JPH06124219A - Data evaluation system - Google Patents

Abstract

PURPOSE:To execute an evaluation processing at a high speed relating to a data evaluation system for generating an evaluation expansion expression by CNF expanding or DNF expanding an evaluation expression and performs a processing so as to evaluate whether the record of an evaluation object satisfies the conditions of the evaluation expansion expression or not. CONSTITUTION:Conditional expressions provided in the evaluation expansion expression are managed corresponding to a form for making them common for same ones and further, evaluated result storage areas 12 for storing sufficiency evaluated results for the conditional, expressions are provided. An evaluation mechanism 14 is provided with a storage means 141 for storing the sufficiency evaluated results for the conditional expressions provided in the evaluation expansion expression in the evaluated result storage areas 12 and a referring means 142 for referring to the corresponding evaluated result storage area 12 before starting the sufficiency evaluation of the conditional expressions and when the referring means 142 judges the storage of the sufficiency evaluated results, the evaluation processing is executed while utilizing the sufficiency evaluated results.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generates an evaluation expansion expression by CNF expansion or DNF expansion of an evaluation expression and evaluates whether or not a record to be evaluated satisfies the condition of the evaluation expansion expression. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data evaluation system that processes in a progressive manner, and particularly to a data evaluation system that enables high-speed evaluation processing.

In a relational database, a search process is executed to extract records satisfying a given search expression from the management records. In order to make effective use of the relational database, it is necessary to be able to execute this search processing at high speed.

[0003]

2. Description of the Related Art A relational database is shown in FIG.
As shown in, when managing the table data T1,
From the user, the value of the C1 item has "20" and C2
Item value has "30" or C3 item value is "40"
And the value of the C5 item of the record having the value of the C4 item is “50” SELECT C5 FROM T1 WHERE (C1 = 20 AND C2 = 30) OR (C3 = 40 AND C4 = 50) When an expression is given, records are read one by one from the table data T1, and it is evaluated whether or not the read record satisfies this evaluation expression, and C5 of the record that satisfies this evaluation expression is held. A search process of storing the item values in the search result table is executed.

In this search processing, W is described by AND connection / OR connection so that it can be processed by a computer.
CNF expansion of the evaluation expression in the HERE clause (O in parentheses
It is described in R condition, and those parentheses are connected by AND condition. Conditional expansion format) or DNF expansion (A in parentheses is A.
It is described by the ND condition, and those parentheses are converted into the expansion form of the condition expansion form connected by the OR condition).

Conventionally, in the CNF expansion / DNF expansion, the structure of managing the contents of the conditional expression corresponding to each conditional expression described in the evaluation expansion expression has been adopted.

[0006]

The CNF expansion / DNF expansion is an expansion method for converting a format according to a certain rule while maintaining the conditions of the entire evaluation expression, and the evaluation generated by this expansion method. The expansion expression may have a plurality of the same conditional expressions.

However, according to such a conventional technique, in such a case, the same conditional expression content is managed separately for each conditional expression although the conditional expression content is the same. Will be taken.

According to the conventional technique, when the evaluation expansion expression has a plurality of the same conditional expressions, the same conditional expression is repeatedly evaluated for the read record. There was a point. That is, according to the conventional technique, there is a problem that it is not possible to quickly evaluate whether or not the search expression is satisfied for the record to be evaluated.

The present invention has been made in view of such circumstances, and an evaluation expansion expression is generated by CNF expansion or DNF expansion of the evaluation expression, and the record to be evaluated satisfies the condition of the evaluation expansion expression. It is an object of the present invention to provide a new data evaluation system capable of executing this evaluation processing at high speed in a data evaluation system for performing processing to evaluate whether or not to perform.

[0010]

FIG. 1 shows the principle configuration of the present invention. In the figure, 1 is a data processing apparatus equipped with the present invention, and 2 is a database such as a relational database that manages records to be evaluated by the data processing apparatus 1.

The data processing apparatus 1 comprises a development mechanism 10, an evaluation development formula management area 11, an evaluation result storage area 12, a reading mechanism 13, an evaluation mechanism 14, and a search result table 15.

The expansion mechanism 10 generates an evaluation expansion expression by CNF expansion or DNF expansion of the evaluation expression for the record. The evaluation expansion formula management area 11 manages the evaluation expansion formula generated by the expansion mechanism 10. Evaluation deployment management area 11
In order to execute this management process, the conditional expression management area 110 is configured to manage the conditional expressions of the generated evaluation expansion expressions according to a common mode for the same one.
And the execution relation of the conditional expression of the evaluation expansion expression is managed according to the satisfaction evaluation result for the conditional expression, and the final evaluation result for the evaluation expansion expression derived from the execution relationship is managed. And a conditional expression evaluation block 111 having the above configuration.

The evaluation result storage area 12 is a conditional expression management area 11
The condition evaluation management area 110 is provided corresponding to 0 and stores the satisfaction evaluation result for the conditional expression managed by the conditional expression management area 110. The reading mechanism 13 reads the record to be evaluated from the database 2.

The evaluation mechanism 14 evaluates whether the record to be evaluated satisfies the condition defined by the evaluation expansion formula. In order to execute this evaluation processing, the evaluation mechanism 14 evaluates the evaluation means 140 configured to evaluate whether or not the conditional expression managed by the conditional expression management area 110 is satisfied, and the evaluation result of the evaluation means 140. A storage unit 141 having a configuration of storing in the corresponding evaluation result storage area 12, and an evaluation result storage area 12
And a reference unit 142 configured to refer to the evaluation result of 1.

The search result table 15 stores the evaluation result of the evaluation mechanism 14.

[0016]

In the present invention, when the expansion mechanism 10 generates an evaluation expansion expression by CNF expansion or DNF expansion of the evaluation expression, the evaluation expansion expression management area 11 manages the conditional expression corresponding to the generated evaluation expansion expression. Area 110 / conditional expression evaluation block 11
By generating 1, the configuration for managing the generated evaluation expansion formula is adopted. At this time, the evaluation expansion management area 11
When the generated evaluation expansion expressions have a plurality of the same conditional expressions, the conditional expression management area 110 is generated so that the conditional expressions are managed in common.

When the evaluation expansion formula is managed in the evaluation expansion formula management area 11 in this way, the evaluation means 1 of the evaluation mechanism 14
The reference numeral 40 traces the management data of the conditional expression evaluation block 111 using the evaluation target record read by the reading mechanism 13 to determine whether or not the evaluation target record satisfies the evaluation expansion expression. . At this time, the storage means 1
41 stores the satisfaction evaluation result for the conditional expression of the evaluation expansion expression in the corresponding evaluation result storage area 12, and the reference means 142 evaluates the satisfaction of the conditional expression of the evaluation expansion expression of the evaluation means 140. Before entering the processing, the corresponding evaluation result storage area 12 is referred to, if the satisfaction evaluation result is already stored, the evaluation means 140 is notified of the satisfaction evaluation result, and if not stored. , That means evaluation means 1
40 will be notified.

When the evaluation means 140 receives the notification from the reference means 142 and omits the satisfaction evaluation processing of the conditional expression, the satisfaction evaluation processing for the conditional expression is omitted. The conditional expression evaluation block 11 is used while using the satisfaction evaluation result notified from the reference unit 142.
By tracing the management data of No. 1, it is judged whether or not the evaluation expansion formula is satisfied.

As described above, in the present invention, the evaluation formula is CNF.
Generate an evaluation expansion expression by expanding or DNF expansion,
When evaluating whether or not the record to be evaluated satisfies the condition of the evaluation expansion expression, and when the evaluation expansion expression has multiple identical conditional expressions, the satisfaction evaluation process for that conditional expression Is configured to be executed only once, and the result of the satisfaction evaluation process is used thereafter.
It becomes possible to evaluate whether or not the record to be evaluated satisfies the condition of the evaluation expansion formula faster than before.

[0020]

EXAMPLES The present invention will be described in detail below with reference to examples applied to a relational database.

The expansion mechanism 10 described with reference to FIG. 1 generates an evaluation expansion expression by CNF expansion or DNF expansion of the evaluation expression represented as a search expression. FIG. 2 illustrates a processing flow of the CNF expansion processing executed by the expansion mechanism 10.

As shown in the processing flow of FIG. 2, the expansion mechanism 10 first expresses the evaluation expression in a description format in step 1. For example, when the evaluation expression is ((conditional expression and conditional expression) or (conditional expression and (conditional expression or conditional expression or conditional expression))) and conditional expression, AND combination block / OR as shown in FIG. It is expressed in the form of a description of the connection of connected blocks.

Next, in step 2, the AND connection block / OR connection block appearing in the description format is expanded into a simple format AND connection / OR connection. FIG. 4 illustrates an AND block expansion process, and FIG. 5 illustrates an OR block expansion process. The expansion processing to the AND connection / OR connection is executed by recursively calling the expansion processing of FIGS. 4 and 5. Here, in FIG. 4A, “(conditional expression a
nd conditional expression) ”as an example, FIG. 4B shows“ (conditional expression)
or conditional expression) and conditional expression "as an example, and FIG.
Is an example of "(conditional expression or conditional expression)"
(B) is "(conditional expression and conditional expression) or (conditional expression
and conditional expression) is taken as an example.

As shown in FIG. 4A, the expansion of the AND connection block is executed by connecting the conditional expressions of the AND connection with vertical pointers.
As shown in (b), when the OR join block is pointed instead of the conditional expression block, the OR join expansion process is called, the OR join is expanded, and then the AND join conditional expression is set to the vertical pointer. Run by connecting with.

Then, as shown in FIG.
The expansion of the R block is executed by connecting the conditional expressions of the OR connection with a horizontal pointer, and FIG.
As shown in (b), when the AND connection is pointed, the Cartesian product of the conditional expressions of the AND connection is obtained and connected by a horizontal pointer. However, if the AND connection block is pointed instead of the conditional expression block, the AND connection expansion process is called to
It is executed by expanding the join, obtaining the direct product of the conditional expressions, and connecting with the horizontal pointer.

When the expansion process of step 2 is executed,
The description format shown in FIG. 3 is expanded to that shown in FIG. Then, in step 3, the AND connection of each level expanded according to the process of step 2 is cut to convert it into the CNF expansion format. This disconnection process
As shown in FIG. 7, which takes “(conditional expression or (conditional expression and conditional expression)” as an example, when disconnecting an M level AND connection, a conditional expression block one level higher is copied to obtain an M level logical block. It is executed by connecting the conditional expression block with an OR connection.

In this way, the expansion mechanism 10 executes the processing flow of FIG. 2 to perform CNF expansion of the evaluation expression to generate the evaluation expansion expression. FIG. 8 shows an evaluation expansion expression for an evaluation expression ((conditional expression and conditional expression) or (conditional expression and (conditional expression or conditional expression or conditional expression))) and conditional expression generated by the expansion mechanism 10. And (conditional expression or conditional expression) and (conditional expression or conditional expression or conditional expression or conditional expression) and (conditional expression or conditional expression) and (conditional expression or conditional expression or conditional expression or conditional expression) and (conditional expression) Illustrate.

As described with reference to FIG. 1, when the expansion mechanism 10 generates an evaluation expansion expression, the evaluation expansion expression management area 11 has a conditional expression management area 110 / which corresponds to the generated evaluation expansion expression.
By generating the conditional expression evaluation block 111, a configuration for managing the generated evaluation expansion expression is adopted. At this time, the evaluation expansion expression management area 11 is configured to generate the conditional expression management area 110 so as to manage the conditional expressions in common when the generated evaluation expansion expressions have a plurality of identical conditional expressions. take. Then, the evaluation result storage area 12 is generated corresponding to the conditional expression management area 110.

FIG. 9 shows an embodiment of the conditional expression management area 110, the conditional expression evaluation block 111 and the evaluation result storage area 12 generated by the evaluation expansion expression of FIG. Here, the “true” column of the conditional expression evaluation block 111 represents the entry number of the conditional expression to be executed next when the record to be evaluated shows a true value for the conditional expression, and the “false” field. Is
It represents the entry number of the conditional expression to be executed next when the record to be evaluated shows a false value for that conditional expression. When "-" is registered in the "true" column, it means that the record to be evaluated satisfies the condition of the evaluation expansion formula, and when "-" is registered in the "false" column, it means that the record is evaluated. , The record to be evaluated does not satisfy the condition of the evaluation expansion formula.

As described with reference to FIG. 1, the evaluation mechanism 14 is
Conditional expression evaluation block 111 using the record to be evaluated
By tracing the management data of, it is determined whether or not the record to be evaluated satisfies the evaluation expansion formula.

FIG. 10 illustrates an example of a processing flow executed by the evaluation mechanism 14. Next, according to this processing flow, one embodiment of the evaluation processing executed by the present invention will be described in detail.

When the evaluation mechanism 14 receives the record to be evaluated from the reading mechanism 13, as shown in the processing flow of FIG. 10, first, in step 1, the evaluation result storage area 1
Initialize 2. Next, at step 2, "1" is set to the variable N which displays the entry number of the conditional expression evaluation block 111. Subsequently, in step 3, it is determined whether or not a satisfactory evaluation result is stored in the evaluation result storage area 12 associated with the conditional expression of the Nth entry.

When it is determined in step 3 that the conditional expression has not been evaluated, the process proceeds to step 4,
It is evaluated whether or not the record to be evaluated satisfies the conditional expression of the Nth entry, and in the subsequent step 5, the evaluation result is stored in the corresponding evaluation result storage area 12. On the other hand, when it is determined in step 3 that the conditional expression has been evaluated, the process proceeds to step 6 and the evaluation result is acquired from the evaluation result storage area 12 associated with the conditional expression of the Nth entry.

When the processing of step 5 / step 6 is completed, subsequently, in step 7, the obtained evaluation result is "true".
Is displayed or "false" is displayed, and when it is determined that "true" is displayed, the process proceeds to step 8 and the "true" column of the conditional expression evaluation block 111 is displayed. Set the registered entry number to the variable N,
When it is determined that "false" is displayed, the process proceeds to step 9, and the entry number registered in the "false" column of the conditional expression evaluation block 111 is set in the variable N.

In the processing of step 8 / step 9,
In some cases, the entry number is not registered in the "true" / "false" column but "-" is registered, so in step 10 that follows, it is determined whether or not this "-" is registered. If "-" is not registered, that is, if the entry number is registered, the process returns to step 3.

On the other hand, when it is judged that "-" is registered, the process proceeds to step 11 and the "-" is "true".
If it is registered in the "false" column and if it is registered in the "true" column, the process proceeds to step 12, and the record to be evaluated is developed. When it is determined that the condition of the expression is satisfied and it is determined that the condition is registered in the “false” column, the process proceeds to step 13 and it is determined that the record of the evaluation target does not satisfy the condition of the evaluation expansion formula To finish.

In this way, when the evaluation expansion expression has a plurality of conditional expressions that are the same, the evaluation mechanism 14 executes the satisfaction evaluation processing for the conditional expression only once, and then performs the satisfaction evaluation. By performing the processing so as to use the processing result, it is determined whether or not the evaluation target record satisfies the condition of the evaluation expansion expression. On the other hand, in the conventional technique, as shown in FIG. 11, even when a plurality of the same conditional expressions are described in the evaluation expansion expression, the common management of these expressions is not adopted. The configuration has been adopted in which the satisfaction evaluation process is repeated for the same conditional expression. From this, according to the present invention, it becomes possible to evaluate whether or not the condition of the evaluation expansion formula is satisfied, faster than in the conventional technique.

Although the illustrated embodiment has been described, the present invention is not limited to this. For example, in the embodiment, C
Although the present invention has been disclosed according to the NF-expanded evaluation expansion formula, the present invention is also applicable to the DNF-expanded evaluation expansion formula. Further, although the present invention has been disclosed according to the embodiment applied to the search processing of the relational database in the embodiment, the present invention is not limited to the search processing of the database, and can be applied as it is to an evaluation system for all data.

[0039]

As described above, according to the present invention,
A data evaluation system that processes an evaluation expansion expression by CNF expansion or DNF expansion of the evaluation expression and evaluates whether or not the record to be evaluated satisfies the condition of the evaluation expansion expression. , This evaluation process can be executed at high speed.

[Brief description of drawings]

FIG. 1 is a principle configuration diagram of the present invention.

FIG. 2 is a processing flow executed by a deployment mechanism.

FIG. 3 is an explanatory diagram of CNF expansion processing.

FIG. 4 is an explanatory diagram of CNF expansion processing.

FIG. 5 is an explanatory diagram of CNF expansion processing.

FIG. 6 is an explanatory diagram of CNF expansion processing.

FIG. 7 is an explanatory diagram of CNF expansion processing.

FIG. 8 is an explanatory diagram of CNF expansion processing.

FIG. 9 is an example of an evaluation expansion management area.

FIG. 10 is an example of a processing flow executed by the evaluation mechanism.

FIG. 11 is an explanatory diagram of a conventional technique.

FIG. 12 is an explanatory diagram of a search process.

[Explanation of symbols]

 1 Data Processing Device 2 Database 10 Expansion Mechanism 11 Evaluation Expansion Expression Management Area 12 Evaluation Result Storage Area 13 Reading Mechanism 14 Evaluation Mechanism 15 Search Result Table 110 Conditional Expression Management Area 111 Conditional Expression Evaluation Block 140 Evaluation Means 141 Storage Means 142 Reference Means

Claims (1)

[Claims] 1. An expansion mechanism (10) for generating an evaluation expansion expression by CNF expansion or DNF expansion of the evaluation expression, and an evaluation for evaluating whether or not a record to be evaluated satisfies a condition of the evaluation expansion expression. In the data evaluation system including the mechanism (14), the conditional expression described in the evaluation expansion expression is managed while following the common mode for the same conditional expression, and further, in association with the conditional expression, The evaluation result storage area (12) for storing the satisfied evaluation result for the conditional expression is adopted, and the evaluation mechanism (14) is configured to satisfy the conditional expression of the evaluation expansion expression. Storage area for the above evaluation results corresponding to the results
A storage means (141) for storing in (12) and a reference means (142) for referring to the corresponding evaluation result storage area (12) before starting the satisfaction evaluation processing of the conditional expression are provided. A data evaluation system characterized in that, when (142) decides to store the satisfaction evaluation result, the evaluation process is executed while utilizing the satisfaction evaluation result.