Abstract structure and evolution of database systems
Standard structure of the database
Three levels of DBMS data management
1.External Level = User Level
a user can see the data processing, in a certain part of the global data
2.Conceptual Level = Logic Level
appreciated that the data / management from a global perspective, containing respective associated constraints
3.Internal Level = Physical Level
data stored on the medium, including the storage path, storage, indexing, etc.
And the data structure of the data - Mode
Mode (Schema)
describes a structured data in the database performed
configuration information of the observed data
view (View) / data (Data)
data manifested in a form of a database
Three two-mode image
Three modes (three views)
External Schema ---- (External) View
a user can see the structure described processing
(Conceptual) Schema ---- Conceptual View
from a global point of understanding the data structure / management is described, containing constraints embodied in the respective associated data the inherent nature of the contact between the
internal schema ---- internal view
data stored on the medium description of the structure containing the storage path, storage, and other indexing
mode refers to the concept default mode, the default view of an external view of means
Two image
EC Mapping: External Schema-Conceptual Schema Mapping
external mode mapped to the conceptual model, to support for data conceptual view to converting an external view of
facilitating a user to observe and use
CI Mapping: Conceptual Schema-Internal Schema Mapping
the conceptual model is mapped to the model, enabling conversion of data into a conceptual view of the internal view
to facilitate storage and processing computer
Two independence
Logical data independence
when the concept of mode change, without changing the external mode (only change EC Mapping), the need to change the application
of physical data independence
when the internal mode is changed, without changing the concept of mode (simply by changing CI Mapping) eliminating the need to change the external mode
Data Model
The difference between the data model, data model, data
What is the data model
models 1. DESCRIPTION OF uniform predetermined pattern, comprising: a data structure, operations and constraints
2. The data model is an abstract model of the structure itself, the pattern data itself is an abstract structure
, such as:
the relational model: The model used abstract may be in the form table (table) [data structure], and each has a specific pattern is different from the specific table column names. This form of data tables and constraints which operate.
(Personal understanding: the relational model defines the structure and performance of all possible operational data, such as union, intersection, and so on, there are constraints on the model data, and model the relational model is the concrete manifestation of the relational model, but not complete data, it only specifies what data attributes)
data model is an abstract data model, data model is an abstraction of the data, so that data model structure is the structure of the data.
Three classical data model
Relational model: data table organized
hierarchical model: data organized in a tree
mesh model of: organizing the data in the form of FIG.
Evolution and Development of the database system
A brief history of the development: four stages
The first stage: database systems technology exploration phase (59-65 / 67)
1. The successful development of the file system format
2. The official presentation of the "Data Base" and began to study
the second stage: the establishment phase of database technology (65 / 68-75)
1. the three databases: hierarchical, network and relational databases camera proposed and conducted in-depth research
2. appeared and commercial database applications, but mostly mesh and hierarchical system
3. database research form the theoretical basis: relational database theory
third stage: database technology maturation phase (76-80s early)
1. the proposed standardized database system structure model of
the rapid development of relations 2. DB system: such as SQL, QBL, system R, ingres, etc.
3. the theory of improving relations, including normalization theory, the relationship between language, design and implementation of RDB, the new relational model, etc.
4. database application has been very popular, penetrated into all aspects of society, many technical branches appears DB, DB towards full maturity, called the 1970s the "decade of the database"
fourth stage: database technology to deepen the development stage (since 1985)
1. database stepwise theory, database design theory ever done
2. The new data model, the model-specific data, special type, new database system, emerging
3. Database technology + == computer technology for other industries dedicated database.
Several Major Developments
By the file system to the database
The method of operating the data management system is a file system:
the advantages of the file system: the user (program) without regard to the physical details of the file is stored, and out of the processing burden on the physical complexity of the access device.
Lack of file systems: data and programs closely integrated, organization and semantic data is closely dependent on the application processes the file, change the structure of the data you must modify the application, no links between files, no links between log files, share poor, engage in redundancy, high inconsistency.
Important development: by the file system database to
database system:
1. Uniform accessed by the DBMS, data maintenance organization and semantics can be strongly independent of the application (logical and physical independence of data independence)
2. Data and define and describe the data structure separate from the application out to the DBMS, such that a plurality of operations applications share data and class data structure, data access may be recorded as a unit, may be set in units of data items and recorded
3 unified data control function, a high degree of data sharing: the system can automatically check the security, integrity, and correctness of concurrent
with a mutual correlation between the overall data structure of 4, between the document (the Table), recording, redundant data of small, easy to expand
5. independent of the application of high-efficiency queries / statistical operations.
By the hierarchical model database, network model database to a relational model database
Hierarchical model database mesh model database and
the association between the data to maintain 1. pointer system by the complex, the complex structure description
2. Data retrieval operation dependent on the path indicated by the pointer system
3. The operation of recording one by one, can not effectively support the set operation
database relationships
relationship between data characterized by the value of the attribute in Table-described structure is simple: Table / relation
data retrieval operation does not depend on the path information or process information, non-support of the process data operation
effectively support a set of records operating
more comprehensive theoretical basis
By the relational database to the object-relational databases, object-oriented database
Relational database
1. organized in rows columns Data: 1 relationship Paradigm
2. The characteristic data can not be divided item
3. relational operators: relational algebra, calculus tuples, domain calculus - "Standard the SQL
4. Relational Database Design theoretical
object - relational database
1 can effectively support the paradigm does not satisfy the relationship between the first data item
2. the items need to be packaged objects decomposed
3. the row and column objects objects: Object structure aggregate object and
object-oriented database
1. Object oriented binding techniques (OO) and geometric / gather operations (SQL) technology
2. support complex data types, data encapsulation with the abstract data structure
3. supports object-oriented features: classes, inheritance, encapsulation, polymorphism. . .