Basic concepts and definitions - Databases

Basic concepts and definitions

A classification is understood as the division of a set into subsets according to an informally suggested feature. Due to the multifacetedness of databases and DBMS (a complex of technical and software tools for storing, searching, protecting and using data), there are many classification features. Classification of the database by the main of them [28] is shown in Fig. 1.12.

Fig. 1.12. Classification of databases

Especially we will talk about static and dynamic databases.

In static databases , the refresh rate is much lower than the read rate. Data is not directly related to time. For example, personal data, which are used much more often than they change. It was for static databases that the database theory was originally built.

In a static database, the data changes more often than is added. In this regard, the main requirement for them was the ease of updating, which is achieved by dividing the table into several in the process of normalization. The first, second and third normal forms are formed most often, the fourth and fifth are the rarer. The use of the second and higher forms indicates a profound normalization.

Recently, we are increasingly turning to dynamic databases , in which the frequency of reading and updating the data is commensurable. In dynamic databases, time appears explicitly in the form of time (date) or time interval (semester, month, year). For example, data on the progress of students in groups during the training (by semester).

For such databases, the change is not typical, but the addition of data and the procedure of deep normalization loses relevance. Normalization can be limited to the first normal form.

Separately, you should classify database management systems (Figure 1.13). Databases can also be classified in terms of economic 181: - free and paid (non-profit, commercial); by the form of ownership (state, non-state); by the degree of availability (public, with a limited number of users).

Fig. 1.13. DBMS classification

Currently, there are additional, uncommon sentences [3] for identifying classes such as active databases (in fact - expert systems, formerly called deductive databases); spatial databases (associated with the storage of graphic files, for example, geographic maps); Temporary DBs in which time is explicitly present as table fields. Since the last two classes are part of other classes and do not yet have independent significance, we will abstain from this classification line.

In the future we will focus on video databases of data (tables) with structured data of documentary nature. It's about open DBMS, usually an operating type.

Until the mid-90s of the XX century. the database was understood as static databases, which were later called operational (transaction) databases, and abroad - OnLine Transaction Processing (OLTP).

By the mid-1990s, there was so much chronological information in OLTP databases that the volume of databases increased sharply, and performance began to fall. For example, in the work of the dean's office, more detailed information is required about the current school year. At the same time, the database also contains historical data for previous years. Such data are needed much less frequently and more often in aggregated form. For example, give the names of students who have received only excellent marks for the last three semesters.

It became clear that the retrospective information should be periodically transferred to a separate database. In addition, it turned out that the retrospective data has a new quality: they allow us to develop strategic solutions. There was the possibility of forming systems for supporting the adoption of strategic decisions (DSS). Such systems were obtained abroad by the name OnLine Analytic Processing (OLAP).

OLTP and OLAP communication is shown in Fig. 1.14. OLAP serves as a complement to OLTP.

Fig. 1.14. OLTP to OLAP ratio

OLAP was immediately appreciated by managers, because it allowed: to significantly increase the efficiency of the work of managers of various ranks; increase the competitiveness of firms; receive an additional high profit.

The above characteristics of the OLAP system allow us to note that when designing this system, as in the design of OLTP, there are difficulties, determined by the following circumstances:

• the need for significant computing resources to handle a large amount of data;

• problems related to the incompleteness (lack of data) or lack of necessary fields in the operating database;

• Increased complexity of maintenance and restructuring;

• A significant duration of creation and filling of OLAP.

OLAP, in comparison with OLTP is characterized by several other properties, as seen from Table. 1.13.

In subsequent discussions, in the absence of a special reservation, we will understand the DBMS as an operational database.

Table 1.13

OLTP data properties and OLAP

thematic pictures