KR102491915B1 - System Providing Attorney Smart Matching Service - Google Patents

Abstract

The lawyer smart matching system according to the present invention,
A case information input module that receives case information including facts and litigation counterpart information from a client; Attorney database storing attorney information for a plurality of attorneys; A matching module for matching any one of the plurality of lawyers stored in the lawyer database with the client based on the input case information; Depending on the trial result, the inquiry request unit that receives the request for debt execution from the attorney and the credit information on the other party from the credit information management server according to the request for debt execution is provided to the attorney. Relative information browsing module including an information transfer unit to; and a collection information generation module for generating debt collection information for the other party to the lawsuit based on the credit information.

Description

Attorney Smart Matching System {System Providing Attorney Smart Matching Service}

The present invention matches the lawyer optimized for the client so that the client can be recommended a lawyer suitable for his or her lawsuit, and provides related information so that the lawyer can perform debt collection, so that the client can use the lawyer appointed by the lawyer. It is about a lawyer smart matching system that allows one-stop progress from lawsuits to debt collection.

When a legal problem arises for an ordinary person who is not a legal expert, it is not easy to solve the problem by yourself. Therefore, it is desirable to appoint a legal expert such as a lawyer (hereinafter, a lawyer) to proceed with the case, but it is not easy to select a lawyer suitable for the case among countless lawyers.

In addition, as society becomes complex and diversified, the work itself in charge of lawyers exists in a variety of cases, such as civil litigation cases, criminal litigation cases, litigation cases, and non-litigation cases, and each of these cases is also divided into various fields, so that the law Despite the need for the help of an attorney, clients have a difficult problem finding the attorney they need.

In particular, in the case of legal services, there is a problem in that a customer does not have many opportunities to find a lawyer suitable for his/her situation or to promote his or her specialty online from the standpoint of a lawyer.

As a result, most people appointed an attorney through their personal connections or by visiting an office close to their place of residence, so despite the lack of prior information about the attorney, the urgency of the legal work or lack of information led to the appointment of an inappropriate attorney. There were times when it was done.

In order to solve such a problem, in recent years, there have been many cases in which information is obtained in advance using the Internet and a lawyer is appointed, and some Internet service companies assign specialized lawyers for each field to provide Internet legal advice.

However, the reality is that Internet service providers simply stop at introducing lawyers, failing to provide information on the lawsuits requested by customers and suitable lawyers.

To improve this, Korean Patent Publication No. 10-2019-0023321 discloses a 'apparatus and method for providing legal services for lawyers'.

The invention is to provide a legal counseling service providing device and method so that lawyers can conveniently publicize their professional fields to the general public, and clients can receive legal advice or litigation information requested by them. A customer terminal that uploads legal consultation contents and related data corresponding thereto using the app, receives legal consultation contents and related data input from the customer terminal, stores them in a related data DB, and based on the received legal consultation contents to classify into one or more legal fields among case-specific legal fields that correspond to the legal categories pre-determined by the customer before writing the consultation article, provide the uploaded legal counseling content to multiple registered lawyers, request the lawyer and approve the customer When this is done, the related data will be shown to the lawyer in the form of a cloud, and at least one legal consultation written through the answer will be delivered to the customer terminal, and based on the related data, at the request of the customer, lawyer information that can accompany the prosecution or police investigation will be displayed. It is configured including a lawyer legal server provided to the customer.

However, in the case of the above invention, it is close to the function of recommending a lawyer to the client according to the case, and furthermore, the client has to prepare separately for separate collection or execution that may occur in civil lawsuits, so there is a problem in that the client's inconvenience follows.

Therefore, in order to solve the problems described above, in addition to recommending a lawyer suitable for the client, it is also possible to process the debt collection following the trial through a matched lawyer so that the client can be treated as a one-stop service. The need to develop a smart lawyer matching service so that various legal services can be provided is emerging.

The main object of the present invention is to provide a lawyer matching system that maximizes the convenience of the client by recommending an optimized lawyer to the client and further enabling the lawyer to provide additional services.

Another object of the present invention is to provide lawyers with debt collection services, but to provide lawyers with information necessary for debt collection.

Another object of the present invention is to add a counseling text analysis process for optimal matching between a client and an attorney.

A further object of the present invention is to calculate keywords in counseling text analysis and to enable matching between a client and a lawyer through this.

In order to achieve the above object, the lawyer smart matching system according to the present invention includes a case information input module for receiving case information including facts and litigation opponent information from a client; Attorney database storing attorney information for a plurality of attorneys; A matching module for matching any one of the plurality of lawyers stored in the lawyer database with the client based on the input case information; Depending on the trial result, the inquiry request unit that receives the request for debt execution from the attorney and the credit information on the other party from the credit information management server according to the request for debt execution is provided to the attorney. Relative information browsing module including an information transfer unit to; and a collection information generation module for generating debt collection information for the other party to the lawsuit based on the credit information.

Furthermore, the counterparty information viewing module includes a content-certification generating unit for generating content-certification information for the counterparty in response to the claim execution request, and the information delivery unit includes the credit information management server in response to the content-certification sending. It is characterized in that the credit information on the litigation opponent is received from and delivered to the lawyer.

In addition, the information delivery unit is characterized in that, in response to the bond execution request, the credit information and real estate information on the other party to the lawsuit are provided from the credit information management server and delivered to the lawyer.

In addition, the event input module receives further input of counseling text composed of text from the client, and the system analyzes the counseling text based on a keyword DB for storing a standard keyword that is a criterion for analysis, and the standard keyword. and a keyword calculation module including a keyword calculation unit that calculates at least one counseling keyword from the counseling text, wherein the matching module selects one of a plurality of lawyers based on the input case information and the counseling keyword. It is characterized by matching with the client.

The lawyer smart matching system according to the present invention,

1) Matching a lawyer optimized for the client's case or situation, fee conditions, etc., and providing one-stop legal services by allowing the lawyer to act as a proxy for debt collection that may be necessary depending on the trial result,

2) Enables debt collection by allowing lawyers in charge of debt collection to easily inquire and browse the debtor's credit information and real estate information owned by the debtor;

3) When matching a client and an attorney, the counseling content is input from the client so that the client's needs are reflected as much as possible, and the attorney is matched based on the input counseling content.

4) It has the effect of increasing analysis efficiency by enabling keyword extraction and filtering of keywords when analyzing the contents of consultation.

1 is a conceptual diagram showing a schematic configuration of the system of the present invention.
2 is a block diagram showing the overall configuration of the system of the present invention.
3 is a conceptual diagram showing an event information input screen according to the present invention;
Figure 4 is a conceptual diagram showing an example of a lawyer database of the present invention.
5 is a conceptual diagram illustrating a matching configuration of the present invention.
6 is a conceptual diagram illustrating a matching completion state guide according to the present invention;
7 is a conceptual diagram showing an example of holding real estate inquiry.
8 is a conceptual diagram illustrating an example of a collection progress information guide;
9 is a table showing a hierarchy for deriving a Euclidean distance;

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The accompanying drawings are not drawn to scale, and like reference numbers in each drawing indicate like elements.

1 is a conceptual diagram showing a schematic configuration of the system of the present invention.

Referring to FIG. 1, the lawyer smart matching system of the present invention consists of a client (1), a litigation opponent (2), a lawyer (3), and a central control server (4).

The client (1) is a person who appoints a lawyer (3) through the lawyer smart matching system of the present invention to proceed with a lawsuit (trial), and the client is a lawyer optimized for his or her lawsuit through the lawyer smart matching system of the present invention ( 3) can be easily recommended to represent the person in the lawsuit, and, as will be described later, cumbersome and additional information-requiring services such as debt collection can also be entrusted as a one-stop service.

The other party (2) is the other party in the lawsuit with the client (1). If the client (1) is a plaintiff, the other party (2) becomes the defendant, and if the client (1) is a defendant, the other party (2) ) becomes the manuscript. In the case of litigation, the basis is not to divide civil and criminal cases. Therefore, the client (1) appoints a lawyer (3) and the other party to proceed with the lawsuit becomes the litigation opponent (2).

The lawyer (3) represents the client (1) and proceeds with the lawsuit with the opponent (2), and in the lawyer smart matching system of the present invention, the lawyer (3)'s brief or history (of the trial win rate), etc. are based on the database. In addition, as lawyers (3) must be legally qualified, the lawyers (3) registered in the system of the present invention are based on their identity being verified.

Therefore, the lawyer (3) of the present invention can enter his/her profile on the lawyer smart matching system of the present invention to enable matching with the client (1), and when entering the profile, information about past trials as well as his or her profile It goes without saying that information, winning rate, etc. can be entered together.

The central control server 4 manages the database for lawyers 3 registered in the system of the present invention, and acts as a subject that actually performs the function of matching the lawyer 3 optimized for the client 1. carry out

The central control server 4 refers to hardware having a CPU and storage means for processing information. In other words, the central control server 4 includes a central processing unit (CPU) and storage means such as a memory and a hard disk. A program that can be executed in the central processing unit, that is, software, is installed on a hardware basis equipped with wired and wireless communication equipment such as Bluetooth, and this software can be executed. It will be described later as a constituent unit called 'Part'.

At this time, the central control server 4 temporarily and / or permanently stores the signals (or data) processed therein (RAM: Random Access Memory, not shown) and ROM (ROM: Read-Only Memory, not shown), may include a processor. In addition, the central control server 4 may be implemented in the form of a system on chip (SoC) including at least one of a graphics processing unit, RAM, and ROM.

The processor may include one or more cores (not shown) and a graphic processing unit (not shown) and/or a connection path (eg, a bus) for transmitting and receiving signals to and from other components.

The memory may store programs (one or more instructions) for executing and controlling modules or parts to be described later. Programs stored in the memory may be divided into a plurality of modules according to functions.

Steps of a method or algorithm described in connection with an embodiment of the present invention may be implemented directly in hardware, implemented in a software module executed by hardware, or implemented by a combination thereof. A software module includes RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), Flash Memory, hard disk, removable disk, CD-ROM, Alternatively, it may reside in any form of computer readable recording medium well known in the art to which the present invention pertains.

That is, the components of the present invention may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium. Components of the present invention may be implemented as software programming or software elements, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, such as C, C++ , Java (Java), can be implemented in a programming or scripting language such as assembler (assembler). Functional aspects may be implemented in an algorithm running on one or more processors.

The configuration of these 'modules' or 'units' or 'interfaces' is a configuration of hardware such as software or FPGA or ASIC executed via CPU and memory in a state installed and stored in the storage means of the central control server 4. it means. At this time, the configuration of 'module' or 'unit' or 'interface' is not limited to hardware, and may be configured to be in an addressable storage medium or configured to reproduce one or more processors. As an example, 'module' or 'part' or 'interface' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, and properties. , procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables.

Functions provided by these 'modules' or 'units' or 'interfaces' may be combined into a smaller number of components and 'units' or 'modules', or may be combined into additional components and 'units' or 'modules'. can be further separated.

Hereinafter, detailed configuration and functions thereof will be described based on the macroscopic configuration of the central control server 4.

2 is a block diagram showing the overall configuration of the system of the present invention.

Referring to FIG. 2, the lawyer smart matching system of the present invention includes a case information input module 100, a lawyer database 200, a matching module 300, a counterpart information viewing module 400, and a collection information generation module ( 500).

3 is a conceptual diagram showing an event information input screen according to the present invention.

Referring to FIG. 3 together with FIG. 2, the case information input module 100 plays a role of receiving case information including the facts of the case and the litigation opponent information from the above-described client 1. .

Here, the facts of the case may include information on litigation cases currently in progress, information on legal advice with the possibility of litigation, information on litigation cases in which litigation has already been confirmed, and precedent information. there is. Information on a litigation case may include information on a current litigation stage (eg, whether it is a first trial case or a second trial case), information on a lawsuit, and the like. Therefore, the relevant contents of the case, including the date and time of the event that caused the lawsuit, the date of the trial, whether or not the plaintiff/defendant, the outline of the case or the circumstances of the case, etc., are facts.

The information on the opponent of the lawsuit is information that includes the personal information of the opponent of the lawsuit (2). information that is the subject of the lawsuit. In other words, the factual relationship and litigation opponent information can be said to be all the litigation-related information necessary for the proceeding of the lawsuit. .

In addition, in the case information input module 100, it is also possible to receive information for debt collection from the client 1. To this end, the client 1 requests a lawyer 3 for a debt collection-related case, in case of a case to be collected. You can also upload the facts and bonds for the case as case information.

4 is a conceptual diagram showing an example of a lawyer database of the present invention.

Referring to FIG. 4 together with FIG. 2, the lawyer database 200 stores lawyer information for a plurality of lawyers 3, where the lawyer information refers to an existing lawsuit represented by the lawyer 3. Information about the lawyer (3), the history including the winning rate of the lawsuit, furthermore, the lawyer (3)'s profile, professional litigation field, etc. are included. In addition, in the case of lawyer (3), since it is a certified qualification, it includes whether or not the qualification is certified.

Therefore, when a new lawyer (3) is registered in the lawyer smart matching system of the present invention, the lawyer information of the lawyer (3) is newly entered and stored in the lawyer database (200).

In addition, since the number of cases currently in progress, fees, etc. may be different for each lawyer (3), the lawyer information includes information about this together, so that the client (1) can obtain information such as the number of ongoing lawsuits and fees of the lawyer (3). It also makes it easy to understand.

In addition, the lawyer database 200 of the present invention may be configured to include a winning percentage database in which the verification of the winning rate for each lawyer 3 is performed and stored in the stored lawyer information.

This winning rate database analyzes the past trial results and winning rates of each lawyer (3) included in the lawyer information, and determines whether the cause of the lawyer (3) winning the case in the last trial is the case itself or the lawyer (3) It analyzes and processes the ability recognition and stores it in the database. In other words, it is to analyze whether the acceptance itself was favorable to winning the case or whether it was the ability of the lawyer (3) that led to the victory.

In addition, if the acceptance itself was favorable to winning the case, the variable that won the case, that is, the cause, can be analyzed and stored together in the victory rate database. ) can be analyzed and stored in the win rate database.

In addition, since the system of the present invention includes information on various trials, it further includes the Supreme Court judgment database, the prevailing judgment database, the losing judgment database, and the original indictment database, etc. It is also possible to store and process the lost judgment and the original text of the indictment based on the criminal standard by making a database.

5 is a conceptual diagram illustrating a matching configuration of the present invention, and FIG. 6 is a conceptual diagram illustrating matching completion state guidance according to the present invention.

Referring to FIGS. 5 and 6 together with FIG. 2, the matching module 300 is based on the case information input from the client 1, any one of the plurality of lawyers 3 stored in the lawyer database 200 is matched with the client (1). At this time, the case information entered by default must match the professional litigation field of the lawyer (3), and if there is a fee range desired by the client (1), the lawyer (3) corresponding to the fee range must be matched. to be.

In addition, basically, the lawyer (3) and the client (1) are matched one-to-one, but after providing the client (1) with a list of multiple lawyers (3) that match the case information of the client (1), one of them Of course, it is also possible to perform matching processing between the client 1 and the lawyer 3 by receiving a selection from the client 1.

It is also possible to have the client 1 select a desired lawyer 3 in the matching configuration through the matching module 300. It is also possible to browse lawyer information and select a specific lawyer (3) to process matching.

In addition, through the matching module 300, at least one lawyer 3 that is optimal for the client 1 is extracted from the system, and the extracted lawyer information is provided to the client 1, and then the client 1 provides the lawyer. It is also possible to make a selection of any one of the information.

Therefore, after such matching processing is performed, the client (1) has the attorney (3) take over the case, and furthermore, procedures such as delegation are carried out and, if necessary, through additional consultation, the attorney (3) is responsible for the client (1). It is to be able to respond to the event information of

In the case of matching, it can be checked as shown in FIG. 6, and the details of the case, such as the time remaining until processing, the name of the client (1), the type of case, and the jurisdiction, as well as information on the bond or case are uploaded to the system. It can be guided to the client (1) and the matched lawyer (3).

When the trial is completed through such a process, according to the trial result, the client 1 becomes a creditor and proceeds with debt collection against the opponent 2 of the lawsuit. Therefore, the smart matching system of the present invention goes beyond the simple lawyer (3)-client (1) matching to provide a one-stop service so that the client (1) does not separately request debt collection to a separate company in relation to debt collection. For this purpose, a relative information viewing module 400 and a collection information generating module 500 are included.

The inquiry request unit 410 included in the counterpart information reading module 400 executes the claim against the client 1, that is, the debtor, from the lawyer 3 representing the client 1, that is, the creditor, according to the trial result. Performs the function of receiving input requests.

In general, when a client (1), who may be a plaintiff, or creditor, wins a trial and proceeds with debt collection against a debtor, or opponent (2), an attorney (3) representing the client (1) A is to enter a request for debt execution against the debtor, that is, the opponent of the lawsuit (2), for the debt collection agency.

Next, the information transmission unit 420 included in the counterpart information viewing module 400, as the request for bond execution for the litigation counterpart (2) is input from the lawyer (3) representing the client (1), the conventional It receives the credit information of the opponent (2) of the lawsuit from the credit information management server and delivers it to the creditor, that is, the lawyer (3) representing the client (1).

Here, the conventional credit information management server is a conventional server that plays a role of credit inquiry and credit management. It receives and manages credit information of general individuals in real time from financial institutions and public institutions, and if necessary, financial institutions and public institutions. It becomes the server of the institution providing information to the institution.

In the case of providing credit information, since it deals with sensitive matters, a power of attorney, etc., which proves that the lawyer (3) has been authorized to certify the attorney's qualifications and has been entrusted with the trial result of the case and the corresponding debt collection authority, is input from the lawyer (3). Of course, it is necessary to prepare in the system and provide it to the credit information management server in advance.

In addition, the credit information provided from the credit information management server includes financial transaction history information, other debt information, debt collection information, multiple debt information, etc. of the debtor, that is, the litigation counterpart (2). Therefore, this data is provided from the credit information management server and provided in real time to the creditor, that is, the lawyer (3) representing the client (1).

The collection information generation module 500 generates debt collection information for the corresponding litigation partner 2, that is, the debtor, based on the provided credit information. At this time, the debt collection information includes the above-mentioned credit information, the above-mentioned case information, and lawyer information, and the integrated debt collection information is provided to the lawyer 3 or encrypted and stored in a database separately prepared on the system of the present invention. It can be.

Therefore, through the counterpart information viewing module 400 and the collection information generation module 500, the credit information of the debtor, that is, the counterparty of the lawsuit (2) is provided, and debt collection information is integrated and generated, and the corresponding creditor, that is, the client (1). ) is provided to the attorney (3) to represent the attorney (3), enabling the attorney (3) to execute lawsuits and judgments as well as debt collection.

Therefore, the client (1) does not need to select a separate company or agent for debt collection after the lawsuit is completed, so the lawyer (3) matched for the lawsuit can be in charge of the execution. Non-professional clients (1) who have difficulty searching for information can more easily and conveniently receive and appoint a customized lawyer (3) for their lawsuit, and conveniently proceed with one-stop collection of their debts. can do.

In addition, through the system of the present invention, the lawyers (3) can not only perform credit information inquiry of the litigation partner (2), but also prepare a lawsuit request contract and power of attorney with the client (1) non-face-to-face. , Of course, it is possible to generate a report recording the results of the property investigation in the process of debt collection and automatically send it to the client (1).

Further here, in the above-described opponent information viewing module 400, the credit information of the debtor, that is, the opponent of the lawsuit (2) is inquired. Therefore, as much as the personal information of others is read, the litigation opponent (2) can be notified in advance whether or not to read the personal information.

To this end, the relative information browsing module 400 may further include a proof-of-contents generator 430. The proof-of-contents generator 430 executes bonds from the lawyer 3 representing the creditor, that is, the client 1. As a request is entered, it plays a role in generating content verification information for the counterparty in the lawsuit (2) and sending it to the counterparty in the lawsuit (2).

Here, the contents proof information includes information on the debts that the litigation partner (2), that is, the debtor has to pay to the client (1), that is, the creditor, and at the same time, the authority for debt collection is delegated to the lawyer (3). In addition, it informs that credit information inquiry is made accordingly, and refers to information including the facts about the receivable and the due date for repayment. Such proof-of-contents information can generally be generated in the form of a document including text, and the generated proof-of-contents information is sent to the litigation partner (2).

Therefore, when the proof of contents is generated and sent, the above-described information delivery unit 420 receives credit information about the other party to the lawsuit (2) from the credit information management server according to the sending of the proof of contents, and transfers the information to the lawyer (3). By providing it to you, you can prepare for actual debt collection.

7 is a conceptual diagram illustrating an example of holding real estate inquiry.

Referring to FIG. 7 , the above-described information delivery unit 420 transmits information from the credit information management server to the counterparty 2 in response to a bond execution request being input, or as a bond execution request is input and a proof of contents is sent. He said he was looking for credit information. However, going further here, it is also possible to receive not only credit information but also real estate information owned by the other party to the lawsuit (2) from the credit information management server and deliver it to the lawyer (3).

To this end, as shown in Figure 3, when the lawyer (3) enters the name, date of birth, gender, contact information, etc. of the litigation partner (2) into the credit information management server, the credit information management server is the litigation partner (2) It is possible to provide information on the real estate held as holding real estate information.

Such held real estate information includes ownership of real estate, some equity-shared real estate, ownership security right, tangible real estate, past owned real estate, and the like. Therefore, it is possible to inquire real estate for which auction real estate and mortgage rights have been set, and to check information on all types of real estate owned, such as land, buildings, villas, shops, officetels, and condominium shares, and to provide them to the lawyer (3) to secure bonds It can be made available for collection.

8 is a conceptual diagram illustrating an example of a collection progress information guide.

Referring to FIG. 8, when the system of the present invention performs matching, acceptance of a case, and agency of collection through the system of the present invention, reporting of property inquiry results to the client 1 and debt collection, Reports on the progress of the case should, of course, be made. Therefore, the report may be guided step by step through examples such as the drawings, and at each step, the client 1 may be informed of the progress through a separate phone call, e-mail, or text message.

In addition, as the lawyer smart matching system of the present invention basically starts with matching the lawyer (3) optimized for the client (1), the most suitable lawyer (3) for the client (1) can be recommended and matched. Of course there must be

To this end, the case information input module 100 included in the system of the present invention may further receive input of counseling text consisting of text from the client 1 . Here, the counseling text refers to a text input of a person's needs related to accepting a case, that is, the content to be consulted by the lawyer 3, and the length may basically be around 100 to 5,000 characters.

In addition, the consultation text may be input as text, but as much as the consultation with the lawyer (3) can be input over the wire, if the consultation text is input in the form of the client's (1) voice, it is recognized as speech (STT: Speech to Text) It can be converted into text using , and the converted counseling text can be input and processed into the system along with event information.

In addition, when the counseling text is input in this way, the input counseling text needs to be analyzed. can

In addition, when the consultation keyword is calculated, any one of the plurality of lawyers (3) stored in the lawyer database (200) and the client (1) are matched based on the case information input from the client (1) and the calculated consultation keyword By doing so, it is possible to enable optimized matching that meets the needs of the client (1).

Referring to the detailed configuration of the keyword calculation module 600, the keyword calculation module 600 may basically include a keyword DB 610 and a keyword calculation unit 620.

The keyword DB 610 stores standard keywords that are the basis for analysis, and words commonly used in litigation, for example, harm, fraud, damage, debt, reimbursement, alimony, etc., can be stored as standard keywords. , In addition, various adjectives that can represent the psychological state of the client 1 or various nouns for identifying facts can be stored as standard keywords.

In other words, there is no restriction on the standard keyword, and as a standard for keyword analysis, not only words commonly used in lawsuits or noun words that can represent facts, but also various adjectives that can represent emotional states are the standard keywords. It can be.

The keyword calculator 620 analyzes the counseling text input from the client 1 based on a plurality of standard keywords stored in the keyword DB 610, and calculates or extracts counseling keywords from the counseling text. This can be called a kind of filtering, and it can be calculated as a counseling keyword by filtering words that match the standard keyword in the counseling text.

Alternatively, it is also possible to analyze the counseling text based on the standard keywords, extract some keywords, and designate only some of the extracted keywords as counseling keywords. It may include a part 622, a normalization part 623, a distance calculation part 624, a proximity value calculation part 625, and a counseling keyword designation part 626.

The keyword calculation part 621 can be said to be the basic configuration of the above-described keyword calculation unit 620, and analyzes and processes the counseling text based on the standard keywords stored in the keyword DB 610 to calculate a plurality of basic keywords from the counseling text play a role Here, the basic keyword refers to all keywords obtained by filtering the counseling text based on the basic configuration of the above-described keyword calculation unit 620, that is, the standard keyword.

7 is a table showing a hierarchy for deriving a Euclidean distance.

If the layering part 622 is described with reference to FIG. 7 together with FIG. 2 described above, the layering part 622 performs a function of layering a plurality of generated basic keywords. Here, a basic keyword with a high hierarchy refers to something having a meaning that can encompass other basic keywords.

Therefore, in the example of FIG. 4 , the first layer means the highest layer, and as the second layer and the third layer gradually go down to the lower layer, it can be said to be a lower layer.

At this time, there is no limitation on the method of hierarchical classification, but basically, when a specific concept can encompass other concepts, it is regarded as a high hierarchy, and hierarchical classification can be used preferably according to the ontology concept. In this case, the ontology concept is a specification explicitly formalized to conceptualize a certain field of interest, and is a concept used for term classification.

For example, when there is a standard keyword called 'dispatch', the first layer may include loan, loan, and flexibility, and the second layer, which is a lower layer of the word 'loan', includes 1 financial loan, There may be 2nd financial loans, 3rd financial loans, private finance, etc. In addition, as the third layer, a lower layer of 'private finance', there may be a keyword called emergency money.

The normalization part 623 performs a function of normalizing a plurality of stratified basic keywords. The reason for performing the normalization is that even if the frequency of occurrence of a specific basic keyword is higher than that of other basic keywords, the basic keyword is unconditionally different from other basic keywords. Because it can't be concluded that it is more important.

Here, the normalization method of the basic keyword through the normalization part 623 is not limited, but most preferably, normalization can be performed through Equation 1 below.

Equation 1,

는

번째 기본키워드가 시스템 상에서 생성된 빈도수,

은 상기 시스템에 입력된 상담텍스트의 전체 개수,

는 시스템에 입력된 전체 상담텍스트 중

번째 기본키워드를 포함하는 상담텍스트의 개수,

는

번째 기본키워드의 정규화 값을 의미한다.here,

Is

The frequency of the th primary keyword generated on the system,

is the total number of counseling texts entered into the system,

Among the entire counseling text entered into the system,

The number of counseling texts including the th basic keyword,

Is

It means the normalized value of the th basic keyword.

This not only analyzes individual counseling texts, but also reflects the frequency of occurrence of individual basic keywords in all counseling texts entered into the system to identify basic keywords that frequently appear in counseling texts in general and reflect them in normalization. did

는 10가 되며, 시스템 상에 저장 처리된 상담텍스트의 전체 개수가 20개이고, 시스템 상에서 ‘급전’이라는 기본키워드가 생성된, 즉 ‘급전’이라는 기본키워드를 포함하는 상담텍스트의 총 개수가 5개인 경우

는 5가 된다.As an example, if the basic keyword 'dispatch' is generated 10 times in the system

is 10, the total number of consultation texts stored and processed in the system is 20, and the total number of consultation texts including the basic keyword 'dispatch' generated in the system is 5 Occation

becomes 5.

은 3이 되며, 시스템 상에 저장 처리된 상담텍스트의 전체 개수가 20개이고, 시스템 상에서 ‘대출’이라는 기본키워드가 생성된, 즉 ‘대출’이라는 기본키워드를 포함하는 상담텍스트의 총 개수가 2개인 경우

는 2가 된다.In addition, if the basic keyword 'Loan' is created 3 times in the system

is 3, the total number of counseling texts stored and processed in the system is 20, and the total number of counseling texts including the basic keyword 'loan' generated in the system is 2. Occation

becomes 2.

Therefore, in this case, if we calculate the normalized value of the basic keyword ‘dispatch’,

can be calculated as,

Calculating the normalized value of the basic keyword 'loan',

is calculated as

This normalization part 623 not only normalizes the frequency through an algorithm, which is not a simple reflection of the frequency of each basic keyword generation, so that more standardized values can be compared. There is an effect of calculating a value based on more objective normalized data by reflecting the number of counseling texts in which the corresponding basic keyword is generated, that is, including the corresponding basic keyword.

The distance calculation part 624 performs a function of calculating a Euclidean distance between normalized basic keywords. At this time, the Euclidean distance is used to calculate the distance between two nodes in an N-dimensional space, and at this time, the distance can be understood as a distance between terms, that is, a concept of similarity or proximity.

Preferably, the Euclidean distance may be calculated based on Equation 2 below.

Equation 2,

는

번째 기본키워드와

번째 기본키워드 사이의 유클리디안 거리,

는

번째 기본키워드가 시스템 상에서 생성된 빈도수,

는

번째 기본키워드의 정규화 값을 의미한다.here,

Is

with the second primary keyword

Euclidean distance between the th primary keywords,

Is

The frequency of the th primary keyword generated on the system,

Is

It means the normalized value of the th basic keyword.

이므로, 서로 다른 계층 사이의 비교에서는 상술한 수학식 3의 두 번째 경우와 같이 계층이 더 작은 기본키워드의 정규화 값 및 빈도수를 기반으로 유클리디안 거리가 산출되는 것이다.together

Therefore, in the comparison between different layers, the Euclidean distance is calculated based on the normalized value and frequency of the primary keyword with a smaller layer as in the second case of Equation 3 described above.

In this case, the layer to be compared at the Euclidean distance indicates the layer with reference to FIG. 4 . In the above-described embodiment, the first layer may include loans, loans, and loans, and the second layer, which is a lower layer of the word 'loan', includes first financial loans, second financial loans, third financial loans, private finance, and the like. There may be, and in addition, as a third tier, a lower tier of 'private finance', there may be a keyword called dispatch.

Therefore, when comparing the strata in the stratification example related to loans,

can become

인 경우를 적용한다. 여기서 상술한 실시예에서와 같이 급전의 빈도수는 3, 정규화 값은 3.3856이고, 대출의 빈도수는 10, 정규화 값은 3.6126이다.Therefore, if the Euclidean distance between 'loan' and 'dispatch' is calculated as an embodiment, since 'loan' is a higher layer than 'dispatch', in Equation 2

applies in the case of Here, as in the above-described embodiment, the frequency of feeding is 3, the normalization value is 3.3856, and the frequency of loaning is 10, the normalization value is 3.6126.

thus,

becomes

Here, if we assume that loans and payments are the same class,

becomes

Accordingly, the Euclidean distance may vary depending on the hierarchy, and it may be desirable to calculate the Euclidean distance according to the hierarchy in a state in which hierarchy classification has occurred.

The closer the Euclidean distance is to 1, the higher the distance between the corresponding basic keywords, that is, the higher the similarity. However, since the basic keywords are not the same, a value of 0 is not calculated. That is, a value exceeding 1 appears in the Euclidean distance, and the larger the distance (the larger the value of the Euclidean distance), the lower the degree of similarity.

Since the Euclidean distance calculated in this way means distance calculation according to a comparison value of normalized values in addition to simple frequency comparison, it can be easily used as a factor for calculating frequency-based similarity between corresponding basic keywords.

The proximity value calculation part 625 performs a function of calculating a proximity center value for each of the basic keywords according to the calculated Euclidean distance. In this case, the proximity center value is a value used to determine which basic keyword is at the center of a corresponding layer when the Euclidean distance between basic keywords for each layer is identified. At this time, there is no limitation on the calculation method of the proximity center value, but preferably, the proximity center value can be calculated based on the following Equation 3.

Equation 3,

는

번째 기본키워드에 대해 산출된 근접중심수치,

는

번째 기본키워드와 다른 기본키워드 사이에 산출된 유클리디안 거리의 최솟값,

는 산출된 기본키워드의 전체 개수를 의미한다.here,

Is

Proximity center value calculated for the th basic keyword,

Is

The minimum value of the Euclidean distance calculated between the th basic keyword and other basic keywords,

denotes the total number of calculated primary keywords.

Here, let's take an example of calculating the proximity center value for the basic keyword 'loan'.

At this time, if the minimum value of the Euclidean distance between ‘loan’ and other basic keywords is considered to be 2.4245 in the calculated Euclidean distance, and the number of all basic keywords calculated on the system is 30,

can be calculated as

Therefore, when the proximity center value is calculated in this way, a specific basic keyword may indicate the lowest proximity center value. Since this means that a specific primary keyword represents the closest distance (proximity and similarity) to other primary keywords, the closest distance to multiple primary keywords can be regarded as representing a plurality of primary keywords. It can.

The counseling keyword designation part 626 performs a function of designating at least one of a plurality of basic keywords as a counseling keyword according to the calculated proximity center value. Preferably, proximity center values are mutually compared, and at least one basic keyword is designated as a counseling keyword according to the height of the proximity center values.

Only one basic keyword with the lowest proximity center of gravity can be designated as a counseling keyword, or it is also possible to designate two or more basic keywords with low proximity center of gravity as counseling keywords by comparing and processing the proximity center of gravity in ascending order. .

However, since counseling keywords should include various contents, it is most preferable to filter some of the basic keywords and extract multiple counseling keywords rather than extracting one counseling keyword.

Furthermore, the keyword calculation module 600 of the present invention may extract the most important keyword, that is, a topic, from the calculated counseling keyword, and perform matching processing between the attorney 3 and the client 1 based on this. To this end, the keyword calculation module 600 may further include a topic calculator 630 that calculates a topic keyword from the calculated counseling keyword.

Here, the method of calculating the topic keyword through the topic calculation unit 630 is not limited, but most preferably, a latent semantic analysis is performed from the calculated counseling keyword, and the topic keyword is calculated through the result of the performed latent semantic analysis. can be calculated

Therefore, when the topic keyword is calculated in this way, one of the plurality of lawyers (3) is selected as the client (1 ) and can be matched.

At this time, the latent semantic analysis used to calculate the topic keyword in the topic calculation unit 630 is introduced to find the latent meaning within the counseling keyword. Therefore, latent semantic analysis can be used to search for potential topics within counseling keywords arranged in parallel. The topic calculation unit 630 to which this latent semantic analysis is applied includes a frequency matrix generation part 631, a transposed matrix generation part 632, a co-occurrence matrix calculation part 633, a singular value calculation part 634, and a final value calculation A part 635 and a keyword extraction part 636 are included.

The frequency matrix generating part 631 generates an appearance frequency matrix for the calculated counseling keywords. Here, the frequency matrix is a matrix representing all the calculated frequencies of each counseling keyword in individual counseling texts, and the frequency matrix can be generated through Equation 4 below.

Equation 4,

는 출현빈도행렬,

는

번째 상담키워드(

)가 시스템에 입력된

번째 상담텍스트(

)에 출현한 빈도수,

은 생성된 상담키워드의 개수,

은 시스템 상에 입력된 상담텍스트의 개수를 의미한다.)(here,

is the frequency matrix,

Is

2nd counseling keyword (

) entered into the system

2nd consultation text (

), the frequency of appearance in

is the number of generated counseling keywords,

means the number of counseling texts entered into the system.)

For example, as described above, let's assume that there are two counseling keywords, 'loan' and 'dispatch', and that there are two counseling texts entered into the system. Here, the counseling keyword “loan” appears twice in the first counseling text, the counseling keyword “emergency money” appears three times, and the counseling keyword “loan” appears once in the second counseling text, and the counseling keyword “emergency money” appears 0 times. Occation,

A frequency matrix is created in the form of

The transposed matrix generating part 632 serves to generate a transposed matrix by transposing the generated frequency matrix. Here, transposing means changing the rows and columns of the frequency matrix, and this transposed matrix is generated in the form of Equation 5 below.

Equation 5,

는 전치행렬,

는 전치행렬의

행

열의 값,

은 생성된 상담키워드의 개수,

은 시스템 상에 입력된 상담텍스트의 개수를 의미한다.)(here,

is the transposed matrix,

of the transposed matrix

line

the value of the column,

is the number of generated counseling keywords,

means the number of counseling texts entered into the system.)

Here, in order to generate the transposed matrix, the aforementioned frequency matrix must be generated, and the transposed matrix is a form in which rows and columns are exactly replaced in the frequency matrix.

That is, as described above, the frequency matrix A is

의 형태로 형성된다., then the transposed matrix B is

formed in the form of

That is, the rows and columns are the same here, that is, the values of row 1, column 1, row 2, column 2, and column 3, row 3 are the same, but the rest of the values are created in the form of swapping rows and columns.

The co-occurrence matrix calculation part 633 serves to generate a keyword co-occurrence matrix and a text co-occurrence matrix, respectively, based on the generated frequency matrix and transposed matrix.

Here, the keyword co-occurrence matrix and the text co-occurrence matrix are represented by the product of the frequency matrix and the transposition matrix, respectively. In the case of the keyword co-occurrence matrix, it has as many rows and columns as the number of generated counseling keywords, and in the case of the text co-occurrence matrix, the system It has as many rows and columns as the number of counseling texts entered in .

First, the keyword co-occurrence matrix is calculated through Equation 6 below.

Equation 6,

는 키워드 동시발생행렬,

는 키워드 동시발생행렬의

행

열의 값,

은 생성된 상담키워드의 개수를 의미한다.)(here,

is the keyword co-occurrence matrix,

is the keyword co-occurrence matrix

line

the value of the column,

means the number of generated counseling keywords.)

The keyword co-occurrence matrix is obtained by multiplying the frequency matrix by the transposed matrix, and has as many rows and columns as the number of generated counseling keywords.

,

라고 가정했을 때,In the above example, the frequency matrix

,

Assuming that

is calculated as

The diagonal matrix value in the keyword co-occurrence matrix generated in this way indicates how many times a specific counseling keyword appeared on the entire counseling text entered into the system, and the other values indicate how many times the above-mentioned two counseling keywords appeared at the same time. indicates cognition.

Next, the text co-occurrence matrix is calculated through Equation 7 below.

Equation 7,

는 텍스트 동시발생행렬의

행

열의 값,

은 시스템 상에 입력된 상담텍스트의 개수를 의미한다.)(here,

is the text co-occurrence matrix

line

the value of the column,

means the number of counseling texts entered into the system.)

The text co-occurrence matrix is a value obtained by multiplying the transposed matrix by the frequency of occurrence matrix, and has as many rows and columns as the number of counseling texts entered into the system.

,

라고 가정했을 때,In the above example, the frequency matrix

,

Assuming that

is calculated as

The diagonal matrix values in the text co-occurrence matrix thus generated represent the total values in which counseling keywords appear on the entire counseling text.

The singular value calculation part 634 performs a function of calculating keyword singular values and text singular values based on the calculated keyword co-occurrence matrix and mountain text co-occurrence matrix. In calculating the singular value here, the eigenvalue of the corresponding matrix is basically used.

Preferably, the keyword singular value is calculated through Equation 8 below, and the text singular value is calculated through Equation 9 below.

Equation 8,

Equation 9,

는 키워드 특이값,

는 텍스트 특이값,

는 키워드 동시발생행렬의 고유값 중 최대값,

는 텍스트 동시발생행렬의 고유값 중 최대값을 의미한다.)(here,

is the keyword singular value,

is the text singular value,

is the maximum value of the eigenvalues of the keyword co-occurrence matrix,

means the maximum value among the eigenvalues of the text co-occurrence matrix.)

Therefore, in order to calculate the singular value, first, each unique value of the keyword co-occurrence matrix and the text co-occurrence matrix must be calculated. This can be calculated through conventional mathematical knowledge.

Keyword co-occurrence matrix

일 때,text co-occurrence matrix

when,

Calculating the eigenvalues of the keyword co-occurrence matrix,

의 해를 구하면 되므로,

Since you need to find the year of

Therefore, the maximum value of the eigenvalues of the keyword co-occurrence matrix is 13.16,

The keyword singular value is 3.63, which is the square root of that value.

Calculating the eigenvalues of the text co-occurrence matrix,

의 해를 구하면 되므로,

Since you need to find the year of

Therefore, the maximum value of the eigenvalues of the text co-occurrence matrix is 13.32,

The text singular value is 3.65, the square root of that value.

Based on the singular value calculated in this way, the dimension of the counseling keyword is compressed, and based on the compressed dimension, a topic keyword, which can be called a topic among counseling keywords, is extracted from the counseling keyword.

Therefore, the final value calculation part 635 calculates a final value based on the keyword singular value and the text singular value calculated for topic keyword extraction. may be calculated, or the final value may be calculated based on the average value of the keyword singular value and the text singular value.

Therefore, the final value calculation method is not limited, and in general, in calculating the final value, the final value is calculated by adding the text singular value and the keyword singular value.

Finally, the keyword extraction part 636 extracts the topic keyword from the counseling keyword according to the height of the calculated final value. At this time, in extracting the topic keywords from the counseling keywords, preferably, the higher the final value, the higher the number of topic keywords extracted from the counseling keywords. The lower the final value, the lower the number of topic keywords extracted from the counseling keywords. can

Therefore, according to the calculated final value, the number of topic keywords extracted from counseling keywords is more or less determined, and accordingly, at the level of the system manager. Alternatively, the system can automatically extract topic keywords from counseling keywords.

As described so far, the configuration and operation of the lawyer smart matching system according to the present invention is expressed in the above description and drawings, but this is only an example and the spirit of the present invention is not limited to the above description and drawings, and the present invention Of course, various changes and changes are possible within the range that does not deviate from the technical idea of .

1: Client 2: Litigation Opponent
3: lawyer 4: central control server
100: case information input module 200: lawyer database
300: Matching module 400: Relative information viewing module
410: inquiry request unit 420: information delivery unit
430: content verification generation unit 500: collection information generation module
600: Keyword calculation module 610: Keyword DB
620: keyword calculation unit 621: keyword calculation part
622: stratification part 623: normalization part
624: distance calculation part 625: proximity value calculation part
626: counseling keyword designation part 630: topic calculation unit
631: frequency matrix generation part 632: transpose matrix generation part
633: co-occurrence matrix calculation part 634: singular value calculation part
635: final value calculation part 636: keyword extraction part

Claims (9)

As a lawyer smart matching system,
A case information input module that receives a counseling text consisting of text and case information including facts and litigation counterpart information from a client;
A keyword calculation module including a keyword DB for storing a standard keyword that is a standard for analysis, and a keyword calculation unit that analyzes the counseling text based on the standard keyword and calculates at least one counseling keyword from the counseling text;
Attorney database storing attorney information for a plurality of attorneys;
a matching module for matching one of the plurality of lawyers stored in the lawyer database with the client based on the input case information and the consultation text;
According to the trial result, the inquiry request unit for receiving a request for debt execution against the other party in the lawsuit from the lawyer, and receiving credit information for the other party in the lawsuit from the credit information management server in accordance with the request for debt execution, and forwarding it to the lawyer Relative information reading module including an information delivery unit;
A collection information generation module for generating debt collection information for the counterparty of the lawsuit based on the credit information;
The keyword calculation unit,
A keyword calculation part that analyzes the counseling text based on the reference keyword and calculates a plurality of basic keywords from the counseling text, a stratification part that stratifies the basic keyword, and normalization of the plurality of stratified basic keywords A normalization part that calculates a normalized Euclidean distance between the basic keywords, a distance calculation part that calculates a Euclidean distance, and a proximity value calculation part that calculates a proximity center value for each of the basic keywords according to the calculated Euclidean distance; A counseling keyword designation part for designating at least one of the basic keywords as a counseling keyword according to the calculated proximity center value;
The normalization part,
Normalization is performed based on Equation 1 below,
The Euclidean distance,
It is calculated based on the following Equation 2,
The proximity center value,
Characterized in that the calculation is based on the following Equation 3, lawyer smart matching system.
Equation 1,

(here,

Is

The frequency of the th primary keyword generated on the system,

is the total number of counseling texts entered into the system,

Among the entire counseling text entered into the system,

The number of counseling texts including the th basic keyword,

Is

Normalized value of the th primary keyword)
Equation 2,

(here,

Is

with the second primary keyword

Euclidean distance between the th primary keywords,

Is

The frequency of the th primary keyword generated on the system,

Is

Normalization value of the th primary keyword,

Is

layer value of the th basic keyword)
Equation 3,

(here,

Is

Proximity center value calculated for the th basic keyword,

Is

The minimum value of the Euclidean distance calculated between the th basic keyword and other basic keywords,

is the total number of calculated primary keywords) According to claim 1,
The relative information viewing module,
In response to the claim execution request, a content verification generating unit for generating content proof information for the other party in the lawsuit and sending it to the other party in the lawsuit;
The information delivery unit,
The lawyer smart matching system, characterized in that receiving credit information on the opponent of the lawsuit from the credit information management server according to the sending of the content verification information and delivering it to the lawyer. According to claim 1,
The information delivery unit,
According to the bond execution request, the credit information and real estate information of the other party to the lawsuit are provided from the credit information management server and delivered to the lawyer. According to claim 1,
The keyword calculation module,
It includes a topic calculation unit that calculates a topic keyword by performing a latent semantic analysis on the calculated counseling keyword,
The matching module,
Based on the input case information, the counseling keyword, and the topic keyword, any one of a plurality of lawyers is matched with the client, and the lawyer smart matching system. According to claim 4,
The topic calculation unit,
A frequency matrix generation part for generating an appearance frequency matrix for the calculated counseling keyword;
A transposed matrix generating part for generating a transposed matrix obtained by transposing the frequency matrix;
A co-occurrence matrix calculation part for generating a keyword co-occurrence matrix and a text co-occurrence matrix based on the frequency matrix and the transposition matrix;
a singular value calculation part for calculating keyword singular values and text singular values based on the keyword co-occurrence matrix and the text co-occurrence matrix;
A final value calculation part for calculating a final value based on the calculated keyword singular value and the text singular value;
Characterized in that it comprises a keyword extraction part for extracting the topic keyword from the counseling keyword according to the high and low of the final value, lawyer smart matching system. delete delete delete delete

Images