book: Modern System Analysis and Design Fourth Edition

Author: Jeffrey H. Hoffer, Joey F. George, Joseph H. Valacich

Reference: LRC Extension (QA 76.9 S88 H64 2005)

Chapter 3

 This chapter shows how to develop the data components of a high level Enterprise model. This model serves as a high-level architectural view of data base which can be elaborated in more details. In this chapter will discuss also how to identify the entities that organization needs to record information about, identity relationship between entities and identify the cardinality of a relationship.

A strategic architecture needs a model of data base. It is translated into the detail definition of files, records and fields that go into physical data base. The enterprise model concentrates on the “Big Picture” rather than on details of a design. This stage of modeling process is designed to identify the basis structure of enterprise, identify the data requirements, identify the activities that need to be supported, and set priorities for implementation. At this stage a model is needed that give us three things, first it should give us the name of necessary data groups in the database, second it should give us indication of which data groups are used together and last it should give us some idea of relative number of occurrences that can be expected in each data group.

A graphic notation is used to record identities and relationships in the database model. Graphic notation is used because it packs so much information into a succinct, easily understood form. The analyst needs to define the basic elements of the data file which contains the number of records. Data about single type of entity are normally stored in rectangular tables. Each row is uniquely identified by one or more key attributes. Entities can though of us molecules or building blocks out of which we construct the organization and all that it contains and support. The relationship ire like intermolecular attractions that binds the entities into an organize unit.

Entities are the primary thing of a business about of which users need to record data. They fall into six classes: People who carry out some function; Things are tangible physical objects or groups of objects; Places that are set aside for use by a people or thing; Organizations are formally organized collection of people, places, or things having a define mission; Events are things that happens to some other entity at a given date and time or as steps in sequence; Concepts are intangible principles or ideas that an organization uses and keep track of. As with all models, entity types are an abstraction of the real world entities their represent.

In naming entities it is convenient to use the plural form of the noun to name the entity. The singular form can be used to refer a single occurrence of the entity.  This increases the need for documentation, increase the time need to learn the system, increases the chances of confusion and increases the likelihood of programming error when the name is eventually used as the name of a data file. The name should be simple, clear, unambiguous, and free of jargon.

Fundamentals and attributes, frequently entities are describe by other subsidiary entities in hierarchal fashion. In such case data are stored about an entity occurrence and some other data are stored about each subsidiary entity occurrence that describes it. It is very important that all of the fundamentals entities be identified during this initial stage of the design. Other methods are used to identify missing attributes entities during later phase of the design.

Summary entities a special kind entity occurs when summary data must be recorded, summary data about individual’s items are frequently needed. After the entities have been identified, the relationships between them need to be identified. A relationship is a business action between two associated entity types. The relationships represent some interaction between the corresponding entities in the real world.

The minimum data needed about a relationship is the key attribute of each entity involved in the relationship This information record the fact that a relationship exist and allow as to link to the related entities. Recursive relationship occurs occasionally a special relationship, an entity related to itself, N-ary relationship a single relationship between more than two entities at the same time.

Cardinality is an important property of a relationship that must be identified. It tells as the maximum number of occurrence of one entity type that is linked to occurrence the other entity.

Finding information, there are five sources of information about entities and relationships: existing forms, existing file structure, published examples, strategic plans, and example for a hypothetical university. Existing forms are the common method of data collection, existing files redesign one or more existing system, these are called legacy system, published example in text  journals and books, strategic plan concentrate on the more global and glamorous topics such as markets and mergers; strategic information systems were initially used to mechanize tedious clerical task. The generators identify five strategic thrusts or direction that can be taken and five targets to which they can apply: product differentiation, cost minimization, innovation, growth, alliance, enterprise suppliers, distribution channels, customers, and rival firms.

CASE tools reduce the amount of work required to produce Entity-Relationship Diagrams, it help to identify errors in the diagrams, and capture information in the diagram in later used.