Search Page

Showing 1 - 2 of 2
DM-34 - Conceptual Data Models

Within an initial phase of database design, a conceptual data model is created as a technology-independent specification of the data to be stored within a database. This specification often times takes the form of a formalized diagram.  The process of conceptual data modeling is meant to foster shared understanding among data modelers and stakeholders when creating the specification.  As such, a conceptual data model should be easily readable by people with little or no technical-computer-based expertise because a comprehensive view of information is more important than a detailed view. In a conceptual data model, entity classes are categories of things (person, place, thing, etc.) that have attributes for describing the characteristics of the things.  Relationships can exist between the entity classes.  Entity-relationship diagrams have been and are likely to continue to be a popular way of characterizing entity classes, attributes and relationships.  Various notations for diagrams have been used over the years. The main intent about a conceptual data model and its corresponding entity-relationship diagram is that they should highlight the content and meaning of data within stakeholder information contexts, while postponing the specification of logical structure to the second phase of database design called logical data modeling. 

DM-36 - Physical Data Models

Constructs within a particular implementation of database management software guide the development of a physical data model, which is a product of a physical database design process. A physical data model documents how data are to be stored and accessed on storage media of computer hardware.  A physical data model is dependent on specific data types and indexing mechanisms used within database management system software.  Data types such as integers, reals, character strings, plus many others can lead to different storage structures. Indexing mechanisms such as region-trees and hash functions and others lead to differences in access performance.  Physical data modeling choices about data types and indexing mechanisms related to storage structures refine details of a physical database design. Data types associated with field, record and file storage structures together with the access mechanisms to those structures foster (constrain) performance of a database design. Since all software runs using an operating system, field, record, and file storage structures must be translated into operating system constructs to be implemented.  As such, all storage structures are contingent on the operating system and particular hardware that host data management software.