JPS583027A - Semantic information processing system on data base - Google Patents

Abstract

PURPOSE:To unifiedly control an application attribute given by an application system and a basic attribute given by a data base system, by holding a semantic attribute of a table and a file in a data dictionary. CONSTITUTION:Basic attributes 11, 13 are defined to be necessary for controlling a data base, by a data base system 2. Also, application attributes 12, 14 are defined originally by an application system 3. The data base system 2 executes its processing by referring to the basic attributes 11, 13 only, and also in case when a new table or field has been generated, each attribute 11-13 and 14 is genetically transmitted for its generated table or field under a prescribed condition. The application system 3 is capable of referring to each attribute 11, 12, 13 and 14.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a data-based semantic information processing method,
a data processing system comprising, in particular, a data pace-system for managing data paces and an application system for utilizing nine search results obtained from said data pace; the basic attributes defined by the data pace system and the applied attributes defined by the application system;
The database system inherits the basic attributes and applied attributes for newly created tables and newly created fields in the /1*ti table under predetermined conditions in response to processing. The present invention relates to a semantic information processing method in a database that centralizes data pace management.

In general, in database management systems that manage databases, data format attributes (field definition information, etc.) and attribute values (data) have traditionally been defined. Defining the semantic attributes of data (data units, comments, key information, etc.) is traditionally an application
It was considered to be a system-side problem and was excluded from the functionality of the database system. In the conventional method, depending on the data processing content, the application system is a database, ! : Ha1l11. Seki 9kl
In addition to the problem of having a file that stores the above-mentioned semantic attributes, there is also the problem of rounding, which involves linking data on a database with its semantic attributes through frequent and complicated processing, and even the semantic attributes. It is extremely difficult to centrally manage databases, including databases.

The present invention aims to solve the above problems,
The semantic information processing method in the database of the present invention is as follows:
1'':) A system that includes a database in which a table constituted by one field or a plurality of fields is stored as one unit, and a system for managing the data and a system for managing data obtained from the database. In the data processing system, the data dictionary unit that holds dictionary information regarding the database includes a table basic 1 property and a table application attribute of the table information regarding the table. The database system stores field semantic information including table semantic attribute information and field basic attributes and field applied attributes of the field information for each of the above fields, and the database system stores the table semantic attribute information and the field basic attributes of the field information. and the application system has a function to define the table applied attributes and the field applied attributes, and the data pace Φ system defines the table semantic attributes and the field semantic attributes in advance. A new table and/or table has a function of inheriting a new field under specified conditions, and the data on the data pace and the semantic information regarding the data are managed together. This will be explained below with reference to the drawings.

Figure g1 is an explanatory diagram for explaining the processing concept of one embodiment of the present invention regarding a database system and an application system, Figure 2 is a configuration of two embodiments of the data dictionary section of the present invention, and Figure 3 is a diagram showing the configuration of the data dictionary section of the present invention. This shows an example configuration of .

2 in FIG. 1, 1 is a data base (or data pace storage 1s), 21rJ: is a data base system that manages the database, and 3 is an application system that stores data.・Processing, analysis, editing, and display of the search results for the base; 4 is the reference function for the database system fJ4c; 5 is the update function; 6 is the same genetic processing function; and 4 is the application system. Reference function 8 also represents an update function.

Furthermore, on the database, as described later, data is 1
The data is stored in the form of a plurality of tables each having one or more fields, and dictionary information regarding the data is stored as a data dictionary. The data dictionary includes a table semantic attribute (or its storage area) 9 prepared for the table (or relation y).
and a field semantic attribute (or its storage area) 10 prepared for the field. Further, stored in the table semantic attributes 9 are basic attributes 11 such as comments and key-field information, and applied attributes 12 that are uniquely defined and used by the user. Further, within the field semantic attribute 10, there are a basic attribute 13 portion and an applied attribute 14 portion that are substantially similar to those of the table semantic attribute 9. In addition, the above semantic attributes (basic attributes,
Application attributes) will be explained in detail with reference to FIG.

In the present invention, the data pace system 2 (1) refers to the table basic attributes 11 and updates and defines them) and inherits the contents, (11) inherits the contents of the table application attributes 12, and (1) inherits the contents of the table application attributes 12; it
) refer to the field basic attribute 13 and update it) and inherit its contents; (1v) inherit the contents of the field applied attribute 14;
At the same time, application-system 3 (V)
Refer to table basic attribute 11, (vi) Refer to table applied attribute 12 and update, (→ Refer to field basic attribute 13, (4) Field applied attribute 1
Please refer to 4 and update it.

That is, basic attributes 11.13 are defined as necessary for the data pace system 2 to manage its own processing, that is, the data pace.

Further, application attributes 12.14 are independently defined by the application system 3. Then, the data pace system 2 performs processing by referring only to the basic attributes 11 and 13, and when a new table or field is created, the data pace system 2 processes the created table under nine predetermined conditions. and fields, each attribute 11, 12,
13°14 will be inherited. Further, the application system 3 is enabled to refer to each attribute 11, 12, 13, 14.

FIG. 2 shows the configuration of an embodiment of the data dictionary section.

For each table (relation), in addition to formal attributes such as ω relation name (REI:LNAMIE) and (11) relation yΦ type (RELTYPg), (-) table semantic attribute part (RIIELAPL) is added.
) or (9) multiple field sections (FtgLD
).

As shown in FIG. 2, each field section (FiELD) includes a semantic attribute section (FLDAPL) in addition to formal attributes such as field name and field type. The semantic attribute section (FLDAPL) includes a basic attribute section including field basic attribute length 15, type 16, and comment 17, and an applied attribute section including field application attribute length 1 day and applied attribute information section 19. It is composed of The basic attribute section corresponds to the basic field attribute 13 shown in FIG. 1, and the applied attribute section corresponds to the applied field attribute 14 shown in FIG.

In addition, the table semantic attribute section (RELA) shown in Fig. 2 (4)
PL) contains a basic attribute section including table basic attribute length 20, comment 21, key field information 22, and table application attribute length 2, as shown in FIG.
3. It is composed of a z-rigidity information section 24t and an application attribute section including. @The basic attribute section corresponds to the table basic attribute 11 shown in FIG. 1, and the applied attribute section corresponds to the table applied attribute 12 shown in FIG.

The basic attribute part 11.13 and applied attribute part 12.14 shown in Figures 2 and (C) are variable-length character strings or variable-length character strings, respectively.
It is stored in a hexadecimal format and is referenced or updated (defined) by the data e-pace system 2 and application system 3 in the manner shown in FIG. Needless to say, the applied attribute information is information for giving meaning in the application system 2 by giving a predetermined comment to each content of a predetermined table or field.

The manner in which the data pace system 2 inherits the table semantic attribute 9 and the field semantic attribute lOt described above will be described below.

Table semantic attributes 9 and field semantic attributes 1 as described above for fields in existing tables and our tables
When 0 is given, these tables and fields are used to generate a new table, or when a field is given and merged in the new table, the new table or new field is It is quite complicated to newly prepare the corresponding semantic attributes 9 and 10. For this purpose, it would be convenient if each semantic attribute could be inherited for a new table or new field under a predetermined condition.

For this reason, the database system 2 performs genetic inheritance as follows. That is, (4) genetic processing regarding table semantic attributes.

Inheritance is made only when a search is performed such that a new table is generated from one table.

Genetic processing is not performed when a search result is generated as a new table by combining a plurality of tables (such as through a jig-in process). This is because there is no predetermined processing method for how a table is generated from a plurality of tables. Genetic processing for key field information among the basic attributes is performed only when the search target appears in the form of a single field, and is not performed when calculation elements such as arithmetic expressions are added.

Even in this case, if the search target appears in the form of a single feed, the search is performed only in the spring. Of course, regarding this inheritance, even if, for example, field α in a table person is set to field m in a new table C, and field β in table B is set to field n in the new table C, the semantic attribute of field α is set to field m. In addition, the semantic attribute of field β can be inherited by that of field n.

When performing genetic processing such as the Queen, the semantic attributes shown in Figure 2 ■ and the semantic attributes shown in Figure 2 0 for existing tables and fields, and the semantic attributes shown in Figure 2 ■ and 0 for newly generated tables and fields are used. It can be considered that the data is transferred to the corresponding location.

However, regarding basic attributes belonging to table semantic attributes,
Key field information (attributes are defined in the form shown in FIG. 2 (D-1). That is, for example, for table A, the field "l l "8 *αrr, HHH-
... Assume that a- exists, and now the field all) E
It has the attribute of φ1 key type and the field smell is ◆2
It has a key type attribute and the field smell is φ3 key.
It has an attribute of type, and field a6 is φ6 key.
When we consider a turtle that has the attribute of type, Figure 2 (D
-1) Key field information is provided as shown.

That is, a bit string indicating the existence of each key type (6 in total)
The attribute is designated by associating the bit (bit) with field information for the key type for which the pit is on, thereby reducing the storage capacity of the basic attribute section 11 shown in FIG. Therefore, genetic processing regarding the key field information 22 is performed in the following manner. That is, field G1 in the current table is set as field rl in new table RK, and field G1 in table AK is set as field rl in new table RK.
Let field rl in table person be field rl in table person, and field rf in new table R.
If we are to generate a new table R such as
2) Key field information as shown is created. In this case, a verbatim copy of the key types φl to φ6 is made and the field rl.

The corresponding key types are indicated for rg and ys, respectively. By doing so, not only the key type to be searched can be reflected in the results, but also the middle type (φ3 in the illustrated case) that is missing during the search process can be detected.

FIG. 3 shows the configuration of an embodiment of the present invention. Code 1 in the figure,
2.3 corresponds to FIG. 1, 25 is a data processing device, 26
is a table definition function section, 2F is a semantic attribute (basic attribute) definition function section, 28 is a search processing function section, 29 is an attribute (applied attribute) definition function section, 30 is a processing/analysis/editing/display processing section, 31 is a data dictionary, 32-1.32-2.・Songs respectively represent tables, 33 represents the data dictionary contents transferred to the buffer of the main storage device, 34 represents the contents of the search target table also transferred to the buffer, and 35 represents the contents of the table that generated the search results. There is.

The table held on the data pace storage unit 1 is, for example, the one shown in the illustrated table 32-1.
NO), employee name (NAME), department/section number (DNO)
), salary (SAL), etc. In the illustrated example, "Mr. Yamada with ENO = 70101 is D.
The following information is summarized in the table: NO = 500, belongs to the department, and 5AL = 350. Of course, information regarding the corpse table 32-1 is prepared in the data dictionary 31 in the form of the 21st prisoner illustrated.

``When the illustrated table 32-1 is retrieved to obtain a new table and the application system uses this, information regarding the table 32-1 on the data dictionary 31 is fetched as the data dictionary contents 33 and The necessary fields in 32-1 are fetched as the search target table contents 34.Then, the data Φ
The base book system 2Fi executes a search process while accessing these to obtain the illustrated search result table contents 35.

It is inherited. Then, the application system 3 uses the search result table contents 35 to refer to and update the data dictionary contents 33. When the processing is completed, these contents 33 and 35 are stored in the data t-verse storage section 1.

As described above, according to the present invention, the semantic attributes of tables and fields are held in a data dictionary so that they can be accessed by the data base system and the application system, respectively.

The data pace system then inherits the semantic attributes. For this purpose, the application
It is possible to centrally manage applied attributes provided by the system and basic attributes provided by the data pace system, and it is also possible to inherit and utilize semantic attributes in search results.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram illustrating the processing concept of an embodiment of the present invention regarding a data pace system and an application system, FIG. 2 is a configuration of an embodiment of a data dictionary section of the present invention, and FIG. 1 shows the configuration of an embodiment of the invention. In the figure, 1 is the data pace (or database storage)%, 2 is the data pace system, 3 is the application system, 9 is the table semantic attribute (or semantic attribute part), loa field semantic attribute (or semantic attribute Department)
, 11 is the table basic attribute, 12 is the table application attribute,
Reference numeral 13 represents a field basic attribute, 14 a field application attribute, 25 a data processing device, 31 a data dictionary, and 32 a table. Patent applicant Fujitsu Limited

Claims (1)

[Claims] It has a data pace in which a table composed of one or more fields is stored as a unit, and utilizes the data pace 0 system that manages the data pace and the search results obtained from the data pace. In a data processing system having an application system that stores dictionary information regarding the data e-pace, table semantic attribute information including table basic attributes and table application attributes of the table information regarding the table, and In addition to storing field semantic attribute information including field basic attributes and field applied attributes of the field information for each of the above fields, the data pace system has a function of defining the table basic attributes and the field basic attributes. In addition, the above applications/
The system has a function of defining the table application attribute and the field application attribute, and the data pace system predetermines the table semantic attribute and the field semantic attribute and creates a new table and /l with nine conditions. Semantic information in a data pace characterized by having a function of inheriting the new C/1 field, and managing both the data on the data pace and the semantic information related to the data. Processing method.