RU2455684C2 - Method of forming information models of information systems - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: in the method of forming information models of information systems, a plurality of models is further formed for determining types of information objects, a set of their properties (attributes) and relationships between objects contained in the data storage machine structure represented by the 'essence-communication' model.

EFFECT: broader functional capabilities of information systems owing to inclusion of a novel information medium therein, which enables storage of arbitrary information objects.

17 dwg

Description

The present invention relates to the field of technical cybernetics and can be used in information (automated, information-analytical, decision support) systems (IP) for organizing storage of information objects.

A known method of organizing data in IP [1], which proposes a classifier of the hierarchical structure, created in the form of an open system that allows its addition and modernization, while existing formalized data in a modernized or existing form include in their data structure.

The disadvantage of this method is the hierarchy of the created data structures, which does not allow you to store data, for example, of a network type.

There is a known method of forming an information model [2], in which the elements of the system are multi-parameter multi-level hierarchically ordered and integrated into a network graph structures and generate information about the direct and through the attributes (parameters) of the relationship of the elements of the system with a description of the relationships represented by simple and meaningful relationships with a description of a meaningful nature each.

The disadvantage of this method is that the system is focused only on information objects with a predetermined structure and it is impossible to organize a database system with an arbitrary data set with a previously unknown structure in a single database.

The closest way to the proposed one is the method of forming an information model and the method of organizing databases, referred to as the "concept of 3 schemes" [3], the implementation of which, depending on the structure of the data required for the subject area (the first scheme is an external scheme - user's point of view), a neutral conceptual scheme is created (the so-called entity-relationship model, or ER-model is the third scheme is a conceptual neutral scheme in which types of information objects are represented, their properties are attributes and relations between objects), in accordance with which a machine (computer) scheme is created (database structure - second scheme - internal scheme - computer point of view).

The disadvantage of this method is that the machine circuit is focused only on information objects with a predetermined structure. If it is necessary to develop IP and enter data of a new structure, one has to change the machine circuit or change it to a new one, which does not allow using the principle of "inheritance" of information.

The aim of the proposed method is to increase the functionality and openness of IP.

The goal is achieved by the implementation of the proposed method of forming information models and organizing the storage of information objects, providing the creation of a new type of information tool for the IS storage of information objects. The storage of information objects provides storage of data of an arbitrary structure. In order to increase the functionality and openness of IP, it is proposed:

- switch to using a concept that is different from the well-known concept of three schemes and consists in the fact that instead of one neutral conceptual scheme, two are used: the first of the new conceptual schemes should reflect the user point of view and determine the set of types of information objects, their properties (attributes) and relationships between objects, and the second new scheme, determining the types of types of information objects, should reflect the point of view of the computer and the structure of the proposed new information tool - stored conductive information objects. It is proposed to call such a concept a “four-circuit concept”;

- use a specific second conceptual scheme (structure of the storage of information objects) that provides storage of arbitrary information objects.

This approach will allow:

- increase the functionality of IP by including in their composition a new information tool that provides storage of arbitrary and predetermined information objects;

- reduce the cost of operating IS by eliminating the need for changes in database structures;

- reduce the cost of operating IS by eliminating the need to transfer accumulated data from existing databases to new ones when replacing the first (implementation of the principle of data inheritance);

- reduce the cost of creating IP due to increased level of unification of information tools; the same type of storage of information objects can be used to organize operational databases (in automation systems), data warehouses and data marts (in information-analytical systems and in decision support systems).

The essence of the invention is as follows.

The first one. A third (new) conceptual scheme is created, which is a set of information models (the name for such models is proposed - "application scheme"), describing the types of information objects with sets of their properties (attributes) and relations between them.

The second one. A fourth (new) scheme is created that defines the structure of the new information tool - the storage of information objects. A distinctive feature of the storage of information objects is that it is not oriented to the storage of information objects of a certain structure and allows you to store information objects with an arbitrary data structure.

This allows, without changing the structure of the storage of information objects, to introduce new objects with a data structure unknown at the stage of creating the storage. With the expansion of functionality and the development of the information system, there is no need to solve inheritance problems - transferring data from the old system (from a database with an old structure) to a new one (with a new database structure).

The proposed method of forming information models of information systems and organizing the storage of information objects is illustrated in figures 1 - 17.

An example of three options for a graphical subset of two applied schemes "Persons" and "Organizational units" (the third conceptual scheme) is shown in Fig. 1, Fig. 2 and Fig. 3, where 1 is the type of information object, 2 is the attribute (property) of the information object or an attribute (property) of the relationship, 3 - the relationship of the information object, 4 - the information object of another application scheme with which there is a relationship.

An example of a text subset of the three variants of the two above-mentioned application circuits is shown in FIG. 4, FIG. 5 and FIG. 6.

An example explaining the first method of organizing the fourth scheme is shown in Fig.7. An example explaining the second method of organizing the fourth scheme is shown in Fig. 8. An example explaining the third method of organizing the fourth scheme is shown in Fig. 9 An example explaining the fourth and fifth methods of organizing the fourth scheme is shown in Fig. 10. 7-10 are indicated: 5 - the entity "Information object" with attributes, 6 - "Unique identifier of the object", 7 - "Identifier of the type of the object" and a group of attributes (properties) 8; 9 - Entity "Object property" with attributes, 10 - "Unique identifier of the property" of the object, 11 - "Unique identifier of the object" (identifier of the object to which the property belongs), 12 - "Identifier of the type of property" and 13 - "Value of the property"; 14 - entity “Object relationship” with attributes, 15 - “Unique identifier of the relationship”, 16 - “Beginning of the relationship. Unique identifier of the object” (identifier of the object that forms the beginning of the relationship), 17 - “End of the relationship. Unique identifier of the object” (identifier the object that forms the end of the relationship), and 18 - “Relationship Type Identifier”; 19 - entity "Relationship property" with attributes, 20 - "Unique identifier of the relationship", 21 - "Unique identifier of the relation property", 22 - "Identifier of the type of relation property" and 23 - "Value of the relation property", 24 - additional entity "Property object "and 25, 26 - additional entities" Relationship property ".

Examples of information objects of the type "Persons" and "Organizational units" are shown in Fig.11.

Examples of relationships between instances of the "Person" type and instances of the type "Organizational unit" and instances of relations between different instances of the type "Organizational unit" are shown in Fig. 12.

Examples of tables of the storage of information objects and data in these tables in the first method of organizing the storage of information objects are shown in Fig. 13.

Examples of tables of the storage of information objects and data in these tables in the second method of organizing the storage of information objects are shown in Fig. 14.

Examples of tables of the storage of information objects and data in these tables with the third method of organizing the storage of information objects are shown in Fig. 15.

Examples of tables of the storage of information objects and data in these tables with the fourth and fifth method of organizing the storage of information objects are shown in Fig. 16, Fig. 17.

To implement the first method of organizing the storage of information objects in accordance with the entity "Information object" (Fig.7) create the first table of the storage of information objects with the first ("Unique identifier of the object"), the second ("Object type identifier") and a group of third (" Object properties ") to fields (columns). Such a structure of the storage of information objects is sufficient to store arbitrary information objects. To record in the storage of an information object of any type (in accordance with the application diagram in figure 1, and the example in figure 11 of specific instances of the type "Person" and type "Organizational unit"), records are created in the first table (Fig. 13) . In the first column, a unique identifier of information objects is generated (1 - in the example in Fig. 13 for the first instance from the list in Fig. 11, respectively 2 - for the second, 3 - for the third, 4 - for the fourth, 5 - for the fifth and 6 - for the sixth one), in the second - an identifier of the type of information object (1 - in the example in Fig. 13 for instances of information objects of the type "Person" and 6 - for instances of information objects of the type "Organizational unit" for the list of specific instances shown in FIG. eleven). In the group of third fields (columns), the values of the properties of information objects and links are formed, which ensure the implementation of relations between objects. For ease of explanation, FIG. 13 shows specific property values of only part of the properties from the example of specific instances in FIG. 11 (birth dates not shown). In the first method of organizing the storage of information objects for the implementation of relations between information objects, a special type of object property is used - a link to the unique identifier of that instance of the object with which this object relates. Such an implementation of relationships allows you to implement relationships with a power of 1-1 or 1-n and without properties. Relations with power nn and relations with properties must be “normalized” in the application scheme and reduced to 1-n relations (for this, the nn relationship must be replaced by an additional “associated” entity and two 1-n relations. If there are properties, the relation property values are indicated as properties of the associated entity). In the example in Fig. 13, relationships of the type "A person works in an organizational unit" are implemented - using the reference property (4, 5 and 6 indicate a unique identifier for an instance of an organizational unit) in the fourth column of the group of third fields - and "The organizational unit is subordinate to the organizational unit" - using the reference property in the second column of the group of third fields (4 also indicates a unique identifier for the instance of the organizational unit).

To implement the second method of organizing the storage of information objects in accordance with the entity "Information object" (Fig. 8), create the first table of the storage of information objects with the first ("Unique identifier of the object") and the second ("Object type identifier"). In accordance with the Entity Property entity, a second information object storage table is created with the first (“Unique identifier of the property”), the second (“Unique identifier of the object”, the identifier of the object to which the property belongs), the third (“Property type identifier”) and the fourth ("Property Value") by fields (columns). Such a structure of the storage of information objects is also sufficient for storing arbitrary information objects. To record in the storage of an information object of any type (in accordance with the application diagram in FIG. 1 and the example in FIG. 11 of specific instances of the “Person” type and “Organizational unit” type), an entry is created in the first table. In the first column, a unique identifier of the information object instance is generated (1 - in the example in Fig. 14 for the first instance from the list in Fig. 11, respectively 2 - for the second, 3 - for the third, 4 - for the fourth, 5 - for the fifth and 6 for the sixth), in the second - an identifier of the type of information object (1 - in the example in Fig. 14 for instances of information objects of the type "Person" and 6 - for instances of information objects of the type "Organizational unit" for the list of specific instances shown in Fig.11 ) After obtaining a unique identifier for an instance of an information object (when creating a record in the first table), records for the properties of an information object are sequentially generated in the second table, for which a unique property identifier is generated for each property in the first column, and a unique identifier of the information object instance is entered in the second column ( obtained during the formation of the record in the first table), the identifier of the property type is entered in the third column (for information object instances in the type “Face” 2 - for the property “Last name”, 3 - “First name” and 4 “Middle name”, for instances of information objects of the type “Organizational unit”, 7 - “Name”), and in the fourth - the value of the property (Specific values in accordance with the examples from the table in Fig.11). For ease of explanation, FIG. 14 shows specific property values of only part of the properties from the example of specific instances in FIG. 11 (birth dates not shown).

In the second method of organizing the storage of information objects, as well as in the first, a special type of object property is used to implement relations between information objects - a link to the unique identifier of that instance of the object with which this object relates. Such an implementation of relationships allows you to implement relationships with a power of 1-1 or 1-n and without properties. Relations with power n-n and relations with properties must be “normalized” in the application scheme and reduced to 1-n relations (for this, the n-n relation should be replaced by an additional “associated” entity and two 1-n relations). In the example of FIG. 14, relationships of the type “Person works in an organizational unit” are implemented - using a reference property with an identifier of the property type - 5 - and “Organization unit is subordinate to an organizational unit” - using a reference property with an identifier of type - 8.

In the third method of organizing storage of information objects, additionally, in accordance with the Entity Relation of Objects entity (Fig. 9), create a third table of information objects storage with the first ("Unique identifier of the relationship"), the second ("Start of relationship. Unique identifier of the object", object identifier , forming the beginning of the relationship), third ("End of the relationship. Unique identifier of the object", identifier of the object forming the end of the relationship) and fourth ("Identifier of the type of relationship") fields (columns). In accordance with the Entity “Relationship Property”, they create a fourth table of information objects storage with the first (“Unique identifier of the relation”), the second (“Unique identifier of the relation property”, the third (“Identifier of the relation property type”) and the fourth (“Relation property value” ) fields (columns). According to the third method of organization, relationships of any power can be indicated in the application schemes and relationship properties can be indicated for the relationship. Example records of specific instances of information objects relations between the tables 11 and 12 shown in Figure 15.

The fourth and fifth way of organizing storage of information objects are distinguished by the introduction of not one, but groups of second and fourth tables, which allows you to store in each of the second and fourth tables only one data type for the properties of objects. This allows us to simplify the data conversion procedures required in the first and second method.

For the third and fourth method, an example of records of specific instances of information objects and the relationships between them from the tables in FIG. 11 and FIG. 12 is shown in FIG. 16 and FIG.

As a result of the application of the present invention, in comparison with the known methods, a primary effect is achieved consisting in increasing the functionality of the IP due to the inclusion of a new information tool that provides storage of arbitrary and predetermined information objects.

The above approach will allow you to get a significant economic effect due to:

- reduce the cost of operating IS due to the elimination of the need for changes in the structures of databases, storages and data marts;

- reducing the cost of operating IS due to the elimination of the need to transfer accumulated data from existing databases to new ones when replacing the first ones (solving the problem of data inheritance).

From the use of the present invention, a secondary effect should also be expected, consisting in reducing the cost of creating IP due to an increase in the level of unification of information tools; the same type of storage of information objects can be used to organize operational databases, data warehouses and data marts.

Particularly relevant is the use of the proposal in automated information systems for organizing, based on repositories of information objects, operational databases. It is possible to use the proposal in decision support systems and information and analytical systems for organizing data warehouses and data marts.

Information sources

1. RU, application No.2000101713, class G06F 17/00, 2001.

2. RU, patent No. 2225033, class. G06N 1/00, 2004.

3. IDEF1X standard.

Claims (1)

A method of forming information models of an information system in which an entity-relationship model is formed, the elements of which are represented in a machine data storage structure, for example, an entity is usually represented by a database table, and a property (attribute) is represented by a column in the table, characterized in that additionally form a set of models for determining the types of information objects, a set of their properties (attributes) and relations between objects contained in the machine data storage structure represented by the model " uschnost - communication. "

Images