JPH04326428A - Control system for knowledge base of information inference device - Google Patents

Abstract

PURPOSE:To generate and recompose the secondary data with high efficiency in an information retrieving system using a knowledge base by providing a conversion part to convert the primary data into the secondary data. CONSTITUTION:When the data are added or changed at a primary data input part 1, the change information is recorded in a change register part 3. An inference arithmetic part 8 has an access to the knowledge base of a secondary data store part 5 by reference to the inference rule stored in an inference rule store part 6 of a knowledge base part 7. A knowledge control part 10 collates the access area with the information given from the part 3. If the part 10 decides that the access part is equal to the change data part, the part 10 reads the changed primary data on the corresponding part out of a primary data store part 2 and outputs a conversion instruction to a conversion part 4. Thus the part 4 converts the changed primary data into the new secondary data to store this data into the part 5 of the part 7 and updates the secondary data on the corresponding part.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique that is effective when applied to the management of secondary data included in a knowledge base in intellectual information processing such as computational reasoning.

0002

[Background Art] With the recent increase in the speed of memory technology and processors, it is expected that intelligent information processing technologies such as expert systems will be used. is difficult, and there are few examples of full-scale practical use.

[0003] For example, factual data that forms the basis of intellectual information processing such as deductive reasoning is often graphic, text, and numerical information in various forms such as drawings, specifications, and operating conditions.

[0004] These primary data need to be input into a computer system as factual knowledge data, but since the amount of this data is enormous, relying on manual input would be time-consuming and labor-intensive. However, the costs involved were enormous, making it extremely difficult to create knowledge data on a practical scale.

On the other hand, with the development of electronic computer systems in recent years, in the design and planning process using CAD, etc., the configuration, connection relationships, shapes, etc. of parts and elements are input into the computer as drawings using figures, and the design Character and numerical information such as conditions and specifications were accumulated as an engineering database, and it became possible to store this in computers as primary data.

[0006]

[Problem to be solved by the invention] However, it is difficult to subject these primary data to intellectual information processing such as deductive reasoning in their original form, and it is necessary to reprocess these primary data to create secondary data. needed to be generated.

[0007] These secondary data are, for example, knowledge data about facts expressed in a predicate format for deductive reasoning, etc., and the processing for converting the primary data into this data relies exclusively on humans. Data had to be processed.

[0008] Furthermore, when change information such as specifications is given to primary data from the outside, it is necessary to repeat the conversion to secondary data each time. Therefore, in this type of system, the main work of intelligent information processing is spent on converting and reorganizing this secondary data, and there is a problem that the efficiency of deductive reasoning processing, which is originally the main task, decreases. Ta.

The present invention has been made in view of the above problems, and its purpose is to provide a technique that can extremely efficiently generate and reorganize secondary data.

0010

[Means for Solving the Problems] The present invention provides a primary data storage unit that stores data (primary data) such as the shape and specifications of figures input through a primary data input unit 1 from input means or other databases, etc. 2, a change registration unit 3 that records change information when data is added or changed in the primary data input unit 1; a conversion unit 4 that converts the primary data into secondary data; A knowledge base section 7 consists of a secondary data storage section 5 that stores data and an inference rule storage section 6 that stores inference rules, and an inference (intellectual The knowledge management unit 10 includes an inference calculation unit 8 that executes information processing) and a knowledge management unit 10 that controls the conversion process in the conversion unit 4. The access is monitored and the accessed location is compared with the information from the change registration unit 3. If the knowledge base unit 7 detects that a location corresponding to the changed data has been accessed, the information is sent to the conversion unit 4. The gist of this paper is a knowledge base management method in an information inference device that instructs conversion processing of corresponding data and storage of the data in the knowledge base unit 7.

[0011]

[Function] Practical intellectual information processing requires a knowledge base that stores a huge amount of secondary data. For example, knowledge bases for equipment maintenance and safety management are
It can be created by converting primary data such as CAD drawings and engineering databases that are created and input for design. In addition, although it is necessary to maintain consistency and consistency between these primary data and secondary data, the time and place at which primary data is entered may differ from the time and place at which secondary data is used. Converting primary data to secondary data in real time places a heavy burden on computers, and may not be possible depending on processing power.

[0012] Furthermore, when performing intelligent information processing, it may be necessary to reorganize secondary data in order to maintain consistency with primary data, but there are a large number of locations that require such reorganization. It may be a small part of the following data or may not be present at all.

[0013] The present invention focuses on the actual state of intelligent information processing using such knowledge bases, and provides a conversion means (like the above-mentioned means) for converting primary data into secondary data.
A conversion section) was provided. Furthermore, even if there is a change in the primary data in the primary data storage unit 2, instead of immediately converting it to secondary data, it may be converted as necessary at the stage of intelligent information processing, or We decided to delay the intelligent information processing by that amount of time and convert it all at once in a batch processing manner.

Therefore, according to the present invention, efficient system operation is possible without the entire system being occupied by conversion processing when primary data is added or changed.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the system configuration of the present invention.

As shown in the figure, it has a primary data input section 1 for inputting primary data such as graphics, numerical values, character information, etc., and a primary data storage section 2 for storing this.

The primary data input section 1 includes a change registration section 3 for registering information on generation, addition, modification, etc. of primary data in the knowledge management section 10.

[0018] Here, the primary data is, for example, data related to drawings created with CAD as shown in Fig. 2, and includes information regarding figures and shapes as well as symbol identification numbers, identification names, and classification codes representing meanings. , connection information indicating the connection relationship between symbols, etc.

The primary data storage section 2 has a drawing database and an engineering database as shown in FIG. 2 as primary data. The minimum unit in the drawing database includes symbol information such as a symbol identifier, classification code, and connection relationship (S1023 and S1023 in the figure).
1024 indicates that it is connected), shape information (image data), etc. are registered. In addition, the minimum unit in the engineering database includes attribute information such as identification number, model, inner diameter, length, design pressure,
Technical information such as design temperature is registered.

The primary data storage unit 2 has a correspondence table with entities that are targets of knowledge-based intellectual information processing (search processing, etc.) to be described later (FIG. 3). Here, in the above example, which targets facility design, facility diagnosis, etc., the entity is a specific device indicated by "D-201" or the like.

[0021] Search area-limiting items, key items, etc. of primary data do not necessarily match when multiple types of primary databases exist, but hierarchical relationships are established before and after the search order in accordance with the inclusive relationship of data areas. Set this in advance.

In other words, as shown in FIG. 2, the drawing database has a hierarchical structure of "drawing database" → "plant" → "drawing" → "symbol identifier", whereas the engineering database has a hierarchical structure of "engineering database". It has a hierarchical structure of → "Plant" → "Equipment" → "Equipment identifier".

The conversion unit 4 performs processing to convert the primary data described above into secondary data as factual data.

Conversion from primary data to secondary data in the conversion section 4 is performed as follows. That is, CAD
The primary data indicating the shapes, connection relationships, and meanings/names of symbols, etc., which are the elemental parts of drawings created in It is converted into secondary data regarding facts, for example, written in predicate format, which represents the correspondence between entity identifiers.

Here, the secondary data storage unit 5 in which secondary data is stored holds data in the format shown in FIG. 4, and these are objects of intellectual information processing as a knowledge base.

Although the search area limiting items and key items of the secondary data are different from those of the primary data described above, the order of search is set in accordance with the inclusive relationship of the data areas. For example, the hierarchical relationship of fact data (secondary data) in the knowledge base shown in FIG. 4 is as follows (→ direction is lower).

"Knowledge base"..."Plant" → "Domain" → "Entity identifier" Here, if there is an addition or change (hereinafter simply referred to as update) to the above-mentioned primary data, the conversion should be 4, the updated primary data is immediately converted again and the secondary data is also updated to maintain consistency between the primary data and the secondary data. In this embodiment, data conversion is performed every time the primary data is updated. is not carried out.

That is, in this embodiment, when primary data is updated, only this updated information is first registered in the change registration section 3. Update information for such primary data is registered in units of data groups of appropriate size, and if the data has a hierarchical structure, update information is added to data items in the hierarchy above the corresponding data. For example, if the primary data has a physical hierarchical structure as shown below (→ direction indicates lower level), "Database" → "File" → "Record" → "Field" Update information is registered in the upper hierarchy including "Record".

For example, when data is updated in units of "records", the update status (flag) of the upper file or database is set, for example, with a bit set as "
"Update" is displayed.

FIG. 5 shows the relationship between these primary data, update information, and secondary data. Also, Figure 6
1 shows an example of items in the data change registration table prepared in the knowledge management unit 10.

Next, the conversion processing procedure in this embodiment will be explained using FIG. First, when the inference calculation unit 8 is given a question search formula, the inference rule storage unit 6 in the knowledge base
The knowledge base in the secondary data storage unit 5 is accessed while referring to the inference rules in the secondary data storage unit 5.

Here, checking whether factual data (secondary data) required for inference (intellectual information processing) corresponds to changed data is performed by checking the update status of updated information stored in the knowledge management unit 10. The data items to which the relevant data belongs are checked in order from the highest to the lowest, and if it is found that the upper data item has not been updated, checking of the update statuses below it is stopped. In this way, if the secondary data remains unchanged, access to the knowledge base by the inference calculation unit 8 continues as is.

In the above procedure, if it is found that the access point corresponds to the changed data part, the knowledge management unit 10 reads out the changed primary data of the relevant part from the primary data storage unit 2 and converts it to the conversion unit 4. Based on the conversion command, the conversion unit 4 converts it into new secondary data, stores it in the secondary data storage unit 5 of the knowledge base unit 7, and updates the secondary data of the corresponding part.

In the above explanation, the conversion process of the changed primary data is sequentially performed by the knowledge management unit 10 while checking the updated status. When a notification is received, the intelligent information processing may be delayed and the conversion and storage of the data range including the changed data may be batch processed. Alternatively, batch processing may be performed in accordance with instructions from an operator. This adds a large amount of primary data and
This is particularly effective when updates are made or when the designer needs to perform step-by-step checks while checking the design contents during the design process. moreover,
A timer may be provided to perform batch conversion at fixed intervals. If a timer is provided, statistics on the frequency of access to the knowledge base section 7 can be recorded, and the batch processing time period can be set to a time period with low access frequency, thereby increasing the efficiency of the entire system. Operation becomes possible.

Furthermore, when converting primary data from another database into factual data, additional data may be added from outside.

In the above embodiment, the system configuration is such that one secondary data storage section 5 is connected to one primary data storage section 2, but the system configuration is not limited to this. It may be a so-called distributed database system in which a secondary data storage section 5 is provided. In such a distributed database system, when data in the first secondary data storage unit 5 is changed, the secondary data in other unrelated secondary data storage units 5 is also subjected to conversion processing. Since the processing is not interrupted due to this, the system can be operated efficiently.

[0037]

According to the present invention, in an information reasoning system that performs intellectual information processing such as deductive reasoning, it is possible to extremely efficiently generate and manage factual data constituting a knowledge base.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the system configuration of an information inference device according to the present invention.

[Figure 2] Explanatory diagram showing the database structure in the example

[Fig. 3] Explanatory diagram showing each correspondence table in the embodiment

[
Figure 4: Explanatory diagram showing the knowledge base in the example

[Figure 5]
Explanatory diagram showing the relationship between primary data, data update registration table, and secondary data in the example

[Fig. 6] Explanatory diagram showing example items of the data change registration table in the knowledge management section of the embodiment

[Fig. 7] Flow diagram showing the conversion processing procedure of the embodiment

[Explanation of symbols]

1. Primary data input section 2. Primary data storage section 3. Change registration department 4. Conversion section 5. Secondary data storage section 6. Inference rule storage section 7. Knowledge base department 8.Inference calculation section 10...Knowledge Management Department

Claims (4)

[Claims] 1. A primary data storage section (2) for storing data (primary data) such as the shape and specifications of figures input through a primary data input section (1) from an input means or other database, etc.; a change registration unit (3) that records change information when data is added or changed in the primary data input unit (1); a conversion unit (4) that converts the primary data into secondary data; A knowledge base section (consisting of a secondary data storage section (5) that stores secondary data and an inference rule storage section (6) that stores inference rules)
7), an inference calculation unit (8) that executes inference (intellectual information processing) on the knowledge base unit (7) based on commands such as question search from the outside, and a conversion process in the conversion unit (4). The knowledge management unit (10) monitors access from the inference calculation unit (8) to the knowledge base unit (7), and registers the access point for the change. When the knowledge base unit (7) detects that a location corresponding to the changed data has been accessed, the corresponding data is sent to the conversion unit (4). A knowledge base management method in an information inference device, characterized in that it instructs conversion processing and storage of the same in a knowledge base unit (7). 2. The conversion and storage of the primary data of the location corresponding to the changed data in the knowledge base section (7) is performed when the knowledge management section (10) receives notification of the location accessed by the inference calculation section (8). 2. A knowledge base management method in an information inference device according to claim 1, wherein the conversion and storage of the data range including the changed data is batch-processed by delaying intelligent information processing. 3. The conversion and storage of the primary data in the knowledge base unit (7) corresponding to the changed data is performed in a batch process using an external command or a timer. A knowledge base management method in the described information reasoning device. 4. The knowledge management unit (10) has a data change registration table corresponding to the hierarchical structure of the primary data based on the change record from the change registration unit (3), and the data change registration table corresponds to the hierarchical structure of the primary data. In the registration table, if there is a change in lower-level data, the update status is also registered in the higher-level data to which this lower-level data belongs, and the knowledge base section (
Verification with the access in 7) is performed in order from the upper hierarchy to the lower hierarchy of this data change registration table, and if it is found that the data has not been changed during the verification, it is verified with the update status of a lower level. 4. The knowledge base management method in an information inference device according to claim 1, 2 or 3, wherein the knowledge base management method is not performed.