KR20180115202A - Hybrid knowledge acquisition method for ripple down rules: combining case based and rule based approaches - Google Patents

Abstract

Disclosed are a method for acquiring multiple types of knowledge using ripple down rules and a system thereof. According to one embodiment of the present invention, the method for acquiring knowledge by the system comprises the steps of: categorizing acquired knowledge according to knowledge learning into a case standard or a knowledge standard; and accumulating the knowledge categorized based on the case standard or the knowledge standard. In the step of accumulating the categorized knowledge based on the case standard or the knowledge standard, the system may perform maintenance on a knowledge base storing knowledge through the accumulated knowledge based on the categorized knowledge based on the case standard and construct cases on the accumulated knowledge based on the knowledge categorized based on the knowledge standard.

Description

HYBRID KNOWLEDGE ACQUISITION METHOD FOR RIPPLE DOWN RULES: COMBINING CASE BASED AND RULE BASED APPROACHES BACKGROUND OF THE INVENTION 1. Field of the Invention < RTI ID = 0.0 >

The following explains the classification of the knowledge acquisition process of the Ripple Down Rules (RDR) theory for the establishment and maintenance of the knowledge base of Expert Systems.

In order to maintain the knowledge base continuously, the RDR theory suggests to manage cases used in knowledge acquisition (cornstone cases) separately. In this respect, the RDR theory is basically case-driven based on Case Based Reasoning (CBR). However, there is a time and cost limit to find the cases necessary for acquiring knowledge in the course of acquiring knowledge using field system experts.

1 is a diagram for explaining case based reasoning (CBR). Case-based Reasoning is a field of artificial intelligence that focuses on human perception of things in the light of past experience. It is a system that applies solutions to past problems .

2 is a diagram for explaining Ripple Down Rules (RDR). The Ripple Down Rules (RDR) theory approaches the knowledge acquisition process of a system from a software engineering concept called maintenance of knowledge. From this point of view, the knowledge base of the system is constantly understood as a maintenance target and a methodology for making such maintenance possible is presented. In the RDR theory, the role of the case used in the knowledge acquisition process (cornstone case) is essential to maintain and maintain the system knowledge base. The system manages Cornerstone cases separately and performs Validation & Verification (V & V) through stored cornstone cases for problems that may arise in the creation of new knowledge (rules) in the knowledge acquisition process.

Referring to FIG. 3, there is a problem in the existing RDR knowledge acquisition process. The Ripple Down Rules (RDR) theory does not distinguish between the initial knowledge building process and the maintenance process because the knowledge base of the system is perceived as a constant maintenance object. However, since there is a limitation on system development time in actual field, it aims to learn the maximum amount of knowledge in the initial knowledge building process. In this regard, the existing RDR knowledge acquisition process should be verified for all cases that exist at the beginning of the field experts. Textbook knowledge already exists in field experts, and it takes a lot of time to acquire knowledge based on the above knowledge. Also, if there are no cases in the existing case base, knowledge related to it can not be learned at that time.

References: KR10-1568346, KR10-1636477

The present invention relates to a method and system for classifying a process of learning knowledge into a case-driven and a knowledge-driven according to a type of knowledge acquisition and accumulating knowledge classified on the basis of a case basis or a knowledge base Can be provided.

The present invention performs maintenance on a knowledge base storing knowledge through accumulated knowledge based on classified knowledge based on a case basis and analyzes the accumulated knowledge based on knowledge classified based on the knowledge base A method and system for enabling the knowledge accumulation method of RDR theory to be used can be provided.

A method of acquiring knowledge performed by a knowledge acquisition system includes: classifying knowledge acquired according to knowledge learning into a case reference or a knowledge reference; And accumulating knowledge classified on the basis of the case basis or the knowledge criterion, wherein, in the step of accumulating knowledge based on the case criterion or the knowledge criterion, Based on the accumulated knowledge, the maintenance of the knowledge base storing the knowledge can be performed, and the accumulated knowledge can be constructed based on the knowledge classified based on the knowledge base.

The knowledge acquisition method may further include generating a virtual case through a virtual case generator when there is no case required for accumulating the knowledge in the case base in which the case is stored.

Wherein the step of accumulating knowledge based on the case basis or the knowledge criterion comprises defining rules according to receiving at least one condition and a conclusion being selected and selecting the case based on the defined rule from the stored case database And searching for the search result.

Wherein the step of accumulating the classified knowledge based on the case basis or the knowledge criterion comprises the steps of generating a virtual case when no case based on the defined rule exists from the case database, And performing validation and verification of the case when there is a case based on the defined rule from the case database.

The step of accumulating knowledge classified on the basis of the case basis or the knowledge criterion comprises the steps of storing the cornerstone case generated according to performing the validation and verification and the rules related to the cornerstone case in the knowledgebase and cornerstone database Thereby accumulating the knowledge.

The step of accumulating knowledge classified on the basis of the case basis or the knowledge criterion comprises the steps of: searching for an Incorrect case from the case database in which the case is stored and defining a rule for the inaccurate case; May be generated by selecting at least one condition and a conclusion, respectively.

Wherein the step of accumulating the classified knowledge based on the case basis or the knowledge criterion comprises the steps of: validating and verifying the inaccurate case based on the defined rule; And accumulating the knowledge by storing rules associated with the knowledge base and the cornerstone database.

The knowledge acquisition system includes: a classification unit that classifies knowledge acquired according to learning of knowledge into a case basis or a knowledge base; And an acquisition unit for accumulating knowledge classified based on the case reference or the knowledge criterion, wherein in the acquisition unit, knowledge that stores knowledge through accumulated knowledge based on the knowledge classified based on the case criterion Perform maintenance on the base, and build an example of accumulated knowledge based on the knowledge classified based on the knowledge criteria.

The knowledge acquisition system may further include a virtual case generator for generating a virtual case through a virtual case generator when there is no case required to accumulate the knowledge in the case base in which the case is stored.

The obtaining unit may define a rule as it receives that at least one condition and a conclusion are selected, and may search for a case based on the defined rule from the case database in which the case is stored.

Wherein the obtaining unit is configured to generate a virtual case when there is no case based on the defined rule from the case database, to perform validation and verification on the generated virtual case, If there is a case based on the case, validation and verification of the case can be performed.

The obtaining unit may accumulate the knowledge by storing the cornerstone case generated according to the validation and verification and the rules related to the cornerstone case in the knowledge base and the cornerstone database.

The obtaining unit includes searching for an Incorrect case from the case database in which the case is stored and defining a rule for the incorrect case, and the rule is generated by selecting at least one condition and a conclusion, respectively .

Wherein the obtaining unit stores the cornstone case created in accordance with the validation and verification of the incorrect case based on the defined rule and the rules related to the cornerstone case in the knowledge base and the cornerstone database, . ≪ / RTI >

According to the knowledge-driven method proposed in the present invention, it is possible to reduce the cost of finding cases necessary for acquiring knowledge, and significantly reduce the time cost of the field expert in the stage of building the knowledge base.

In addition, the present invention can generate a case virtually through a virtual case generator when there is no case required by the rule defined in the case database stored in the case database, Validation & Verification (V & V) process of acquiring knowledge.

In addition, the present invention classifies the knowledge acquisition process of the RDR theory into a case standard and a knowledge standard, thereby making it possible to acquire knowledge in the form required according to the request of the field expert and the type of knowledge acquisition. Accordingly, knowledge acquired through case standards can be used for maintenance / maintenance of knowledge base, and knowledge acquired through knowledge base can be used for knowledge building.

1 is a diagram for explaining case based reasoning (CBR).
2 is a diagram for explaining Ripple Down Rules (RDR).
3 is a diagram for explaining a problem of the conventional RDR knowledge acquisition process.
4 is a block diagram illustrating a configuration of a knowledge acquisition system according to an embodiment.
5 is a flowchart illustrating a method of acquiring knowledge in a knowledge acquisition system according to an embodiment.
FIG. 6 is a diagram for explaining a method of acquiring case-based knowledge in a knowledge acquisition system according to an embodiment.
7 is a diagram for explaining a method of acquiring knowledge base knowledge in a knowledge acquisition system according to an embodiment.
8 is a diagram for explaining a virtual case generator according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings.

The following embodiment relates to a knowledge acquisition system and method for classifying a knowledge acquisition process of Ripple Down Rules (RDR) theory for managing knowledge base construction and maintenance. For the continuous maintenance of the knowledge base that stores knowledge, the cases of cornerstone used in the knowledge acquisition process should be managed separately. From this point of view, RDR theory basically acquires knowledge on case basis based on case based reasoning (CBR). However, there are time and cost limitations in the field expert 's acquisition of knowledge using the system.

Accordingly, a method and system for classifying a process of learning knowledge into a case basis or a knowledge base and accumulating knowledge classified on the basis of a case basis or a knowledge base will be described. In addition, we will explain the process of knowledge base method to use knowledge accumulation method of existing RDR theory.

FIG. 4 is a block diagram for explaining a configuration of a knowledge acquisition system according to an embodiment, and FIG. 5 is a flowchart for explaining a knowledge acquisition method in a knowledge acquisition system according to an embodiment.

The processor 400 of the knowledge acquisition system 100 may include a classifier 410 and an acquiring unit 420. In addition, the processor 400 of the knowledge acquisition system 100 may further include a virtual case generator (not shown). The virtual case generator can create a virtual case through a virtual case generator if there is no case required to accumulate knowledge in the case base where the case is stored.

Further, the knowledge acquisition system 100 may include a storage unit (not shown) including a case database storing a case, a cornerstone case database storing a cornerstone case, and a knowledge base storing knowledge.

The components of the processor 400 and the processor 400 may control the knowledge acquisition system to perform the steps 510 through 520 that are included in the knowledge acquisition method of FIG. At this time, the components of the processor 400 and the processor 400 may be implemented to execute an instruction according to code of an operating system and code of at least one program included in the memory. Here, the components of the processor 400 may be representations of different functions performed by the processor 400 in accordance with control commands provided by the program code stored in the knowledge acquisition system 100.

The processor 400 may load the program code stored in the file of the program for the knowledge acquisition method into the memory. For example, when a program is executed in the knowledge acquisition system 100, the processor can control the knowledge acquisition system to load the program code from the file of the program into the memory under the control of the operating system.

In step 510, the classifier 410 may classify the acquired knowledge into a case basis or a knowledge basis by learning the knowledge. For example, the classifier 410 may classify the knowledge acquisition process of the RDR theory into two types including a case basis and a knowledge base. At this time, it is possible to acquire knowledge in the necessary form according to the request of field experts and the type of knowledge acquisition.

In step 520, the acquiring unit 420 may accumulate classified knowledge based on a case basis or a knowledge base. The acquiring unit 420 performs maintenance on the knowledge base storing the knowledge based on the classified knowledge based on the case basis, and acquires the accumulated information based on the knowledge classified based on the knowledge base Knowledge can be built as an example.

The acquiring unit 420 can define the rule by selecting at least one condition and a conclusion, and can search the case stored in the case database based on the defined rule. At this time, the case database may refer to the store storing the case. If there is no case based on the rule defined in the case database, the obtaining unit 420 may generate a virtual case and perform validation and verification on the generated virtual case. The acquiring unit 420 may perform validation and verification of the case when there is a case based on the rule defined in the case database. The acquiring unit 420 may accumulate knowledge by storing rules and cornerstone cases for cornerstone cases generated according to the validation and verification in Cornerstone case database and knowledge base.

The acquiring unit 420 can search for an Incorrect case from the case database where the case is stored, and define rules for an incorrect case. At this time, the rule may be generated by selecting at least one condition and a conclusion respectively. The acquiring unit 420 stores the generated Cornerstone case and the rules associated with the Cornerstone case in the Cornerstone case database and the knowledge base by performing validation and verification on the incorrect case based on the defined rule, Can accumulate.

FIG. 6 is a diagram for explaining a method of acquiring case-based knowledge in a knowledge acquisition system according to an embodiment.

The knowledge acquisition system may search for an incorrect case from the case database 601 where the case is stored (610). For example, the knowledge acquisition system may select and inaccurate cases that are present in the case database 601 may be searched by field experts, thereby seeking and obtaining inaccurate cases.

The knowledge acquisition system may define rules for inaccurate cases (620). At this time, the rule may be generated by selecting at least one condition and a conclusion respectively. For example, the knowledge acquisition system may define rules by combining each of at least one condition and a conclusion selected by a field expert.

The knowledge acquisition system may perform validation and verification of an inaccurate case based on the defined rules (630). At this time, based on the cornerstone cases stored in the cornerstone case database 602, validation and verification of an incorrect case is performed. For example, the knowledge acquisition system may perform validation and validation by comparing Cornerstone cases stored in the Cornerstone Case database 602 with incorrect instances including defined rules.

The knowledge acquisition system may store the generated cornstone case and the rules for it in the cornstone case database 602 and the knowledge base 603 as the validation and verification of the incorrect case is performed. As a result, knowledge including Cornerstone cases and rules can be obtained.

The knowledge acquisition system may store in the cornerstone case database 602 and the knowledge base 630 rules related to the generated cornerstone case and the generated cornerstone case as the validation and verification of the incorrect case is performed . By doing so, knowledge can be stored (640). Thus, knowledge may be obtained that includes rules relating to cornerstone cases and cornerstone cases that are stored.

7 is a diagram for explaining a method of acquiring knowledge base knowledge in a knowledge acquisition system according to an embodiment.

The knowledge acquisition system may define rules 710 based on at least one or more conditions and conclusions. For example, the knowledge acquisition system may define the rules as received by the field specialist that at least one condition or conclusion is selected.

The knowledge acquisition system can search whether the case stored in the case database exists based on the defined rule. In other words, the knowledge acquisition system can extract cases that fit the defined rules from the case database (720). Furthermore, the knowledge acquisition system can determine cases that are misclassified based on defined rules, and may add additional exception rules other than the defined rules.

The knowledge acquisition system may perform validation and verification of the extracted case (740) if the case stored in the case database exists (730) based on the defined rule. At this time, the validity and verification of the extracted cases are performed based on the cornerstone cases stored in the cornerstone case database 602. For example, the knowledge acquisition system may perform validation and verification by comparing the coronstone cases stored in the coronstone case database 702 with the extracted cases.

The knowledge acquisition system may store the generated cornstone case and rules for the cornstone case database 702 and the knowledge base 703 as the validation and verification of the extracted case is performed (750). As a result, knowledge including Cornerstone cases and rules can be obtained.

On the other hand, the knowledge acquisition system may generate a virtual case (760) if the case stored in the case database does not exist (730) based on the defined rule. At this time, the virtual case can be generated by the virtual case generator, and a case suitable for the defined rule can be created as a virtual case.

Referring to FIG. 8, since the role of the case in the knowledge acquisition process of the RDR theory is essential for verification (validation and verification) of the knowledge base for the virtual case generator, , And a virtual case generator 800 (virtual case generator).

In order to create a virtual case, it is possible to generate the structure of the cornstone case stored in the cornstone case database 810 and the rule defined by the expert in the knowledge acquisition process. More specifically, a case structure can be taken from the cornerstone case stored in the cornerstone case database 810. [ A rule may be defined 820 as at least one condition and conclusion selected by the expert is selected. A virtual case can be created based on the structure and rules of the cornerstone case obtained from the cornerstone case database 810. [ At this time, the numerical value of case structure can be processed as Missing value type.

The knowledge acquisition system may perform validation and verification of the generated virtual case (740). At this time, validation and verification are performed on the extracted cases based on the cornerstone cases stored in the cornerstone case database 702. FIG. For example, the knowledge acquisition system determines the fitness for a virtual case. After the fitness of the virtual case is determined based on the conditions set in the knowledge acquisition system, the final verification can be performed by the field expert.

The knowledge acquisition system may store the generated cornstone example and the rules for it in the cornstone case database 702 and the knowledge base 703 as the validation and verification of the virtual case is performed (750). As a result, knowledge including Cornerstone cases and rules can be obtained.

The knowledge acquisition system according to an embodiment classifies the knowledge acquisition process of the RDR theory into two types including a case basis or a knowledge base, do. In this case, since all of the classified criteria accumulate knowledge in the same manner, it can be used without discrimination between the existing case reference type and the knowledge reference type. In addition, the knowledge acquisition system according to an exemplary embodiment may use the form of the knowledge base to build knowledge, and use the form of the case base to manage maintenance and maintenance of the knowledge base.

Therefore, the knowledge acquisition system performs validation and verification, which is the final stage of the knowledge acquisition process, in the knowledge acquisition process, and therefore, after the fitness of the virtual case is verified, the virtual case is stored It becomes usable without any additional verification of the generated virtual case.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA) , A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device As shown in FIG. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (14)

A knowledge acquisition method performed by a knowledge acquisition system,
Classifying the acquired knowledge into a case basis or a knowledge basis by learning the knowledge; And
Accumulating knowledge that is classified based on the case basis or the knowledge criteria,
In the step of accumulating the classified knowledge based on the case reference or the knowledge reference, maintenance is performed on the knowledge base storing the knowledge based on the classified knowledge based on the case reference , Build a case for accumulated knowledge based on knowledge categorized based on the knowledge criteria
Wherein the method comprises the steps of: The method according to claim 1,
If there is no case required to accumulate the knowledge in the case base in which the case is stored, a step of creating a virtual case through a virtual case generator
Further comprising the steps of: The method according to claim 1,
Wherein accumulating knowledge based on the case basis or the knowledge criteria comprises:
Defining rules according to receiving at least one condition and a conclusion selected, and searching for a case based on the defined rule from the case database in which the case is stored
≪ / RTI > The method of claim 3,
Wherein accumulating knowledge based on the case basis or the knowledge criteria comprises:
If a case based on the defined rule does not exist from the case database, a virtual case is created, validation and verification of the generated case are performed, and a case based on the defined rule is extracted from the case database If yes, perform the validation and verification of the case
≪ / RTI > 5. The method of claim 4,
Wherein accumulating knowledge based on the case basis or the knowledge criteria comprises:
Accumulating the knowledge by storing in the knowledge base and the cornerstone database rules related to the cornerstone case and the cornerstone case generated by performing the validation and verification;
≪ / RTI > The method according to claim 1,
Wherein accumulating knowledge based on the case basis or the knowledge criteria comprises:
Searching for an Incorrect case from the case database where the case is stored and defining a rule for the incorrect case, the rule being generated by selecting at least one condition and a conclusion respectively,
≪ / RTI > The method according to claim 6,
Wherein accumulating knowledge based on the case basis or the knowledge criteria comprises:
Accumulating the knowledge by storing in the knowledge base and the cornerstone database rules related to the cornerstone case and the cornerstone case generated by performing validation and verification of the incorrect case based on the defined rule
≪ / RTI > In a knowledge acquisition system,
A classification unit for classifying the acquired knowledge according to learning of knowledge into a case basis or a knowledge base; And
And an acquiring unit for accumulating the classified knowledge based on the case reference or the knowledge reference,
Wherein the acquisition unit performs maintenance on a knowledge base storing knowledge through accumulated knowledge based on the classified knowledge based on the case criterion and accumulates information based on the knowledge classified based on the knowledge base To build a case for knowledge
Wherein the knowledge acquisition system comprises: 9. The method of claim 8,
If there is no case required to accumulate the knowledge in the case base in which the case is stored, a virtual case generator
Further comprising: 9. The method of claim 8,
Wherein the obtaining unit comprises:
Defining rules according to receiving at least one condition and a conclusion being selected, and searching the case based on the defined rule from the case database in which the case is stored
Wherein the knowledge acquisition system comprises: 11. The method of claim 10,
Wherein the obtaining unit comprises:
If a case based on the defined rule does not exist from the case database, a virtual case is created, validation and verification of the generated case are performed, and a case based on the defined rule is extracted from the case database If yes, perform the validation and verification of the case
Wherein the knowledge acquisition system comprises: 12. The method of claim 11,
Wherein the obtaining unit comprises:
Storing the cornstone case created by performing the validation and verification and rules related to the cornerstone case in the knowledge base and the cornerstone database to accumulate the knowledge
Wherein the knowledge acquisition system comprises: 9. The method of claim 8,
Wherein the obtaining unit comprises:
Search for incorrect cases from the case database where the case is stored and define rules for the inaccurate case
/ RTI >
The rule is generated by selecting at least one condition and a conclusion, respectively
Wherein the knowledge acquisition system comprises: 14. The method of claim 13,
Wherein the obtaining unit comprises:
Accumulating the knowledge by storing in the knowledge base and the cornerstone database rules related to the cornerstone case and the cornerstone case generated by performing validation and verification of the incorrect case based on the defined rule
Wherein the knowledge acquisition system comprises:

Images