KR101642031B1 - An expert system for determining penalty and operating method thereof - Google Patents

Abstract

A sentencing expert system according to an embodiment of the present invention includes a first database that stores sentencing information corresponding to a type of crime, a second database that stores a correspondence relationship between the sentence and the sentence, , A knowledge base that stores result data that includes a method of applying to at least one of a legal, precedent, or supreme court sentencing criterion determined by experience or knowledge of a legal professional; And a control unit for calculating the sentence by combining input information for the sentence factor and information stored in the first database, the second database, and the knowledge base.

Description

≪ Desc / Clms Page number 1 > AN EXPERT SYSTEM FOR DETERMINING PENALTY AND OPERATING METHOD THEREOF &

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sentencing expert system for determining a sentence for an offense, and more particularly to a sentencing expert system using an artificial intelligence technology and an operation method thereof.

Sentencing means determining the type of punishment and sentence in a criminal trial. The sentence shall be decided by the judge in personal judgment, taking into account the facts and provisions of the law. As such, when the sentence is influenced by the rule of the judge, social disputes have arisen due to inconsistent sentence, even though it is sometimes similar.

In order to solve this problem, the Supreme Court of Korea has established a sentencing committee and proposes guideline standards that can be used by judges to derive rational sentences for 23 crimes.

The sentencing guidelines proposed by the Commission are more specific than the provisions of the law and provide a narrower range of penalties than the provisions of the Act. However, It is true that the sentence is not drawn straight away, and the scope of the sentence is still wide. Therefore, judging the final sentence still requires a considerable degree of judicial judgment.

In the end, the sentencing guidelines of the Supreme Court provide guidelines to assist judges in some subjective judgments, but they do not offer a narrow range of punishments or specific punishments. The judge is concerned about the specific application of a number of sentencing factors in the exercise of broad discretion. Moreover, in the case of the general public without legal knowledge, it is practically impossible to predict the type and amount of punishment expected of the perpetrator

The present invention proposes a screen for an information about sentencing factors to a user. When a user inputs sentencing information about a specific crime issue, the user receives the inputted information and applies it to an expert system composed of a database and a knowledge base in the system, A sentencing expert system for presenting a sentence and an operation method thereof.

A sentencing expert system according to an embodiment of the present invention includes a first database that stores sentencing information corresponding to a type of crime, a second database that stores a correspondence relationship between the sentence and the sentence, And the result data corresponding to the condition data, wherein the result data is information that reflects the application method of at least one of laws, precedents, or sentencing criteria determined by experience or knowledge of legal experts Base; And a control unit for calculating the sentence by combining input information for the sentence factor and information stored in the first database, the second database, and the knowledge base.

A method of operating a sentencing expert system according to an embodiment of the present invention includes receiving input information on a crime and a sentence factor, a first database storing sentence factor information corresponding to a type of a crime, A second database for storing the correspondence relation, a second database for storing the correspondence relationship, a second database for storing the correspondence relation, Calculating the sentence score by referring to the application result, calculating the range of punishment based on the sentence of the Supreme Court with reference to the application result, And determining the sentence by combining the calculated sentence score and the range of the sentence.

The sentencing expert system and its operation method according to the present invention allow a user to easily input information on a sentence factor through a screen and apply input information to an internal expert system to provide concrete and consistent sentence information, It can be a tool to provide direct reference to the general public.

The sentencing expert system and its operation method according to the present invention can provide a simulation function to analyze the sensitivity to the sentence while changing each sentence factor, thereby grasping the relative importance of sentence factors.

The sentencing expert system and its operation method according to the present invention can automatically generate a result equivalent to a judgment sentence by applying inputted information to an expert system.

1 is a block diagram of a sentencing expert system in accordance with an embodiment of the present invention.
2 is an example of data stored in the first database of Fig.
Figures 3a and 3b are examples of data stored in the second database of Figure 1;
Figure 4 is an example of data stored in the knowledge base of Figure 1;
5 is a flowchart showing the operation of the system control unit of FIG.
6 is an example of a sentence input screen provided to the user;
7 is a flowchart showing the operation of the sentence control unit of Fig.

Hereinafter, embodiments according to the present invention will be described with reference to the accompanying drawings. In the following, like reference numerals designate substantially identical components.

1 is a block diagram of a sentencing expert system 1000 in accordance with an embodiment of the present invention.

The sentencing expert system 1000 according to an embodiment of the present invention includes a control unit 100, a first database 200, a second database 300, and a knowledge base 400.

The first database 200, the second database 300, and the knowledge base 400 are logical identifiers, which may be implemented to be physically contained in one database or embodied in a separate database, depending on the embodiment It is possible. For example, the first database 200 and the second database 300 may be included in separate database servers, may be separate database structures included in one database server, or may be separate database structures included in one database structure It may be a table structure.

The sentencing expert system 1000 according to an embodiment of the present invention includes an input and output interface 500 for transmitting and receiving data to and from a local user 10 and an input and output interface 500 for transmitting and receiving data to and from the remote user 30 via the network 20. [ And may further include a network interface 600. The remote user 30 collectively refers to a terminal of a remote user connected to the network 20.

In the present embodiment, the input / output interface 500 collectively refers to an input device such as a keyboard and a mouse, an interface for the input device, an output device such as a monitor and a printer, or an interface therefor.

The network 20 may include a local area network (LAN), a wide area network (WAN), the Internet, and the like.

The control unit 100 controls the operation of setting the basic information necessary for the operation of the sentencing expert system 1000 in the first database 200, the second database 300 and the knowledge base 400, The information on the crime provided by the local user 10 or the remote user 30 connected to the first database 200 and the information on the second database 300 and the knowledge base 400, And control the operation of outputting it.

The users 10 and 30 can be authorized by the system administrator or the general user. For example, the operation of setting up the first database 200, the second database 300, and the knowledge base 400 may be limited to a user having administrator authority.

The control unit 100 may include a system control unit 110 and a sentential control unit 120.

The system control unit 110 can control the operation of setting basic information necessary for operation of the sentencing expert system 1000 in the first database 200, the second database 300 and the knowledge base 400. [

The system control unit 110 may provide an input screen for basic information to the local user 10 or the remote user 30 through the input / output interface 500 or the network interface 600. [

When the local user 10 or the remote user 30 inputs basic information requested through the screen and issues a storage command, basic information is provided to the system control unit 110 via the input / output interface 500 or the network interface 600 .

The system control unit 110 may divide the received basic information into the first database 200, the second database 300, or the knowledge base 400 according to the type of the received basic information. Also, the system control unit 110 receives the crime selection information provided by the users 10 and 30 connected to the sentencing expert system 1000 and can provide an input screen necessary for inputting necessary sentencing factor information for the selected crime have.

The sentence control unit 120 receives the crime selection information provided by the local user 10 or the remote user 30 connected to the sentencing expert system 1000 and the sentence factor information inputted for the selected crime, Output result applied to the first database 200, the second database 300 and the knowledge base 400 and outputs the result to the local user 10 or the remote user 30 through the input / output interface 500 and the network interface 600. [ Lt; / RTI >

As described above, the first database 200, the second database 300, and the knowledge base 400 store basic information necessary for operation of the sentencing expert system 1000.

More specifically, the first database 200 stores various types of sentencing factors defined for each type of crime and information on sentencing factors associated with each type of sentencing type.

In the following, to avoid confusion of terms, the type of sentence type is referred to as the first stage sentence type, and the sentence type associated with the type of sentence type is referred to as the second stage sentence type.

One crime is associated with one or more first-stage sentence factors, each of which can be weighted.

The second database 300 stores the relationship between the second-stage sentence defined in the first database 200 and the legal statement and punishment. In addition, the second database 300 may store the relationship between the two or more second-stage sentence factors defined in the first database 200, the legal statement, and punishment by applying the knowledge base 400. [ The second database 300 can store information on the types of crime classified by the Supreme Court sentencing standard and information on the range of the sentence to be applied in basic, weighted, and sensitized cases according to each type.

The knowledge base 400 structurizes and stores at least one application method of at least one of laws, judgments, and sentencing standards based on experience and experience of legal experts required in the application process of the law, for example, in the form of conditional statements.

For example, the knowledge base 400 may store legal knowledge about how to apply each legal text if a plurality of legal texts can be applied by the second-stage sentence factors selected by the sentential factor information input by the user .

For example, the knowledge base 400 may be configured to select and apply penalties to one of a plurality of legal texts having high priority (imaginary contention) or to combine and apply penalties defined in a plurality of legal texts (substantive contention) Can be structured and stored in conditional form.

In addition, knowledge base 400 may include a method for selecting a type considering the culpability of a legal statement if a specific combination of the second-stage positive factors is satisfied, a method for determining the type of the sentencing criterion, The method of selecting the reduction / weight / basic sentence, and the determination of whether to suspend execution can be structured and stored as conditional form data.

Hereinafter, the structure of data stored in the first database 200, the second database 300, and the knowledge base 400 will be described with reference to FIGS.

FIG. 2 illustrates a data structure stored in a first database 200 according to an embodiment of the present invention.

The first database 200 stores information on the type of crime, the first-stage sentence for each crime, the weight given to the first-stage sentence, the second-stage sentence under the first-stage sentence, Corresponding weighting and reduction / basic / weighting information.

In this embodiment, the crime is divided into 23 categories including murder, sex, bribery, fraud, violence, traffic, and firearms, and an identification code may be assigned to each crime. For example, the identification code 1 may be assigned to the murder, the identification code 2 may be assigned to the gender, and the identification code 19 may be assigned to the traffic.

Each crime may correspond to one or more first-level sentence factors. For example, nine types of first-instance factors can be assigned to traffic: victim's personal information, suspect status, pre-suspect status, type of accident, type of violation of law, victim's fault, level of consensus,

Each of the first stage sentence factors may be assigned an identification code, which may include identification code information for the crime associated with the first stage sentence factor. For example, an identification code may be assigned to the suspect person (19, 1), and to the suspect situation (19, 2).

Figure 2 discloses only the first stage sentinel factors associated with traffic crime, but the first stage sentence factors may be related to other crimes in a similar fashion.

Weight information is associated with each first-step positive factor. For example, 0.1 for victims and 0.1 for suspects. The sum of the weights corresponding to the first-step positive factors associated with one crime is set to one.

The first stage positive determinant may be associated with one or more second stage positive determinants. The second stage sentence factor may also be assigned an identification code, which may include information about the crime, the identification code of the first stage sentence factor.

For example, in the present embodiment, the second stage of sentencing is associated with the first stage sentence (19, 1), which is the first stage sentence factor, and is related to the mind and sick person, the mental and physical handicapped person, 1, 1), and (19, 1, 2) are assigned to the minor and minor.

Weight information may be added to each second stage sentence factor. In this embodiment, a real number from 1 to 10, for example, is given as a weight to the second-stage sentence factor.

In addition, each stage 2 sentence factor may be associated with sentence / weight information of the Supreme Court sentence standard. Sensitivity / weighted information can be used to determine the sentencing range of the Supreme Court sentencing standard. Sensitivity / weighting information can be used for statutory reduction. In the present embodiment, the second step-type sentinel, a person with mental or physical disabilities, is subjected to a total reduction, a sick person, a deaf-mute person, and a juvenile offender. Basis or weight information determined by the Supreme Court sentencing criteria, which can be judged by combining a plurality of second sentence sentence factors, unlike the sentence by law, can be stored in the knowledge base 400.

In the present embodiment, only the reduction / basic / weight information is exemplified, but other types of information corresponding to the second stage positive factor may be stored in the first database. For example, each second stage sentence factor may be associated with information on whether the Supreme Court sentence is on probation.

The type of crime included in the first database 200, the first-stage sentinel factors associated with each crime, the weights corresponding to the first-stage sentinel factors, the second-stage sentinel factors associated with the respective first-stage sentinel factors, Weight / weight information corresponding to the second-stage sentence factor, and weight / weight information corresponding to the second-stage sentence factor may be added, changed or deleted by the information setting operation of the system control unit 110. [

3A and 3B illustrate data structures stored in a second database according to an embodiment of the present invention.

The second database 300 may store data related to the relationship between the second-stage sentence factor and the legal statement.

The sentence control unit 120 can inquire the second database 300 by applying the knowledge base 400 to be described later when the second-stage sentence factor is selected, thereby confirming the applicable text and the penalties defined therein.

One legal statement may correspond to one second stage sentence factor and may correspond to more than one second stage sentence factor. Each legal text has a corresponding penalty.

For example, referring to FIG. 3A, the traffic accident handling special procedure (hereinafter referred to as the school special law) Article 3 (exception of punishment) corresponds to one lethal factor (19, 7, 7) Article 5-3 of the Weighted Punishment Act (hereinafter referred to as the Special Price Act) corresponds to two lethal factors (19, 7, 8) and escapes (19, 9, 11).

For example, referring to FIG. 3B, Article 148-2 of the Road Traffic Act corresponds to one drunken driving factor (19, 5, 15) as a second stage driving factor, and 11 of the special price method 11 corresponds to two second- (19, 5, 15) and dental images (19, 7, 8).

For example, the sentence control unit 120 may inquire the second database 300 when the information related to the lethal (19, 7, 7) is inputted as the second-stage sentence factor, . Also, the sentence control unit 120 inquires the second database 300 when the lethal sentences 19 and 78 and the escapes 19, 9 and 11 are simultaneously input as the second-stage sentence factors, Can be determined.

Although not shown, the second database 200 may further include types of crime types determined by the sentencing standards of the Supreme Court's Committee, and information on the range of sentences applied to weight / reduction / basic sentence included in each type. However, it is preferable that information on how to determine a type from the inputted second-stage sentence factors or how to apply the weighted / reduced / basic sentence is stored in the knowledge base 400 described later.

FIG. 4 illustrates a data structure of a knowledge base 400 according to an embodiment of the present invention.

For example, when information on a plurality of second-stage sentence factors is provided, one or more applicable legal statements can be derived by querying the second database 300 using the information.

The knowledge base 400 stores the method of deriving the legal statement in various conditions by reflecting the knowledge of the legal experts. Also, if one or more legal texts are derived, there may arise a problem of how to apply the penalties specified therein.

For example, if a plurality of applicable legal statements are derived as a result of querying the second database 300, the knowledge base 400 may select one of the legal statements having a higher priority to obtain applicable penalties Or a combination of two or more of these texts may be used to store information on the conditions for deriving applicable penalties.

In addition, the knowledge base 400 may store information on conditions under which a penalty will not be applied, clues on a penalty, conditions under which the penalty will be applied, and the like, depending on the cues of the legal text.

Thus, the knowledge base 400 comprehensively stores the conditions related to the application method of the regulations, thereby enabling an accurate sentence calculation by artificial intelligence technology.

The knowledge base 400 may include the condition data IF on the condition and the resultant data THEN on the penalty to be applied when the condition is satisfied.

The condition data can be defined by a logical expression composed of a plurality of second-stage positive-off factors and a plurality of logical operators such as and, or, not, as shown in FIG.

For example, the condition data "(19,5,15) and (19,7,8) or (19,7,9) (19, 7, 8) or (19, 7, 9) are input.

The result data corresponding to the condition data may include the content of the penalty derived from the application of a number of laws. For example, the result data corresponding to the above-mentioned condition data is "Penalty = (imprisonment for 10 years or fines of 5 million won or more and 10 million won or less, Can be stored together.

The result data corresponding to the condition data may include information about whether or not the request is filed. In the second and third cases of the four cases shown in FIG. 4, the result data corresponding to the condition data is stored as "not filed" and indicates that filing is not performed if the condition is satisfied.

In the case of the third case, the condition data and the result data reflecting the provisions of Article 4 of the Kyoiku Law and the precedents related thereto are exemplified.

Article 4 (Exceptions in case of joining insurance) of Article 4 of the Special Education Act (1) means that a car causing a traffic accident is classified under the Insurance Business Act Articles 4 and 126 to 128, Land Improvement Act Article 8 or Article 51 In the case of insurance or deductible pursuant to the provisions, a public prosecutor shall not file a complaint against the driver of the car who committed the crime provided for in the main sentence of Article 3, paragraph 2. However, This shall not apply to cases where the contract or deduction contract is nullified or terminated, or the liability of the insurer or the deductible provider for insurance or deductible payment is no longer obligated by the contractual exemption, etc. <Amended on Aug. 8, 1984, Aug. 30, (1) of Article 4 (1) of the Act, a person shall be deemed to be a victim of a business accident or serious traffic accident, And the part that stipulates that it can not be filed in cases where it has reached the gold medal damages is a violation of the Constitution. "

In the case of the third case, the condition data is a logic combination of the second stage sentence factors, in which the above-mentioned judgments and the judgment conditions of the judgments are not allowed to be filed.

The knowledge base 400 may divide the conditions for determining the type determined in the sentencing criterion of the Supreme Court and the corresponding application results into condition data and result data and store them.

The Supreme Court 's guideline classifies each type of crime into a plurality of types, and divides each type into three categories: basic, weighted, and reduced. It is preferable that the sentencing expert system according to the present invention is also determined so as not to deviate from the sentencing standards of the Supreme Court.

The knowledge base 400 may express the condition data to be applied to each type presented by the sentencing criterion of the Supreme Court in a logical combination of the second-stage sentence factors, similarly to the above-described examples. The fourth example in FIG. 4 illustrates the condition data to which the fourth type of traffic crime is applied by sentencing criteria.

Also, the knowledge base 400 can express the condition data for applying the basic, reduced, and weighted for each type presented by the sentence criterion by a logical combination of the second-stage sentence factors.

5 is a flowchart showing a basic information setting operation by the system control unit 110 of FIG.

The sentencing expert system 1000 according to the present embodiment can perform a sentence operation only when basic information is set in the first database 200, the second database 300, and the knowledge base 400.

The basic information setting operation can be performed before the system is used, or when there is a change of laws, a change of precedent, or a change of the sentencing guidelines of the Supreme Court during use of the system. It can also be performed for fine adjustment of the sentence calculated in the sentencing expert system 1000.

The system control unit 110 may indicate to the local user 10 or the remote user 30 through the input / output interface 500 or the network interface 600 to select an object to be set up (S100). The basic information setting operation is preferably performed by the local user 10 and the remote user 30 having the administrator authority.

If the user selects to set the first database 200, the system control unit 110 provides the first database setting screen to the users 10 and 30 (S111).

The first database configuration screen may be a web page comprising a plurality of HTML input elements.

When the users 10 and 30 enter data on the setting screen and complete the setting using the input button, the correction button, the delete button, etc., the system control unit 110 can receive the command and information inputted by the user (S112 ).

The system control unit 110 may add new information to the first database 200 using existing data, modify the existing information, or modify existing information using the transferred data according to a command type such as input, modification, (Step S113).

The basic information setting operation for the second database 200 and the basic information setting operation for the knowledge base may also proceed in a similar manner as described above.

For example, if the user selects to set up the second database 300, the system control unit 110 provides the second database setting screen to the users 10 and 30 (S121).

When the users 10 and 30 enter information through the setting screen and issue a command, the system control unit 110 can receive commands and information (S122).

The system control unit 110 may store information in the second database 300 according to the received command and information (S123).

If the user selects to set the knowledge base 400, the system control unit 110 provides the knowledge base setting screen to the users 10 and 30 (S131).

When the users 10 and 30 input information through the setting screen and issue a command, the system control unit 110 can receive commands and information (S132).

The system control unit 110 may store information in the knowledge base 400 according to the received command and information (S133).

The system control unit 120 provides a crime selection screen through the input / output interface 500 or the network interface 600. When the user 10 or 30 selects the crime, the system control unit 120 applies the first database 200 to the crime selection information and provides a corresponding sentence input screen.

FIG. 6 shows a sentence input screen according to the present embodiment.

FIG. 6 is a sentence input screen corresponding to a traffic crime; the first stage sentence factors listed on the horizontal axis at the top of the screen, and the second stage sentence factors corresponding to the first stage sentence factors listed in the vertical direction of the screen.

In the illustrated embodiment (alternatively) indicates that one of the plurality of second-stage positive factors corresponding to the first-stage positive determinant is selected and the (optional) one of the plurality of second- Indicates that more than one is selected and input.

In this embodiment, when the user 10 or 30 moves the cursor on the second-stage sentence factor, the information-describing window corresponding to the second-stage sentence factor is popped up, thereby assisting the user.

When all the required information is input, the user can press the "Determination of the sentence amount" button at the top of the screen to transmit information about the sentence factor to the sentence control unit 120.

The input screen for the above input may be provided in a form capable of inputting all the positive factors in one screen or may be provided in a form capable of being input by the first-stage positive factor or the second-stage positive factor. In the latter case, you can provide more explanation on the input screen.

The sentence control unit 120 determines the sentence by combining the information of the sentence factors transmitted according to the input of the user and the information of the first database 200, the second database 300, and the knowledge base 400.

The sentence control unit 120 can transmit the determined sentence information to the users 10 and 30.

The operation sequence of the control unit 100 will be described in detail below.

FIG. 7 is a flowchart showing an operation procedure of the control unit 100. FIG.

The system control unit 110 provides an input screen to the users 10 and 30 (S201).

The sentence control unit 120 receives the sentence information about the second-stage sentence factors input by the user 10, 30 (S202).

The sentence control unit 120 processes three operations in parallel or sequentially using the sentence factor information input by the user.

Steps S210 to S212 of FIG. 7 illustrate the process of reflecting the sentencing criteria set by the Supreme Court sentencing committee during the operation of the control unit 100.

First, the sentence control unit 120 analyzes the first database 200, the second database 300, and the knowledge base 400 based on the received sentence factor information, and determines whether any of the types of crimes currently committed by the Supreme Court sentencing committee (S210).

As shown in FIG. 4, knowledge base 400 may determine the type of sentencing criteria by logical combination of a plurality of second stage sentence factors.

Thereafter, the sentence control unit 120 determines whether the weighted / reduced / basic type is determined for each type determined by the sentence criterion (S211). For this, the sentence control unit 120 may refer to the sentential factor information and the first database 200, the second database 300, and the knowledge base 400, which are input.

For example, in the knowledge base 400, if the evaluation result for the special weighting factor indicates that there are two or more specialists in the event corresponding to the weighted area, or if the specialist is more than two special exception factors, (The number of special weight factors) - 2 ≥ (the number of special reduction factors) "and the result data" (the range of the sentence) = 1.5 (the upper range of the sentence range) " .

The correspondence relationship between the second stage sentence factor and the special weighting factor or special reduction factor can be stored in the first database 200. However, if a logical combination formula of a plurality of second-stage positive factors is required to determine a special weight factor or a special reduction factor, this can also be stored as condition data and result data in the knowledge base 400.

The sentence control unit 120 can check the number of special weighting factors and the number of special reduction factors by referring to the input sentential information and the information of the second database 200 or the knowledge base 400, It is possible to determine whether the weighting condition is established and the contents of the result data according to the weighting condition.

Steps S220 to S221 of FIG. 7 of FIG. 7 show the operation of calculating the number of jumping points of the sentence control unit 120. FIG.

The sentence control unit 120 quantifies the sentence factor by comparing the sentence factor information with the information of the first database 200 (S220).

The quantification of the sentinel factor is performed by weighting the weight w _i given to the i th first-stage sentence factor stored in the first database 200 inputted before the sentence score calculation to the weight w _ij of the j th second sentence grade factor Based process.

More specifically, in the present embodiment, the sentence factorization operation is to determine the maximum influence (A) when only one first-level positive affect acts and multiply it by the original weight to set the new weight.

In this embodiment, the quantification operation of the sentence factor is performed as follows.

As described above, the first database 200 stores the values (w _i ) of 0 to 1 in each of the first-stage positive factors, and the sum of the weights of the first-stage positive factors corresponding to one crime is For example, 1. Also, the first database 200 stores relative importance (w _ij ) for each of the second-stage sentence factors as a real number from 1 to 10.

The sentence control unit 120 analyzes the inputted sentence factor information and assigns a value (di) of 0 or 1 to each of the first-stage sentence factors. This value di is set to 0 if the value of the corresponding second-stage positive factor is selected as the minimum value for the first-stage positive factor, and to 1 otherwise.

Next, the sentence control unit 120 calculates the number d of the first-stage positive sentence factors that are not selected as the minimum value.

(N is the total number of first-stage positive factors associated with the crime)

Next, the sentence control unit 120 obtains the sum (Y) of the weights that substantially affect the sentence.

Next, the sentence control unit 120 calculates the influence index e (D) when d is not the minimum value and acts.

(D is the total number of first-stage positive affecting the sentence.

Next, the sentence control unit 120 calculates the reduction ratio E (d) of the weight sum when d first-stage positive factors act.

Finally, the sentence control unit 120 decreases the value of the weight w _i of the first-step positive sentence factor by the decrease rate.

When the quantification step of the sentence factor ends, the sentence control unit 120 calculates the sentence score (S221).

In this embodiment, the calculation of the sentence score can be determined by the following equation.

K is the number of second-stage positive factors corresponding to the i-th first-stage positive affirmative factor, i _j is 1 if the _j-th negative second-stage positive definitive factor corresponding to the i-th first positive affirmative factor is input, = 0

The above Equation (6) may be applied without applying Equations (1) to (5) according to the embodiment.

Steps S230 to S234 of FIG. 7 are associated with the determination sentence creating operation of the sentence control section 120. [

In criminal trials, judgments generally have a formalized format that includes:

(1) the legal text, (2) the choice of the type, (3) the weight of liquor, (4) legal relief, (5)

The steps S210 to S212 of FIG. 7 are performed by combining the input sentential information and the first database 200, the second database 300, and the knowledge base 400 to generate information .

First, the sentence control unit 120 selects an associated legal statement (S230).

As described above, the second database 300 stores the relationship between the sentence in which the penalty is defined and the second-stage sentence factor. The knowledge base 400 stores a method of selecting a legal statement to be applied by the logical combination of the second-stage sentence factors. For example, the knowledge base 400 can be applied to cases in which a problem of imaginary competition occurs, such as when a plurality of legal statements can be applied to one action.

Thus, the sentence control unit 120 combines the inputted second-stage sentence factors with the contents of the second database 300 and the knowledge base 400, thereby generating (1) the text information Can be determined.

Next, the sentence control unit 120 selects the type of the mold (S231).

The sentence control unit 120 can select one species by referring to the knowledge base 400 when the selected legal text defines a plurality of species. Therefore, the sentence control unit 120 can determine the text information that can be input to the selection item of the type (2) of the judgment sentence.

Next, the sentence control unit 120 performs a lounging weighting operation (S231).

Article 35 of the Criminal Code reads as follows: "① A person who has received a sentence of more than a sentence or has committed a crime equivalent to more than a sentence within three years after the exemption of the sentence is exempted from punishment as a loser. And weigh up to two times. "

In the present embodiment, the second database 300 stores the correlation between the statutory text of the Article 35 of the Criminal Procedure Act and the corresponding second stage sentence factor. In addition, the knowledge base 400 stores result data for applying the condition data for the case of judging as a nuisance and the case in this case twice as long as the term.

Therefore, the sentence control unit 120 may transmit the sentence information, which is input to the sentence weighting item of the judgment sentence (3) by comparing the input sentence information with the sentence information of the first database 200, the second database 300 and the knowledge base 400 You can decide.

Next, the sentence control unit 120 performs a legal reduction operation (S233).

The legal limit is to carry out the reduction operation prescribed in the law, and the second database 200 can store the reduction information determined by the legal statement. Information on the method of legal reduction can also be stored in the knowledge base 400.

The sentence control unit 120 performs legal reduction using the input sentence information and the information of the first database 200, the second database 300, and the knowledge base 400, thereby inputting a legal reduction item of the judgment sentence Can determine the textual information that can be generated.

The sentence control unit 120 determines the range of the legal disposition type determined by the law as a result of the selection of the type, the weighting of the loser, and the legal reduction of the type determined before (S234).

The decision further includes the reason for the decision by the Supreme Court sentencing standards and the order for which the final sentence was determined.

Information on the determination of crime types by the Supreme Court sentencing standards and the reduction / default / weighting of the determined types is already determined and can be used to generate textual information corresponding to items on the sentencing guidelines of the Supreme Court.

Also, the order of the judgment sentence is the part in which the contents of the final sentence are written. In the sentencing expert system according to the present invention, the sentence is automatically determined and presented in step S203 to be described later. By using this, text information corresponding to the order of judgment can be generated.

Step S203 to step S204 of FIG. 7 show the final operation of the sentence control unit 120.

First, the sentence control unit 120 determines the sentence by applying an interpolation method to the result of the above-described left and middle actions (S203).

In this embodiment, the interpolation method can be applied by the following three methods. In each case, the method of determining the amount of the punishment can be changed.

The first method is a method of correlating the jamming score 0 and the jamming score 1 for the lowest jam and the highest jam according to the type determined according to the guideline in step S210.

The second method is a method of correlating the declination score 0 and the declination score 1 with respect to the lowest value and the highest value among the reduction, basic or weighted amount determined in step S211.

The third method is to score the sentence scores for the six sentences determined in steps S210 and S211 (the lowest sentence, the lowest sentence, the lowest sentence, the highest sentence, the highest sentence, the highest sentence) , P3, P4, P5, and P6.

For example, P1 represents the sentence score obtained in the sentence score calculation step (S221) when a combination of typical second-level sentence factors that can yield a sentence corresponding to the lowest sentence level is input.

In this embodiment, when the first and second interpolation methods are applied, the sentence is determined as follows.

In Equation (7), the sentence score uses the value determined in step S221.

In the present embodiment, when the third interpolation method is applied, the sentence is determined as follows.

In the equation (8), the sentence score uses the value determined in step S221.

Although the method of determining the punishment according to the interpolation method is exemplified in this way, the concrete method for determining the interpolation method and the punishment amount can be changed according to the embodiment.

The sentence control unit 120 provides the final result to the users 10 and 30 (S205).

The final result provided by the sentence control unit 120 to the user may include only information on the final sentence.

The final result that the sentence control unit 120 provides to the user may be provided in a form similar to a judgment sentence as described above. At this time, the determination reason related to the sentencing of the Supreme Court can be determined while proceeding from step S210 to step S211, and the content of the final sentence corresponding to the order is determined based on the sentence information determined through step S203 .

As described above, the weight information corresponding to the first-stage sentence factors and the second-stage sentence factors of the first database 200 may have a large influence on the final sentence decision despite minor differences.

The system controller 110 according to an embodiment of the present invention can perform a calibration operation. In the calibration operation, the system control unit 110 receives the same sentence factors as those in the related examples using the conventional sentencing examples as a standard.

Then, the sentence control unit 120 determines the sentence by the above-described method. The system control unit 110 prepares the results of the standard sentencing cases and the sentence determined by the sentence control unit 120 while modifying the weight information set in the first database 200 according to a predetermined algorithm, The set can be determined.

At this time, it may be repeatedly performed by an ordinary descriptor to determine an algorithm for changing the weight information or an optimal weight set, and a conventional statistical analysis method may be applied.

The calibration operation can also be applied to the weights, but can also be applied to selecting the interpolation method applied in step S203. The concrete method can be performed in a similar manner to the process of selecting the optimum values of the weights described above.

In the foregoing detailed description, embodiments of the present invention have been specifically disclosed with reference to the drawings. The above description is for the explanation of the present invention, and the scope of the present invention is not limited by the above description. The scope of the present invention is defined by the scope stated in the claims and equivalents thereof.

1000: Ops Expert System
100:
110:
120:
200: first database
300: second database
400: Knowledge base
500: Input / output interface
600: Network interface
10: Local users
20: Network
30: Remote users

Claims (20)

delete delete delete A first database for storing sentencing information corresponding to the type of crime;
A second database for storing a correspondence between a legal statement and a logical combination formula of one or more of said sentential factors;
And the result data is stored in the form of at least one of a legal, precedent, or sentencing criterion determined by experience or knowledge of a legal expert Knowledge base reflecting application method; And
A control unit for calculating a sentence by combining input information for the sentence factor and information stored in the first database, the second database, and the knowledge base;
, &Lt; / RTI &
Wherein the sentence information includes information on a first stage sentence factor associated with the crime and information on a second stage sentence factor associated with the first sentence sentence factor,
Wherein the information on the first stage positive sentence includes weight information of each of the first stage positive sentence factors, the information on the second stage positive sentence factors includes weight information on each of the second stage sentence factors,
Wherein the second database comprises one or more of the second-stage sentence factors selected and logically combined to comply with the requirements laid down in the statute. 5. The method of claim 4, wherein the result data stored in the knowledge base comprises selection information for one of the legal texts of one or more of the legal texts, selection information of a type for two or more species determined in the legal text, Information on whether or not to file a complaint, information on whether or not to file a complaint, choice information on crime types prescribed by the Supreme Court sentencing standards, decision on crime types based on the Supreme Court sentencing standards, selection information on reduction / A sentencing expert system that includes either one. 6. The apparatus of claim 5,
A system control unit for controlling an operation of receiving the input information,
A sentence control unit for calculating the sentence by combining the input information with the information of the first database, the second database and the knowledge base,
The sentence expert system. 7. The method of claim 6, wherein the sentence control unit comprises: a sentence score calculated from the input information, the weight information corresponding to the first-step positive sentence factor stored in the first database, the weighted information corresponding to the second- The sentencing expert system determines the sentence by combining the input information with the sentence range determined by applying the information of the crime type selected by the sentence of the Supreme Court sentence stored in the knowledge base or the selected information of the sentence / basic weighted. [Claim 7] The method of claim 7, wherein the sentence control unit quantifies a weight corresponding to the first-stage sentence factor on the basis of the input information, and calculates weights from the weighted weights and weight information corresponding to the input information and the second- A sentencing expert system that calculates scores. The system control unit according to claim 7, wherein the system control unit provides the standard input information on the predetermined sentence to the sentence control unit according to a standard case, and compares the calculated sentence with the standard sentence determined in the standard case, Wherein the weighted information of the first stage positive sentence factor or the weighted information of the second stage sentence factor is adjusted in the database. 8. The method according to claim 7, wherein the sentence control unit applies the input information to the first database, the second database, and the knowledge base to select an associated legal statement, an act to select an applicable statement and penalty, Determines text information from an operation of determining whether the punishment will be weighted, an operation of determining whether the sentence is reduced by law, an operation of determining a range of the punishment type by law, and combines text information on the determined sentence to determine document information The sentence expert system generates the sentence. delete delete Receiving input information on crime and sentence factors;
A first database for storing sentence factor information corresponding to the type of crime, a second database for storing a correspondence relationship between the sentence sentence and a logical combination formula of one or more sentence factors, a condition expressed by a logical combination of sentence factors Applying the received sentence factor information to a knowledge base that stores result data on the data and how to apply the corresponding law, precedent, or Supreme Court sentencing standard;
Calculating a sentence score by referring to a result of applying the sentence factor information;
Calculating a range of punishment based on the sentencing of the Supreme Court with reference to a result of applying the sentencing factor information; And
Determining the sentence by combining the sentence score and the range of punishment
, &Lt; / RTI &
Wherein the sentence information includes information on a first stage sentence factor associated with the crime and information on a second stage sentence factor associated with the first sentence sentence factor,
The step of calculating the sentence score
Calculating quantized weight information for the first step positive sentence factor from the input information and the weight information for the first step positive sentence factor;
Calculating a sentence score by combining the input information, the weight information for the second-stage sentence factor and the quantified weight information,
Wherein the sentence expert system comprises: 14. The method of claim 13, wherein calculating the penalty range comprises combining the input information with information in the first database, the second database, and the knowledge base
Selecting the type of crime determined by the Supreme Court sentencing criteria;
Selecting a reduction / default / weight determined by the Supreme Court sentencing standard according to the selected type; and
Determining the scope of the punishment according to the type of the crime and whether or not the weighting and reduction reason is applied
Wherein the sentence expert system comprises: 14. The method according to claim 13, wherein the step of determining the sentence includes determining, as the sentence, any one of values falling within the scope of the sentence according to the sentence score. 14. The method of claim 13, further comprising: generating document information including the determined sentence according to the input information. The method of claim 16,
Applying the input information to the first database, the second database, and the knowledge base;
Determining textual information in the process of selecting legal texts related to the law;
Determining text information in a process of selecting a penalty to apply;
Determining text information in a process of determining whether punishment will be weighted according to whether or not it is leaked;
Determining text information in a process of deciding whether or not to be closed by law;
Determining textual information in a process of determining a range of the disposition according to the law; And
Determining textual information from the information about the determined sentence;
Generating document information by combining the text information
Wherein the sentence expert system comprises: 14. The method of claim 13, further comprising at least one of modifying, adding, or deleting weight information corresponding to the first step positive factor or weight information stored in the second database. 19. The method of claim 18,
Receiving standard input information about the sentence factors corresponding to a standard case and a standard sentence determined in the standard case;
Applying the standard input information to the first database, the second database, and the knowledge base;
Calculating a sentence score corresponding to the standard input information;
Calculating a scope of penalty based on the Supreme Court sentencing standard corresponding to the standard input information;
Determining the sentence by combining the calculated sentence score and the range of punishment; And
Updating the weight information corresponding to the first step-type positive factor or the weight information corresponding to the second-stage positive factor in comparison with the determined value and the standard value,
Wherein the sentence expert system comprises: 14. The method of claim 13,
Providing a crime selection screen; And
Providing an input screen for the sentence factors corresponding to the selected crime
Further comprising the steps &lt; RTI ID = 0.0 &gt; of: &lt; / RTI &gt;

Images