Levels of representation of data models - Databases: design

Data Model Representation Levels

The concept of multi-level database representation is the basis of modern database technology. These ideas were first formulated in the report of the working group on databases of the Committee for Standardization of Standards of the American National Standards Institute (AYZG/HZ/ZRIAZ), published in 1975. This report proposed a generalized three-level model of the database architecture, including the conceptual, external and internal levels (Figure 1.36).

Fig. 1.36. Levels of data representation (A ^ 1/X3/5PA11C)


The conceptual level of the AK51/5REAS architecture serves to support a unified view of the database structure common to all applications that use it, and independently of them. The conceptual level is a formalized information-logical model of the database.

The internal layer of the ANSI/8APX architecture determines the support for the database view in the storage environment. At this level, it appears to be completely materialized form.

The external architecture level of the ALI/5PAC11 is designed to reflect database support at the level of individual user groups. Descriptions of such database views are called external schemas. In this case, database management systems support several external database schemas for different groups of users or tasks.

Such a presentation characterizes common approaches to the implementation of databases in the process of their development and implementation, but the development of technologies has led to the need to allocate additional architectural levels of data representation, thereby defining the basic rules and concepts for designing and implementing databases (Figure 1.37 ).

Fig. 1.37. Data Model Representation Levels


By highlighting such an architecture, we cover the entire range of currently existing types of database models and data models that guide the design and implementation of the database. At the same time, such separation of levels allows more accurate determination of individual architectural elements of modern relational databases implemented in database management systems.

The analytical level of the data model presentation provides a clear understanding of the data and objects that are discussed in the subject area, which they describe. The basis of this level of representation of the data model is the objects, their attributive composition and the relationships between them.

The concept of database development determines the need for primary analysis of the domain from three positions:

• business processes and data flows, where tasks are defined that must be implemented using databases;

• documents and their structures, which defines the basic list of attributes, which should describe data in databases;

• Objects that together with documents give an idea of ​​the main data elements that should be stored in the database, and additional attributes that were not highlighted at the document review level.

Given these features of database development, the analytical level of data model representation contains three main models, where data is a key element: a data flow model based on business processes being implemented, a workflow scheme, and a model of object domain relationships. The main difficulty in constructing data models at the analytical level calls for the construction of all three models simultaneously, realizing that the information reflected in one of the models is accounted for and/or reflected in another model, in accordance with certain presentation rules (notation).

Considering the representation of data about the data at the analytical level, it is necessary to start from the terminology used in describing the subject area in business processes and documents, namely: object, attribute, function/process/task, connection, object instance, classifier and t . The terminology used allows, without going over to the level of computer systems, to provide information about data structures and their characteristics in a language understandable to functional specialists and specialists in the field of information technology.

The analytical representation of the data model is formed when performing the domain analysis and is the basic model for further design and implementation of the database.

The conceptual level of the representation of the database model reflects the abstract representation of domain data in structures that reflect the attributive composition of each element (object) of the model and the features of the links between them.

One of the main variants of the conceptual representation of the database model is the model in Peter Chan's notation, proposed in 1976. At that time, this presentation enabled the developer to specify all the necessary components of the future database and was considered at the joint conceptual-logical level of the database representation .

For modern databases that include a much larger number of elements of the structure than it was in 1976, Peter Chen's model can be attributed only to the conceptual level, displaying the main elements of the domain using the following terminology: entity (object) , communication, attribute, communication type, cardinality, relationships.

Here it is worth considering that the term Relation in the notation of Peter Chen has a somewhat excellent sense in comparison with the same term in relational algebra, where the relation was understood as any representation of the data. At the model level, the term Relation from relational algebra is specified by marking it in a separate terminology element "Entity" Relationships describing a particular domain object with a fixed set of attributes, and the term Relation called representations obtained by the interaction of related entities.

Thus, the term Relation for the model of Peter Chen (Figure 1.38) determines precisely the features of the connection between entities, expressing it with a verbal form, corresponding to the activity in the subject area. For example, when binding entities, the "Student" and Discipline establish the "Learns" relationship, illustrating the subject-matter of the interaction between learners and the disciplines they are studying.

Fig. 1.38. Example of the model in the notation of Peter Chen


The logical level of the database model representation is an extension of the conceptual model and is complemented by a set of additional elements, namely data types, constraints, defaults, etc. The logical level is a more complete representation of the database model, but compared to the conceptual view, and is formed to translate models from the domain plane into the future database.

The representation of the model in Peter Chen's notation is quite cumbersome, which is not very convenient for subsequent analysis and consideration of the database model. Therefore, for logic and subsequent levels, we use simpler notations, in terms of visualization, such as the Crow's Foot notation used in most database modeling tools

Because the logical database model is an extension of the conceptual model, all elements: entities, attributes, relationships, relations, etc., are also displayed in the diagrams of the logical model. The orientation of the logical model to the level of the abstract representation of the database and the addition of elements of the implementation level to it: default, constraints, etc., relate to specialized models with which most IT specialists work.

Further development of database models will focus on obtaining a physical database on the hard disk, which translates the developer to the level of reflecting the software and technical features of the database implementation. Nevertheless, the continuity of models from analytical to physical level allows to talk about the accuracy and correctness of database implementation in accordance with the requirements of the subject area. But do not forget about such a feature that the conceptual and logical level of work of a functional specialist over models is not limited and continues at subsequent levels in terms of clarifying individual rules for working with submitted data, being reflected in the visualized or hidden elements of models of subsequent levels: datalogical, physical, internal, external.

The data level of the database model representation reflects the features of implementing the database structure, taking into account the requirements of the selected database management system (DBMS). In addition to the database structure view at the datum level, the rules of data integrity constraints are specified, which determine the possible actions with the data when performing certain operations (addition, modification, deletion).

Moving from logical to datum level, the database developer changes the terminology used, namely: from the term Entity the transition to the term Table & quot ;, from the term Attribute - to the term Field ( Column ), etc. This transition is due to the need to consider the model in terms of DBMS, which operates at the physical level.

When passing to a datum level, the database developer must accept the basic namespace definition rules that will be applied to the database. This namespace reflects the basic requirements for how individual database objects will be named (Table 1.10).

Table 1.10

Example of a namespace description


The object/










& lt; name

tables & gt;

1 _

Name of the object Table consists of a functional name that reflects the essence of the data stored in the table, using the prefix






The object/












& lt; name

fields & gt;


The primary key of the table is made up of a functional name that reflects the essence of the identifiable instances, the table data, with the addition of the postfix


This description of the namespace will allow you to standardize the names of database objects and unambiguously interpret the names used when writing software processing and data sampling codes.

At the datum level, database developers should define the following information that is used later when implementing a physical database:

• tables and fields denoting the separation of data into atomic information and grouping into a set of fields that are inseparable by function;

• data types that define the basic rules for the representation and processing of data stored in fields, taking into account the features of the DBMS;

• restrictions on the values ​​of fields, represented as logical expressions and used in DBMS when determining the possibility of saving the corresponding value in the required field of the table;

• Defaults that are set if the value added to the table field is not provided by the add or change command;

• the primary key field corresponding to the attributes of the primary keys of the logical model of the database, but imposes restrictions on the physical layer on the possibility of making corresponding values ​​in this field;

• the foreign key field corresponding to the attributes of the foreign keys of the logical model of the database, but also, as for the primary key zero, imposing special restrictions on storing the corresponding values ​​in it;

• Value calculation formulas used for calculated fields, where mathematical or linguistic calculations apply the values ​​of the fields of the same table and records in which the field is used;

• referential integrity rules that define possible actions (Restrict, Cascade, Set NULL, Set default, No action, etc.) when performing data modification operations (adding, changing, deleting).

As modern database modeling tools have good enough capabilities to move from one level to another and to a physical database, the correctness and accuracy of compiling


The data model of a database is the most important step in the development of a database. It is at this level that the last opportunity to correct many mistakes made during the logical modeling of the database is given, and to proceed to the implementation of the physical database in the DBMS.

The physical layer of the database model is an extension of the datum level, presenting a description of the rules for processing data taking into account the characteristics of the selected DBMS. Depending on the database modeling tool used, the degree of deepening to the physical level of modeling can be different.

In general, the physical modeling layer assumes a description of the data structures, but the views used in the database, program code templates for stored procedures and triggers. In fact, the physical layer is represented by the software level of the database representation.

In the process of developing a database, users' information needs in the data are analyzed using fixed data for sampling (static queries). Such user needs require the developer to compile a model of standard samples that are implemented in a database model in the form of a presentation description. However, it should be borne in mind that the database model does not reflect the process of obtaining data, but only shows the structure (set of fields) of the view and the links to the source tables that must provide information for entering into the view. Often this is sufficient, since static representations are usually used to implement a selection of data from one table (usually, reference classifiers) or several tables to select a complete set of information from all the participating tables.

Also, views can be used to provide user access restrictions to certain data. Taking into account the possibility of modifying some variants of representations, such form of access delimitation will give flexible opportunities to restrict users' work only to the information that they should be available.

Dynamic views that involve sampling data, but the user's request, based on the original information supplied by the user, can not be implemented using the standard view engine and the stored procedure mechanism is applied to them. This mechanism involves working in two modes (depending on the DBMS):

- Creation of the code for defining the view and its subsequent use in the user's requests;

- execution of the query using the language with the result returned in accordance with the user-defined data types corresponding to the set and field properties, the values ​​for which should be returned to the result of the selection.

The use of these mechanisms also simulates the creation of a database model at the physical level by using templating tools included in the appropriate database modeling tools.

In addition to representations, database modeling tools are often provided with the ability to define software code templates for data processing, not only using sampling requests, but also using various programming language constructs, among which are: conditions, loops, etc. Such a representation of procedures processing allows developers to reorient the physical model of the database to any supported DBMS with minimal costs and to transition from one DBMS to another.

The internal layer is represented by the database implementation in the DBMS and is formed on the basis of the physical model of the database. The implementation of all the elements of the database at the internal level allows the developer to perform not only the structuring and description of the rules of the database, but also to make basic basic information in the database, adjusting it to the initial work of the user.

As a basic data, information on classifier directories and some domain-specific data that were previously defined at the analytical level of the database modeling are usually considered.

The external layer represents the implementation of the database, taking into account the availability of individual objects and information to specific users and their groups. Usually this is realized through the organization of database roles, which determine the capabilities of performing operations (adding, modifying, deleting, selecting), obtaining individual data by views, accessibility of interaction with tables and fields (columns), etc.

Because database modeling tools do not provide external modeling capabilities, all modeling is considered at the analytical level and represented by appropriate descriptions that are subsequently transformed into external database presentation rules.

Given the need to review the database at all levels and use different terminology, one of the most important tasks of the developer is to correctly apply terminology when describing the elements of the corresponding models (Table 1.11).

thematic pictures

Also We Can Offer!

Other services that we offer

If you don’t see the necessary subject, paper type, or topic in our list of available services and examples, don’t worry! We have a number of other academic disciplines to suit the needs of anyone who visits this website looking for help.

How to ...

We made your life easier with putting together a big number of articles and guidelines on how to plan and write different types of assignments (Essay, Research Paper, Dissertation etc)