KR20170122146A - A method of case learning and comments generation for clinical pathology examination using rule optimization - Google Patents

Abstract

The present invention provides a technique which builds experimental knowledge (procedural knowledge, rule) of a process of giving an opinion by an expert from a number of existing clinical pathology examination examples as a procedural knowledge base (rule base), and automatically generates the opinion by a machine by using the same. According to an embodiment of the present invention, a method for learning expert knowledge and generating an opinion with respect to a clinical pathology examination through a rule optimization comprises the following steps of: collecting a clinical examination result and a corresponding expert opinion sentence as input data; building an example base by converting the input data into a database; calling an example from the built example base, and generating the opinion sentence for each example by using the prebuilt procedural knowledge base; processing adding/correcting/deleting by the expert with respect to the opinion sentence which requires modification among the generated opinion sentences; and reflecting the modified opinion sentence, and reflecting an optimized procedural knowledge with respect to the opinion modification to the procedural knowledge base.

Description

TECHNICAL FIELD The present invention relates to a method for generating knowledge and learning of expert knowledge on a clinical pathology test using rule optimization,

The present invention relates to a method of generating expert knowledge and knowledge for clinical pathology testing through rule optimization.

The Decision Tree is one of the decision support tools that represents a tree-like graph or model. It is often used to represent hierarchical rules. The basic form of the multiple-classification ripple-down rules (MCRDR), which is mainly referred to in the present invention, may be implemented in the form of a decision tree.

MCRDR is an extended version of RDR that allows multiple classification as one of the incremental knowledge acquisition techniques. It is a technique that is used to build a rulebase by extracting rules from a case-based expert system. The RDR technology is a rule structure that is useful for adding rules by removing the else structure and the negation (not) condition in the form of a rule consisting of if-then-else. This allows you to add / modify / delete rules by adding new rules. The advantage of this is that the intervention of the knowledge engineer is reduced because the rule base can be constructed only by adding a simple rule without a complex logic structure.

A case-based expert system and decision support system is a system of health information technology that helps doctors and other health professionals make decisions. It was mainly studied in artificial intelligence. Because of the nature of the medical domain, statistical methods can not be used for all cases. Therefore, the case-based approach has been tried. The present invention also uses the approach of extracting procedural knowledge from the case data and constructing a procedural knowledge base have.

Korean Patent Publication No. 10-2005-0008021 (published on Jan. 21, 2005) discloses a method of providing a clinical test result in a medical information providing system. The result of a clinical test of a user is transmitted to a doctor, a nurse , Laboratory staff, inspection equipment, general equipment terminal, external authentication system.

(Patent Document 1) KR1020050008021 A1

The present invention establishes empirical knowledge (hereinafter, procedural knowledge and rules) of a process of finding out experts from a large number of existing clinical pathology test cases into a procedural knowledge base (hereinafter referred to as a rule base) It is intended to provide a technique by which a machine automatically generates a remark.

According to an embodiment of the present invention, there is provided a method comprising: collecting a clinical test result and an expert's opinion statement as input data; Establishing a case base by converting the input data into a database; Calling a case in the constructed case base and generating a sentence statement for each case using a pre-built procedural knowledge base; Performing an addition / correction / deletion by a specialist on a sentence which needs to be corrected among the generated sentence sentences; And reflecting the modified refutation sentence and reflecting the optimized procedural knowledge of the refutation to the procedural knowledge base.

Recently, there is an increasing need to request the results of tissue tests as a result of faster clinical tests from primary medical institutions.

Usually, the primary medical institution is a medical institution that can be accessed for the first time as well as easily accessible by patients.

However, most of the primary medical institutions do not have specialized analysis systems for clinical pathology and histology, and they usually collect results from professional analytical companies and deliver the results to the relevant institutions.

It takes about two to three days to receive the test results from a specialized analysis company in the existing primary medical institution.

Therefore, professional analysts need to handle requests of many medical institutions efficiently in a short period of time.

By using the present invention, the machine can automatically generate the findings in the same manner as the experience knowledge of the existing experts by using the procedural knowledge base when generating the findings of the clinical pathology test cases.

Therefore, since the expert does not have to generate the repetitive sentences frequently, the efficiency of the work of creating the findings increases, and since the expert requires additional procedural knowledge only for the findings to be corrected or the missing findings, From the standpoint, procedural knowledge can be easily added without knowledge engineer involvement.

The procedural knowledge base is constructed in a structure similar to the Decision Tree structure,

When the procedural knowledge is added to the base, it analyzes the conditions of the rule based on the sentence sentence, finds the optimal position in the established rule base, and adds the procedural knowledge base. have.

The advantage of constructing an optimal scale procedural knowledge base can greatly reduce the frequency with which experts input duplicate procedural knowledge when adding / modifying / deleting sentences, So that a faster search can be performed.

According to the embodiment of the present invention, in the step of reflecting the optimized procedural knowledge on the correction of the findings into the procedural knowledge base, the procedural knowledge of the experts on the generation of the findings is optimized using the rule- Is the most important feature of the present invention.

By using the present invention, it is possible to increase the efficiency of doctors and pathologists in specialized analytical companies and ultimately receive faster clinical pathology tests and histological examination results in primary medical institutions.

In particular, the knowledge of experts such as medical diagnosis documents such as radiology, dentistry, and oriental medicine as well as medical diagnosis documents, nursing statements, procedures for medical procedures such as surgery, It is possible to apply in the field where the procedural knowledge base should be built from all the document cases that can be done.

The present invention, which is applied in the medical field, is characterized in that it is difficult to easily apply the related technology to other industries or companies because it closely uses personal medical information due to the nature of the medical industry.

However, the present invention can be widely applied in an industrial field such as the applied application field in which a procedural knowledge base is constructed by extracting expert procedural knowledge and applied to it.

In case of non-medical field, application of the present invention enables the expert knowledge of on-site experts in industrial sites and enterprises that require accumulation of experience knowledge of field experts, such as production line management in factory and error repairing process of nuclear control module, It can be constructed as a base.

FIG. 1 is a flow chart of a method for generating expert knowledge and knowledge for clinical pathology testing through rule optimization according to the present invention.
FIG. 2 is a block diagram illustrating the configuration of a system for generating expertise knowledge knowledge and findings for a clinical pathology test according to the present invention.
FIGS. 3 to 7 are flowcharts illustrating a method of generating expert knowledge and knowledge for the clinical pathology test through rule optimization according to the present invention, in more detail.

The following detailed description of the invention refers to the accompanying drawings, which illustrate, by way of example, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the present invention are different, but need not be mutually exclusive. For example, certain features, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the invention in connection with one embodiment. It is also to be understood that the position or arrangement of the individual components within each disclosed embodiment may be varied without departing from the spirit and scope of the invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is to be limited only by the appended claims, along with the full scope of equivalents to which such claims are entitled, if properly explained. In the drawings, like reference numerals refer to the same or similar functions throughout the several views.

The present invention relates to a method for generating expert knowledge and knowledge for clinical pathology testing through rule optimization, and a method for constructing a rule base based on MCRDR to obtain a physician and pathologist's opinion on the clinical pathology test from case data And a technique for accumulating the experience knowledge of field experts, and constructing an optimized procedural knowledge base using rule optimization techniques.

In the process of making this knowledge base, the present invention minimizes the intervention of the knowledge engineer, and it is essential to learn experiential knowledge inferring the opinions of a physician or a pathologist and construct a rule-based procedural knowledge base.

In the present invention, a rule base is constructed using MCRDR. MCRDR is a methodology that allows expert knowledge to be extracted in the form of a Decision Tree. This methodology is well suited to building expert systems because it allows the gradual addition of expert knowledge and the process of adding new knowledge does not affect existing knowledge.

Particularly, the present invention relates to a technique for a physician or a pathologist to provide an explanation to a patient or a test requester about a disease or symptom that may cause a normal or abnormal condition of a sample in a clinical pathological examination. For example, in the case of a blood test, a blood sample from one inspector can be taken and analyzed for more than 500 blood components, which can be used to describe the expected diseases and symptoms according to the values of the particular test item. As described in the above-mentioned expectation effect item, the number of inspectors and the test items become more complicated than the number of doctors and pathologists, and thus it is necessary to improve the efficiency of the duplicated work for frequently repeated observations. This paper focuses on techniques that mechanics learns the knowledge of physicians or pathologists and deduces the findings of new pathologic examinations or histological examinations.

Hereinafter, a method for generating expert knowledge and knowledge for clinicopathological examination through rule optimization according to an embodiment of the present invention will be described with reference to the accompanying drawings.

1 is a flow diagram of a method for generating expert knowledge and knowledge for clinical pathology testing through rule optimization in accordance with an embodiment of the present invention. A method for generating a paraphrase sentence according to an embodiment of the present invention is recorded in a programming language that enables each step to be executed in a computer and can be executed in a computer.

Referring to FIG. 1, an expert experience knowledge learning and finding generation system 100 for clinical pathology examination through rule optimization according to an embodiment of the present invention includes a step 200 of collecting case input data, A case base construction step 300 for storing the test results and the expert opinion texts in a database form, a step 400 for creating a sentence statement for each case by calling the cases from the accumulated case base, a field expert such as a doctor or a pathologist A step 500 of adding / correcting / deleting a sentence by reviewing a sentence refuted by the computer, and reflecting the added or modified sentence, and optimizing the rule in the step of inputting the reason why the sentence is made It is necessary to apply the optimization technique for the changed condition or the duplicated condition to reflect it in the existing procedural knowledge base. System 600 as shown in FIG.

In the present specification, the case data is data in which a result of a clinical pathology test such as a blood test result of a test requester is recorded, and is composed of hundreds of test items.

Case base S700) is a recording device in which case data is stored in a formalized form, and a sentence is a sentence in which a doctor or a pathologist describes the disease or analysis result about the case data.

The procedural knowledge base is a logical rule of the if-then structure, and consists of the conditions of the test items for inferring the sentence and the sentence.

In the embodiment, the expert knowledge knowledge learning and observation generation system 100 for the clinical pathology examination can be expressed by a program or an interface implementing the following embodiments.

It can be part of a module form of an existing clinical pathology system and can be constructed as an independent program.

The step of collecting the case input data 200 in the embodiment is a step for converting the unprocessed output form output from the existing pathology inspection system into a formalized form.

Clinical pathologic examination items and the results (numerical values, nominal values, sentence data) of the examinees are clinical examination items. Expert opinions written by experts are indexed in the form of a table, and the expert opinion sentence items . The analyzer (parser) should be implemented according to the output method of the existing clinical pathology analysis system, and the sentence table should be composed in an expandable form.

A case base construction step (300) for storing the collected case data into a database form by dividing the collected case data into a clinical test result and an expert opinion statement includes a clinical test case base (700) Each instance has a table column for each clinical test item, and the last expert opinion sentence column can contain several testimonials.

In the step 400 of generating the sentence statement for each case by calling the cases in the accumulated case base, the step of inferring the sentence sentence by calling the instances where the sentence item of the constructed clinical test case base is empty, See procedural knowledge base 800.

The procedural knowledge base consists of a set of rules of the if-then structure mainly because the clinician's test items that are the basis for the expert's findings and the conditions according to the numerical values of the items are specified in the procedural knowledge base Through these rules, the sentence can be deduced.

Thus, building the procedural knowledge base 800 is at the heart of the present invention. This procedural knowledge base can be expressed in the form of a rule-tree based on conditions with a hierarchy of parent and child rules.

In step 500, a field expert such as a doctor or a pathologist examines a sentence refuted by the computer and adds / corrects / deletes the sentence, the result of the sentence deduced by the machine as an existing procedural knowledge base becomes the actual data This is a step of reviewing whether there is a match, other findings, or inferred findings, thereby reducing errors in the clinical pathology test to be delivered to the test requesters or hospitals and creating a procedural knowledge base that is continuously improved .

The rules are optimized at the stage of reflecting the added or modified sentence and inputting the reason why the findings are made, so that they are reflected in the existing procedural knowledge base by applying the optimization technique for the changed condition or the duplicated condition In step 600, procedural knowledge made up of new conditions is added to a new or modified sentence, and the basis of the deleted sentence is defined as a rule, and then the add / modify / delete This is a step to reflect the matter.

This process requires an optimization process for the rules. There is a problem in the rule that duplicate rules for duplicate findings must be generated several times depending on the order of the conditions or the combination method. This problem is caused by the inefficiency of the work and the complexity of the rule to the field experts, Section.

In order to solve this problem, the rules are arranged in a hierarchical structure by sorting the conditions into minimum units, and the rules are made to be reflected as child nodes in a rule-tree constituting an existing rule base in an optimized form The giving includes steps.

As described above, according to the embodiment of the present invention, it is possible to learn the basis of the findings of the doctors or pathologists about the cases of the clinical pathology test, and to deduce the sentence about the new case. Using this, it is possible to prevent repetitive work on repetitive feature generation tasks in similar industries (medical and analytical fields), to improve the efficiency of errors or add-on work, Knowledge base can be constructed.

FIG. 2 is a block diagram illustrating the configuration of a system for generating expertise knowledge knowledge and findings for a clinical pathology test according to the present invention.

In the clinical examination case base 10 of the expert experience knowledge learning and finding generation system 100 for clinical pathology examination through rule optimization according to an embodiment of the present invention, the case input data according to the embodiment of the present invention is collected (200), and a case base construction step (300) of storing the collected case data into a database form by dividing the collected case data into clinical test results and expert opinion texts.

The stored case base 10 generates a sentence sentence (400) through the sentence generation unit (20), provides a doctors or pathologist with a sentence correction unit (30), adds a sentence sentence from a doctor or pathology , Modification or deletion.

The procedural knowledge base 40 is mainly used in the step of generating a sentence statement 400 and is used again by a doctor or a pathologist through a rule optimization step 600 And is reflected in the procedural knowledge base 40.

The above-mentioned contents will be described by way of example.

Experimental Experiences on Clinical Pathology Testing by Rule Optimization According to the Present Invention Experiential knowledge learning and generation methods include case data formatting, case base construction, inference inference and rule management interface, inductive PDR-based procedural knowledge base, A knowledge base, and rule disassembly and inspection that generate complex knowledge.

FIGS. 3 to 7 are flowcharts illustrating a method of generating expert knowledge and knowledge for the clinical pathology test through rule optimization according to the present invention, in more detail.

Referring to FIG. 3, the TXT file issued from the main system is analyzed to extract case information (patient code, sex, hospital code, and office information) of the patient and inspection items by code to form the case data.

Referring to FIG. 4, data having a schema is stored in a database as shown in FIG.

Referring to FIG. 5, a screen for providing a reasoning inference and a rule management interface for constructing a procedural knowledge base is shown as an example. The interface screen of FIG. 5 shows an example linked to the PDR rule by reading a TXT file as an example of a blood test.

Referring to FIG. 6, there is shown a structure of an induct PDR-based procedural knowledge base, which shows a method of generating a procedural knowledge base based on a machine learning.

As an example of an induct PDR-based procedural knowledge base, the following knowledge base can be generated by applying an anemia-related rule to the procedural knowledge base structure of FIG.

"Root THEN does not have anemia."

[1] "IF (Iron == nc) THEN There is no anemia and iron is not enough."

[ 2] "IF (Hb  == NA ) THEN  I do not have enough iron. "

[3] "IF (Hb == lc) & (MCV == nc) THEN Anemia has been found.

[4] "IF (Ferritin == lc) THEN Corresponds to iron deficiency anemia.

[5] "IF (Hb == hc) THEN Hemoglobin is high. Fasting or dehydration can result in blood congestion or erythrocytosis (smoking, intact, etc.)."

[6] "IF (Hb == lc) & (MCV == hc) THEN Corresponds to the state of treatment of anemia or anemia.

[7] "IF (Hb == nc) & (Ferritin == lc) THEN corresponds to an iron deficiency state."

[8] "IF (TIBCVu> 332.0) & (feVu <= 61.0) THEN Iron is normal, but transferrin saturation is low.

[9] "IF (Hb == lc) & (Ferritin == NA) THEN Anemia has been found.

[10] "IF (feVu> 51.0) & (TIBCVu <= 400.0) THEN There is no anemia and iron is not lacking."

[11] "IF (TIBCVu> 355.0) & (feVu <= 55.0) THEN Iron is normal, but transferrin saturation is low.

[12] "IF (feVu <= 46.0) & (TIBCVu> 300.0) THEN Iron is normal, but transferrin saturation is low.

[13] "IF (Hb == lc) & (sex == M) THEN Anemia has been found.

[14] IF (Iron == NA) THEN Anemia has been found. When I am anemic, I feel weak and I feel fatigued. "

[ 15] "IF (MCV  == nc ) THEN  Ferritin ( Ferritin ) Has increased . Since ferritin reflects the stored iron Iron excess  It may mean, As an acute reactive substance  Infection, inflammation, etc. As a cancer marker  Liver cancer, lung cancer or leukemia. "

[16] "IF (Hb == lc) THEN Anemia has been found. If you have anemia, you feel weak and boredom."

[17] "IF (Hb == hc) THEN Hemoglobin is high. Fasting or dehydration can result in blood concentration or erythrocytosis (smoking, intact, etc.)."

[18] "IF (Ferritin == lc) THEN corresponds to an iron deficiency state."

[ 19] "IF (Hb  == nc ) THEN  Ferritin ( Ferritin ) Has increased . Since ferritin reflects the stored iron Iron excess  It may mean, As an acute reactive substance  Infection, inflammation, etc. As a cancer marker  Liver cancer, lung cancer or leukemia. "

[20] "IF (Hb == lc) & (MCV == hc) THEN Corresponds to the state of treatment for anemia or anemia.

[21] "IF (Hb == hc) THEN Hemoglobin is high. Fasting or dehydration can cause blood to become enriched or red blood cells (smoking, intact, etc.)."

[22] "IF (sex == M) & (Ferritin == nc) THEN Ferritin is normal Ferritin reflects stored iron, so it can be said that iron is normal. Cancer markers, liver cancer, lung cancer, or leukemia, which can increase iron deficiency. "

However, this Induct PDR-based procedural knowledge base is learned about the whole test result, and there is a problem that there is no relation between the sentence and the test items, and the accuracy of the inferred sentence is much lowered. There is a problem that takes a lot of time in. Therefore, in order to solve this problem, it is necessary to build expert knowledge manually from the beginning into the procedural knowledge base.

Referring to FIG. 7, an expert manual addition based procedural knowledge base is shown.

And constructs a procedural knowledge base for manually acquiring expert knowledge based on the Induct PDR as shown in FIG.

An example based on the expert manual addition based procedural knowledge base shown in FIG. 7 is as follows.

Root THEN

[1] IF (11005> = 12.0) & (11005 <= 16.0) & (sex == F) THEN "There is no anemia."

[2] IF (11001 <4.0) THEN "White blood cell count (WBC) has been somewhat reduced.

[273] IF (11003_03> 44.0) THEN "WBC decreased and lymphocyte increased. Lymphocytes are increased by various infections (especially viruses and tuberculosis) and immune stimulation."

[3] IF (00802_06! = Negative) THEN "A ketone has been detected in the urine. It increases in diabetes, high fat meals, fasting in Guiana, severe exercise, and fever."

[91] IF (00802_06 == na) THEN

[122] IF (00802_06 == yet) THEN

[4] IF (00802_08! = Negative) & (00822_01! = 0-2) THEN "Urine hemoglobin and red blood cells have been detected. It is associated with urinary tract diseases such as menstrual bleeding, urinary tract stones, or severe exercise or asymptomatic hematuria "

[92] IF (00822_01 == na) & (00802_08 == na) THEN

[123] IF (00822_01 == yet) & (00802_08 == yet) THEN

[285] IF (sex == M) THEN "Urine hemoglobin and red blood cells have been detected. It is related to urinary tract diseases such as urinary calculi, severe exercise or asymptomatic hematuria."

[5] IF (00822_06 == amp.phosphate) THEN "Crystals have been found in the urine. Amorphous phosphate crystals can also be found at the top."

(6) IF (11003_01 <= 75.0) & (11001 <= 10.0) & (11003_05 <= 5.0) & (11001> = 4.0) & (11003_04 <= 10.0) & (11012> = 150.0) & (11003_02 = empty) & (11003_06 <= 3.0) & (11003_03 <= 44.0) & (11012 <= 450.0) THEN "

IF (00802_07 == negative) & (00822_06 == not_found) & (00822_01 == 0-2) & (00802_08 == negative) & (00802_06 == negative) & (00802_10 == negative) & (00802_04 == negative) & (00802_05 == negative) & (00822_03 == 0-2) & (00822_05 == not_found) & (00822_02 == 0-2) & (00822_07 == not_found) & (00802_02 == negative) & (00822_04 == not_found) & (00802_09 == norm) THEN "Urine test result is normal."

[8] IF (00802_09! = Norm) THEN "Urine urinary bilignon has been detected.

[93] IF (00802_09 == na) THEN

[124] IF (00802_09 == yet) THEN

[188] IF (00802_07 == positive) THEN "Urine bilirubin and urobilinogen have been detected and are associated with liver disease."

[9] IF (00822_02! = 0-2) & (sex == F) THEN "Urine is leukocyte-positive.It is related to urinary tract infection.In women, contamination by vaginal secretion is common.

Referring to Fig. 8, a rule exploded view for generating a composite feature is illustrated.

By decomposing the rule generation rules including two or more test codes individually, it is possible to efficiently classify the inefficient rules that define all 2 ⁿ rule generation rules into 6 rule sets (2 * 3).

Referring to FIG. 9, generation of inspection-related integrated findings is shown as an example.

It is possible to generate integrated findings for specific tests by integrating individual findings. The example of FIG. 9 illustrates the liver function findings as an example.

The features, structures, effects and the like described in the embodiments are included in one embodiment of the present invention, and are not necessarily limited to only one embodiment. Further, the features, structures, effects, and the like illustrated in the embodiments can be combined and modified by other persons having ordinary skill in the art to which the embodiments belong. Therefore, it should be understood that the present invention is not limited to these combinations and modifications.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of illustration, It can be seen that various modifications and applications are possible. For example, in the embodiment, the procedural knowledge base is expressed in the form of a rule-tree, which can be implemented in an ontology or a graph form. Solution procedures for problems in factory production lines, and the like. Each of these components can be modified and implemented. It is to be understood that all changes and modifications that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

100: Expertise knowledge knowledge and learning system for clinical pathology
10: Clinical test case base
20:
30: Remedy correction
40: procedural knowledge base
50: procedural knowledge optimizer

Claims (8)

A collection step of collecting the results of the clinical test and the experts' sentences as input data;
A building step of building a case base by converting the input data into a database;
A generation step of generating a sentence statement for each case by calling a case in the established case base and using the established procedural knowledge base;
A correction step of performing an addition, correction or deletion correction by a specialist on a recognized sentence which needs to be corrected among the generated sentence sentences; And
And a reflecting step of reflecting the corrected remedial sentence and reflecting the optimized procedural knowledge on the correction of the findings into the procedural knowledge base. The method according to claim 1,
In the collecting step,
A method for generating expertise knowledge knowledge and findings for a clinical pathology test, wherein the input data type is separated from the type of data output from the clinical examination system and analyzed and collected. The method according to claim 1,
In the constructing step,
A method for generating expertise knowledge knowledge and findings for clinical pathology testing, wherein the clinical test result and the expert opinion statement are instantiated from the input data and inserted into the case base. The method according to claim 1,
In the generating step,
Expert experience knowledge of the clinical pathology test, which inference or generate the findings according to the rules that are invoked with reference to the pre-built procedural knowledge base for each case or case of the constructed case base How to create learning and observations. The method according to claim 1,
In the correction step,
Clinical pathology, which adds, modifies, or deletes procedural knowledge by adding a new rule based on the case to the field expert (physician or pathologist) for errors or omissions in the speculation or generated sentence Knowledge Expertise Knowledge Learning and Creation Method for. The method according to claim 1,
In the reflecting step,
After extracting the procedural knowledge of the correction findings from the corrected findings, we optimize the conditions of the procedural knowledge and search for the optimal position in the RDR type decision tree and add it to the child node Experiential knowledge knowledge and learning method for clinical pathology. In the expert experience knowledge learning and observation generation method for the clinical pathology test described in claim 3,
And storing the constructed case instance. In the method for generating expertise knowledge knowledge and findings for the clinical pathology test according to claims 4 and 6,
A computer readable medium in which the clinical pathology test item, the condition of the numerical value, and the sentence corresponding to the condition are recorded in the form of procedural knowledge (rule) so that the above findings can be inferred or generated.

Images